Professional Documents
Culture Documents
Most organizations are production systems – not-for-profit and service organizations also have
production systems.
Most organizations convert a set of inputs like equipment, labor, and material into outputs such as air
conditioners, electricity, and health care services.
Production system can be defined in terms of environment, inputs, transformation system, outputs,
and mechanism used for monitoring and control – these are subsystems of the production system.
There is interaction and interdependence among these subsystems.
System view focuses on how the individual subsystems that make up a production system interact
among themselves. It emphasizes the relationships between subsystems of the production system –
for example, if a company invests in new machines, quality of its products improves.
Suboptimization occurs when one part of the system is improved to the detriment of other parts of
the system, and perhaps the organization as a whole – for example, productivity increases when a
company uses low-quality components and materials to make its products, but quality of its products
suffers. This happens when the relationship between subsystems is ignored.
Boundary of the system determines what a decision maker will or will not consider since things outside
the system boundary are considered to be part of the environment and beyond decision maker’s
control – for example, if cost of materials increases, cost of products will also increase, but product
managers cannot do anything about it, but if a competitor installs a new production system, and hence
improves quality of products, it can do the same and maintain its competitiveness. Important
relationships among components may be omitted if the boundary is defined too narrowly, whereas
complexity and cost increase when the boundary is extended – for example, quality of inputs affect
quality of products, and howsoever good a company’s production processes may be, it will not be able
to produce good quality products if its suppliers continue to supply bad quality components.
1. Environment: Environment includes those things that are outside the actual production
system, but influence it in some way. Environment exerts a great deal of influence on
production system, but it is beyond the control of decision makers. For examples, a competitor
develops new process technologies to make products of higher quality at lesser cost, and the
company must respond by investing in its process technologies, but if banks increase their
rates it cannot do anything about it, though it costs go up.
3. Transformation system: This part of the production system adds value to inputs. This can be
done in following ways:
Alter: Something can be changed structurally: there would be physical change in the
raw materials / components – for example, two components are welded together or
a component is painted. Sensual alteration is very important in service industry –
sensual alteration happens when air conditioner is switched on in a restaurant, and it
increase the value of the meal the guests are having. Psychological alteration can
have value such as feeling of worth from obtaining a masters degree, or feeling of
importance when someone opens the door for a guest entering a restaurant.
Transport: An entity may have more value if it is located somewhere other than where
it is currently located – for example, when a component is assembled into a product,
it becomes much more valuable than when it was lying in the store, or when a
bouquet is delivered to a guest, it becomes more valuable than when it was lying in
the store.
Store: The value of an entity may be enhanced if it is kept in a protected environment
for some period of time – for example, value of a fixed deposit certificate increases in
value with time.
Inspection: An entity may be more valued when its properties are understood better
– for example, when a product is inspected and certified as OK, it becomes much more
valuable that when it had not been certified as OK. Similarly, when jewelry is
appraised and certified, customers are willing to buy them at higher prices.
4. Output: Two types of outputs commonly result from a production system – products and
services. Products are physical goods, whereas services are abstract or non-physical. But,
difference remains ambiguous in cases like hotel industry – for example, to guests in a
restaurant, quality of meals may be as important as demeanor of employees serving them.
Products are easy to comprehend, but service has to be understood either as an exclusive
offering from a company that is primarily intangible, or as a part of the service-product mix
that a company offers. Services are bundle of benefits, some of which may be tangible and
others intangible, and they may be accompanied by facilitating goods.
5. Monitoring and control: Consistency is a big problem in operations – no two products are of
exactly the same dimension, and no process variable is of same value at any two times. A
product’s specified length may be 10.5 mm, but no two products will be of exactly 10.5 mm,
and a machine may be required to run at 100 rpm, but at no two times will it be running at
100 rpm.
Monitoring process must tell when significant changes are occurring and apply corrective
measures to alter inputs / transformation system, and thereby the outputs – for example, if
the machine is not running at 100 rpm, it must be stopped and rectified.
Customer needs, underlying technologies of products, and process technologies are constantly
changing, and hence it is necessary to continuously monitor the production system and take action
when the production system is not meeting its goals like those of achieving certain level of quality at
a certain cost. It may also be that current goals are not appropriate, indicating a need to revise the
goals, or the goals may be fine, but the inputs or the transformation system or both have to be
changed. For example, reducing cost may not be an appropriate goal when customers are demanding
high quality products or new process technologies may have emerged which the company has not
incorporated in its production system or new materials may have been synthesized which the
production system is not using in its products.
Operations management is a broad discipline that consists of many different tasks. Two major tasks
are (i) designing the production system and (ii) operating and controlling the production system. The
production system is designed once in many years, and is mostly the responsibility of senior managers.
Operation managers are mainly responsible for controlling and operating the production system –
scheduling of equipments and manpower, quality control, maintenance, and purchase of materials
are its typical tasks.
Product design
The design of product largely determines the design and operation of production system. The design
of the product will mandate the processes that have to be employed to produce the product, which
in turn will determine the equipments and skills that will be required. The design of the product will
mandate the materials that have to used to produce the product, which in turn will determine the
kind of suppliers that have to be contracted to supply the required materials. Therefore, product
designers must consider how their designs will be produced when they are designing the product –
for example, will its production require complex processes and hence sophisticated equipments or
will it be made of expensive materials or will it be produced only by highly skilled workers. Product
designers should not be wary of redesigning their product to make it producible – difficult-to-produce
products will have high manufacturing costs, and the market may not be able to bear the
corresponding high prices.
The design of the production process will determine the extent of customization, costs, and quality
of product. Two extreme choices of production processes are (i) Line layout and (ii) Job shop layout.
In line layout, the sequence of operations in the product determines the sequence of equipments –
for example, if welding is the first operation that has to be carried out on the product, then welding
machine will be the first equipment in the line layout, and so on. All products have to be processed on
all the equipments in the same sequence. An automobile assembly line is a typical example of a line
layout. Line layout can produce only standard products, but in large quantities and at low costs. In job
shop layout, similar equipments are placed together in departments – for example, all welding
machines are at one place in welding department. Products move to various equipments depending
on whether they are required to be processed on that equipment – for example, product X may go to
welding shop while product Y may not go to welding shop. A hospital or a tool room is a typical
example of a job shop layout. Large variety of products can be produced, but only in limited numbers
and at a higher cost.
Capacity planning
Capacity is the ability to produce in quantitative terms – for example, how many cars can be produced
in a day. Capacity should be linked to demand, but while demand changes frequently and
incrementally, capacity can be added only in larger quantities, and it almost always takes time to install
capacity. There will always be some gap between demand and capacity – sometimes demand may be
more than capacity in which case short term strategies like overtime may be adopted, and sometimes
capacity may be more than demand in which case counterintuitive strategies like producing a
competitor’s products may have to be adopted. Capacity can be kept more than demand if per unit
profit is high, and capacity can be kept less than demand if production cost has high fixed cost
component.
Facility location
A manufacturing facility may be near its competitors – it gets access to competitors’ suppliers and
employees if it is near its competitors. In most cases, manufacturing facilities are located where the
costs of factors of production are minimum. A service facility would not like to be near its competitors
– its competitors will take away its customers if it is near its competitors. But if it is a strong service
brand, it can locate near its competitors – the place having all major providers of a service will attract
large number of customers, but it will take most of them due it being a strong brand. In most cases,
service facilities are located near customers.
Job design
A job is sum of tasks, and a job holder should like all the tasks of his job. A job can be designed in
primarily two ways: (i) A job consists of a single task or very few tasks. The job holder specializes in
the small number of tasks that have been assigned to him, and he enjoys doing them over and over
again – he is not bored. Such narrow jobs are common at both high and low end of job spectrum –
workers working on assembly lines carry out only a few tasks, and super specialized knowledge
workers like doctors again carry out only a few tasks. (ii) A job consists of many tasks. Such broad jobs
are of two types: (a) managers typically have broad jobs – they are expected to deal with variety of
issues, though they may not specialize in any one of them. (b) Workers are given large number of tasks
under the pretext that they would get bored doing a few tasks over and over again – they set up their
machines, bring material to their machine, carry out the required operation, carry out inspection, and
fill a report. Another aspect of job design is the degree of control that workers have on factors that
affect their performance in their jobs – can they influence the purchase of the equipments that they
will operate or do they have a role in deciding the process they will use to produce their assigned
component? A worker is said to be empowered if has significant control over factors that affect his
work.
Quality management
From being a competitive advantage, quality has become a hygiene factor – customers will not even
consider a company’s products if they do not match with their expectations of what a quality product
should be. Though customers’ expectations are shaped by what they pay for a product and how the
company’s competitors are faring on the quality frontier, it is a irrefutable fact that customers are
more finicky about quality today than any time in industrial history. Quality has multifarious links: (i)
Does the design of the product deliver the benefits that the customers expect the product to deliver?
(ii) Can the installed production processes faithfully achieve the dimensions such as length and width
as prescribed by product design? (iii) Are the equipments capable of achieving the tolerances as
prescribed by product design – for example, length of rods should be between 10.4 and 10.6
millimeters with target length of 10.5 millimeters (iv) Are workers suitably trained? (v) Are equipments
properly maintained – is there a preventive maintenance schedule in place? (vi) Do suppliers
understand the company’s philosophy, and do they have the technologies, equipments, and the skills
to meet the company’s expectations on quality? Since quality has such multifarious links, it is
important that each employee is able to understand as to how his job impacts the quality of products
/ services that their company is producing. Every employee is responsible for quality, and it is not the
responsibility of a few people in the company – this is the fundamental principle of Total Quality
Management.
Materials management
Materials should be available as and when, and where they are required, and in the number they are
required – no more than they are required , no earlier than they are required, and no where other
than where they are required. For example, if 100 gear shafts are required on assembly line at 8 am,
material should not have arrived the previous night, and more of them should not be lying in stores.
Inventory should not be held at all, and if it is held at all, its purpose should be clearly defined – for
example, a company requires 100 gear shafts, but orders 110 of them because it is not sure if all 100
of them would be of good quality. Therefore, the company maintains inventory to cover up for bad
quality, and it would not have maintained inventory if it would have been sure of all the products
being of right quality. Similarly, companies maintain inventory because they expect some of their
equipments to fail or they expect their suppliers to falter on their commitments or they have set up
times or lead times. Therefore, companies can reduce their inventories only when they have (i) robust
quality system (ii) rigorous preventive maintenance schedule (iii) competent supplier base and (iv)
small set up or lead times – these are the pillars of Just-in-time production system.
Scheduling
Scheduling allocates available resources like equipments and workers to components and customers
– for example, customer X would be served by service representative A, and component Y would be
manufactured on equipment B. Scheduling is done in ways so that customers / components do not
have to wait to be served / processed, and workers and equipments are not idle – good scheduling
results in low waiting times for customers and components, and less idle times for equipments and
workers.
1. Product is a tangible output that can be touched, held, and stored – examples are cars, air
conditioners, and burgers. Service is an intangible output that satisfy some needs of
customers. It enhances customers’ life in some way, but it cannot be touched or stored –
examples are health care services and legal assistance. Some services also enhance usefulness
of products – for example, after sale service and delivery. Most companies provide mixture of
product and service – for example, restaurants produce the tangible product of food along
with intangible services such as pleasant ambience.
2. Production systems that produce products are called manufacturing systems, and the
production of products is called manufacturing. Production systems that produce services are
called service systems.
3. There is very little customer contact during production of product. There is direct customer
contact during provision of service.
4. There is no customer participation during production of product. Most customers for products
have no contact with the production system. In contrast, there is frequent customer
participation during provision of service. Customers may be required to carry out part of the
process as is the case in quick-service restaurants. In other cases, customers have to
collaborate with service providers to enable the latter to provide good service as is the case
in medical care – for example, patients who are able to explain their medical conditions clearly
are likely to receive better treatment than those who are not able to do so.
5. In manufacturing, division of labor is widely employed, and each worker does only a small part
of the total task of manufacturing a product. Tasks are repetitive – a worker carries out the
same task each time the product is produced. Division of labor is not so widely employed in
service systems, and an employee may provide the entire service all by himself. Also,
customers’ requirement may vary even for the same basic service, and employees of service
systems must be flexible and creative.
6. Response times vary. A manufacturer can take days or even weeks to meet customer demand,
but a service provider must be able to provide service within minutes of a customer arriving
in his facility. Service providers also have difficulty in matching capacity with demand as
customers usually arrive at times of their choosing – most of them may choose to come at the
same time, as they do when most of them decide to watch evening show on a weekend. Also,
arrival patterns may fluctuate daily or even hourly, creating even more short-term demand
uncertainty – for example, most people like to dine at around 9 pm rather than at 7 pm or 11
pm, and hence restaurants face maximum rush at 9 pm, while they are relatively empty at 7
pm and 11 pm..
8. Customer needs and corresponding quality standards are defined by customers themselves,
and hence are difficult to identify and measure – each customer may define them differently.
Also, services may have to be customized for individual customers – for example, doctors and
lawyers must customize their services for individual customers.
9. Bad quality products can be recalled and replaced, whereas bad quality service can be rectified
by reparations and apologies – this process of compensating and appeasing customers is
known as service recovery.
10. Production of products can be separated from the customer in space and time more easily
than can the production of services. A manufacturer can produce at a constant rate, and keep
its products in inventory when rate of demand is less than rate of production, and serve
customers from inventory when rate of demand is more than rate of production. In contrast,
service is produced and consumed simultaneously.
11. Product is consumed over time – for example, a customer uses a car for a few years. Service
is consumed immediately – for example, a customer enjoys the hospitality of a resort only till
he is there.
12. Products can be stored for later use and transported over space before use – Products are
manufactured at few locations, and are transported to customer locations. There is time gap
between when a product is produced and when it is used. Service cannot be stored – there is
simultaneous production and consumption of service.
13. Manufacturing facilities often serve regional, national, or even international markets, and
hence they require larger facilities, more automation, and more capital investment. And since
most services cannot be transported to where customers are, larger number of smaller service
facilities have to be established.
14. Less labor and more equipment is used in creation of product. Since a product can be
produced at a few places, and transported to customer locations, equipments of high capacity
can be employed to produce them. More labor, and less equipment is used in creation of
service. Hence, behavior and morale of service providers play very important roles in provision
of services – for example, hospitals and banks have found out that behavior of their employees
is a strong determinant of customer satisfaction.
15. Some services like air transport are equipment based, while other services like hair styling are
people based.
16. Manufacturing systems use more raw material inputs than service systems – products are
made of materials and components, while services are rendered mostly by acts of service
personnel.
17. Sophisticated methods of measurements are used to measure outputs of conversion process
of product. Elementary methods of measurement are used to measure outputs of conversion
process of service – customer satisfaction measures are often used to measure effectiveness
of service delivery system.
18. In service operations, it is difficult to distinguish output and throughput. Output is a generated
service. Throughput is an item going through the process. In a hospital, output is the medical
service to the patient, who by going through the conversion process is also the throughput.
At a restaurant, customer does not go through the conversion process. The output is food and
throughput is raw materials used in its preparation.
PRODUCT SERVICE
Product can be stored for later use and Service is produced and consumed
transported over space before use. simultaneously
Bad quality product can be recalled and Bad quality service can be rectified by
replaced reparations and apologies
Quality is assessed against design specifications
Quality is assessed against customer
preferences and expectations
There is no customer participation during There is frequent customer participation
production of product. during provision of service.
PRODUCT DESIGN
Companies launch new products to serve customers’ expressed needs. Companies also launch new
products when they believe that they have the technology to serve customers’ latent needs. New-to-
the-world products like mobile phones are successful when new technologies are appropriately
applied to serve customers’ latent needs. Ideas for new products can come from customers,
marketers, R & D department, and production and engineering departments.
A company’s ability to design and launch products will determine its competitiveness. Product life
cycle is shortening – customers’ changing preferences and emergence of new technologies make
existing products redundant. Therefore, companies need to design and launch products quickly as well
as efficiently: A product may become outdated by the time it is launched because its underlying
technologies and / or customers’ preferences change as the product is being designed – therefore,
the time lag between conceptualization of a product idea and its launch must be compressed. And
since the company launches many more products than it did earlier, it must expend less resources in
each launch than it did earlier – its product development process must be efficient.
Design cannot be the sole prerogative of designers – a design must be produced easily and
inexpensively. Personnel from production must give feedback on the producibility of the design before
it is finalized so that modifications can be avoided – when a design it modified to make it producible,
it may compromise on some of the customer needs that it was originally designed to serve. Suppliers
give feedback on availability of components and materials – a supplier may already be producing a
component similar to the one that the designers are conceptualizing. Technologists must be actively
involved in the design process – they must regularly update the design team on the product’s
underlying technologies. Similarly, marketers must regularly update the design team on customers’
preferences. And then social scientists must keep the designers updated on how people’s needs are
changing due to changes in the social and economic environment.
Design is as important for services as it is for products. Since service is intricately linked to its delivery
system, design of service will involve simultaneous design of its delivery system. New technologies
enable a service to be delivered differently than it was done earlier – delivery system changes, but the
service remains the same. For example, people have been transferring money to each others’ accounts
through cheques, but now it can be done through online banking.
Need identification
It must be demonstrated that the new product will serve some customer needs, and also that existing
products do not already serve the same customer needs.
Generation of ideas
Because of their proximity to customers, marketers can identify customer needs that the company
can serve by designing and launching new products. It is very important that marketers spend time
with technologists and designers of the company, so that they understand as to what type of customer
needs the company can serve. Unfulfilled customer needs are always in plenty, but marketers will
make positive contribution to the product development process only when they are able to pluck
customer needs that the company is capable of serving.
Designers know what customer needs the company can serve well. They can come up with product
ideas based on the technologies and other competencies that the company may possess. If designers
spend enough time with customers, they would instinctively know which of these product ideas would
serve their needs.
A company may follow second-to-market strategy, and may choose to imitate a proven new idea or it
may purchase someone else’s invention. Imitation gives an opportunity to study any possible defects
in original product and rapidly develop a better design, often at a better price. Purchasing an idea or
an invention eliminates the risks inherent in research, but it still requires the company to develop and
market the product.
Competitors are also a source of ideas for new products. A company can use perceptual maps,
benchmarking, and reverse engineering to learn from their competitors.
Perceptual map: The company compares its product with those of its competitors.
Benchmarking: The company compares its product with the best product in the market.
Reverse engineering: The company dismantles and inspects a competitor’s product to see if
it can incorporate some of its benefits and features into its own product.
Market and economic analysis: Customer surveys are carried out to find out if the proposed
product would serve the customer needs that it is meant to serve. Customers are also asked
whether the proposed product would meet their needs as regards its price, performance,
quality, and reliability. The company also tries to assess if the proposed product will have
enough demand to justify investments in its development, production, and launch – will it
earn enough profits?
The company tries to assess if the proposed product is similar to an existing product, and
hence will cannibalize its sale. It also tries to assess as to how the proposed product fits in the
company’s product portfolio: Will it increase the sale of other products or will it reduce the
sale of other products? Will it enhance the overall reputation of its product portfolio or will it
damage the overall reputation of its product portfolio?
The company tries to assess if the proposed product can be produced easily and inexpensively.
It also tries to assess if the proposed product can be produced in the existing production
facility or a new production facility has to be installed to produce it. And, that if a new
production facility has to be installed, will it generate enough sales to justify the investments
in it?
The company tries to identify existing products that would compete with the proposed
product, and tries to gauge the probable reactions of the manufacturers’ of these products to
the launch of the proposed product – will they reduce price or will they increase advertising
intensity or will they load more features and benefits in their products?
It also determines the possibility of patenting the product as well as possibility of patent
infringement.
It is important to remember that there might be clash of views between business and technical
personnel – good technical ideas are often adjudged to have insufficient market potential, and
hence are not pursued.
Financial analysis: The company calculates the investment that would be required to design,
produce, and market the product. It also tries to determine per unit profit of the product. It
also calculates payback period, net present value, and return on investment to evaluate the
financial viability of the project. It carries out breakeven analysis to determine the number of
units that it would have to sell before it starts earning profits.
Analysis of Product Life Cycle: The company tries to determine the length of each phase of
the product life cycle of the proposed product. This is important, because if introduction phase
of product is long compared to its maturity phase, the company might not be able to recover
its investments.
The idea is developed into several product concepts. A product concept is a particular combination of
features, benefits and price. Alternate product concepts are evaluated by customers. Though it may
still be a description rather than the actual product, customers have something tangible to react to.
This process allows customer feedback to seep into the new product development process early
enough for designers to evaluate the degree of acceptance of the proposed product. As the physical
product may not be available at this stage, companies go in for a verbal or pictorial description of the
product to let customers have an idea about the actual product. Prospective customers present
feedbacks regarding the attractiveness of the features and benefits offered by the proposed product.
Usually, the intention of the company is to gauge the most desirable combination of benefits and
features that customers are willing to pay for.
Engineering design
The company zeroes on a product concept that it will design, produce, and market. Therefore, the
company has decided what features the product would have, what benefits would it provide, and at
what price would it sell. Design engineers now design the components that would constitute the
product so that it has the required features and gives the desired benefits. Technical expertise of the
company comes to the fore during this stage of the product development process. Large number of
different types of technical specialists may be required to collaborate for long periods of time to come
up with specifications of components.
Though design engineers dominate this phase of product development process, operations managers
should provide design engineers with information about the producibility of the components being
designed. Operations managers should be included in decisions about equipment design, since they
know more about equipments than do design engineers. If the product requires a new material that
has special properties, materials managers and suppliers should be consulted.
Design engineers must be keenly aware that the product must be designed for functionality, reliability,
maintainability, and producibility.
Design for functionality – The product should serve the customer needs for which the product
is being designed.
Design for reliability – The product should function adequately over a defined period of time.
Design for maintainability – The maintenance processes should not be cumbersome, time-
consuming, and expensive.
Design for producibility – The product should be easy to manufacture.
Prototype testing
The design is converted into a prototype. The prototype may take the form of physical model, a
computer simulation, or a real product. Tests are conducted and the design may be accepted,
modified, or rejected.
Final design
The design is fine tuned to accommodate the results of prototype testing, and any changes in
product’s underlying technologies and / or customer needs that might have taken place while the
product was being designed.
The detailed product design is now available. The focus shifts to designing production processes,
installing equipments, and enrolling suppliers.
The company produces the required units of the product, and makes them available to its channel
partners like distributors and retailers.
Customers’ feedback is tracked, and failure data is analyzed. Formal research is conducted to
understand customers’ experience with the product. The company keeps itself updated on product’s
underlying technologies and process technologies used to produce the product. Product is redesigned
if customers’ feedback are persistently negative, or if new technological breakthroughs enable design
and production of a better product.
Customers are trained to use the product. It is important that customers use all the features
of the product, and get all the benefits that the product is capable of providing. Customers
will not rate the product highly if they do not get all the benefits that the product is capable
of providing just because they cannot use them well. For example, less tech-savvy customers
use their smart phones just to make calls and send messages.
Customers are provided with warranty and repair services. The company should provide
comprehensive warranty and prompt repair services for those components of the product which are
most likely to fail
DESIGN PRINCIPLES
Design principles guide the designers. Following design principles are useful:
Common component: A common component is used by many products of the company – for
example, an automobile company may use the same steering wheel in four of its brands of
cars. Designers should always evaluate if a component that is already being used in some
other product of the company can be a good substitute for the component that they are
contemplating designing. Since the component is already being used, its product and process
technologies are understood, and the company can work to improve its quality and reduce its
cost by using the budget that was reserved for designing and developing the component. And
since a common component is produced in large volumes, it can reap the benefits of
economies of scale.
Simple assembly process: Components should be easy to stack – if components are designed
by separate teams, the teams should meet to discuss how the components would be stacked
in the product, and if some specifications of some components have to be changed to make
their stacking easier, they should always be done. Component designers should always
remember that the assembly process will be carried thousands / millions / billions of times
depending on the success of the product, and hence any amount of time they can save on the
assembly time will have huge impact on the profitability of the product during its entire life.
Ease of assembly is probably as important as quality or cost of components, but since
components are designed by separate teams, ease of assembly mostly falls through the
cracks. The project leader must ensure that component designers meet to discuss how the
assembly process can be made simple.
There are four areas of concern: (i) Designers should ensure what is being called mistake-proof
assembly – the idea is that designers should anticipate what can go wrong in the assembly
process, and instead of relying on intuitiveness and steadfastness of operators to avoid such
mistakes, they should design the assembly process in such a way that those mistakes cannot
occur. For example, if two components look similar, and there is a fear that an operator may
insert a component in the wrong place, the components should be of different colors, and
their colors should match with the color of the place where it is to be inserted – for example,
blue component will go in the blue colored slot and red component will go in the red colored
slot. (ii) Tests have to be carried out during the assembly process, and also after the product
has been assembled. Designers have to specify locations in the assembly process where the
tests would be carried out, but more critical issue is to decide the actions that would be taken
if a test fails – if a failed test requires lot of de-stacking and extensive rectification of
components, the assembly process would be slowed down. Therefore, designers should
simplify the process of de-stacking and rectification when a test fails – the process should take
as little time as possible. And there should be no final test of the product since it might involve
extensive de-stacking and rectification, which is wasteful. The tests should be planned in a
way that the so called final test is the just the last test in the series of tests that have to be
carried out – no more elaborate than the other tests, involving no larger de-stacking and
rectification than other tests. (iii) Stacking of components should not require much effort –
stacking should be from top, several components should not have to be held together at one
point, and once a component has been stacked it should not have to be removed or adjusted
to stack another component. (iv) Large numbers of components have to be fastened together
during the assembly process, hence fasteners should be convenient to use – clamps should be
used as often as possible, and small screws of high quality should be used rather than long
screws which takes lot of time in fastening.
Reasonable specifications: Designers are always tempted to use the best – best raw material
and best component should be used. Most designers over-specify – for example, a lap top
should not be able to withstand a fall from The Mount Everest, and a class room should not
be expected to survive a raging wire. Most products are not normally used in such extreme
circumstances, and hence they should not be designed for such usages. Over-specification has
cost, and it can escalate if designers are not questioned frequently if a lower specification
would serve the functionality desired by customers. Designers should always bear in mind that
specifications do not have a life of their own – they are there to provide the functionalities
that customers desire.
Reasonable tolerance: It is known that lower tolerance results in better quality than higher
tolerance – a product with a target specification of 10 cm is of higher quality if it is produced
within 9.9 – 10.1 cm than if it is produced within 9.8 – 10.2 cm. But producing products of
lower tolerance is expensive – it requires more sophisticated equipments, and chances of
rejects are higher. Designers are tempted to specify lower tolerances because lower
tolerances assure them of high quality, but tolerance should be linked to functionality of the
product. Designers should be encouraged to specify larger tolerances as long as such
tolerances do not compromise the functionality of the product.
CONCURRENT DESIGN
Companies have been following what is now famously called ‘over-the-wall’ design process. Designers
would work on a design till they were satisfied that their design would meet customers’ functional
and aesthetic needs, and then they would pass on the design to production department. It never
bothered the designers if their designs could be produced, or how difficult or easy, or how expensive
or inexpensive would it be to produce it. The production personnel would often discover that they
have to buy new equipments and tools to be able to manufacture the product, and that its
manufacturing would involve complex processes and expensive materials. Since all these factors
would raise the cost of manufacturing of the product, production personnel would send back the
product to the designers asking them to redesign the product. Designers set down to redesign the
product with explicit intention of making it producible. And such redesigning can become dangerous.
In their obsession to make the product producible, they may make changes that may compromise the
functionality and aesthetics of the product – the key reasons why customers would buy the product.
Therefore, designers may end up designing a producible product, but it may be a product that
customers would not like to buy.
Producible design: The idea is to motivate the designer to consider how the product would
be manufactured as he is designing the product. He needs to understand that his best design
would no serve no purpose if it cannot be brought to the market fast, and at a price that the
customers find acceptable – the designer has to accept the responsibility for producibility of
the product.
Concurrent design: The idea is to design and test the production process while the product is
being designed. As designers release specifications, engineers set up production process to
check if they are easily producible. Engineers also determine if they have to buy new tools
and equipments to be able to manufacture the product. If some specifications have to be
changed to make the product producible, engineers sit with designers to explain the matter
to them. Designers have been often reluctant to release specifications of their modules early
in the fear that they would have make changes in them either to accommodate specifications
of other modules or to accommodate new finding from the market or to incorporate new
technologies that might have emerged. It is important that designers are made to understand
that these are very valid reasons for making modifications in specifications, and that having
to make modifications in one’s design is indication of their lower designing acumen – they are
not sufficiently aware of what their fellow designers are doing and how their customers’
needs and the product’s underlying technologies are evolving. Concurrent design can be
successful only if designers release their specifications early. It is important to clarify that
designers still control the design process – if designers insist that certain specifications cannot
be changed because they would compromise the product’s functionality and aesthetics, they
cannot be bulldozed into changing those specifications to make the product producible.
Design team: Design team should include personnel from manufacturing, purchasing,
marketing, and finance. It should also have a senior manager among its midst. The idea is to
prevent the designers from designing extravagant products which would be difficult to
manufacture and sell. Personnel from manufacturing, purchasing, and finance keep
reminding the designers that the product design has to satisfy criteria other than functionality
and aesthetics, and they are available to give the designers any feedback that they would
require on their designs – purchaser would immediately tell them the cost of the material
that they might be contemplating using, and the process engineer would immediately tell
them if the specification that they were thinking of finalizing could be inexpensively
produced. The senior manager would keep reminding designers that that the product design
be in line with company’s strategy. Such a team structure is not really an intrusion into the
work of the designers – their work is facilitated as they have people to guide them in
processes like manufacturing and purchasing, which they do not understand very well, but
which are critical for the success of the product.
Supplier involvement: Many companies are outsourcing extensively, and many suppliers
have developed sophisticated competencies in many product areas – they may know more
about latest product and process technologies in their area of competence than the buyer.
Companies that are outsourcing manufacturing of components are losing the technological
grip of the components whose manufacturing they are outsourcing since they do not have
engineers in these product areas – it is futile to believe that companies can keep excelling in
designing of components that they have been outsourcing, because they simply do not have
the technological bandwidth to do so. Therefore, it is best to involve suppliers early in the
design process, and let them design components that they are ultimately going to
manufacture. But, suppliers cannot be left alone to design components since they need to
understand the philosophy and rationale behind designing the product. Therefore, suppliers
have to be intrinsic part of the design team – they are outsiders to the company, but for all
practical purposes, they belong to the company. Suppliers are also helpful in another way –
since they are also working for other companies, they may suggest components that they
may already have designed for another company. If the design is acceptable to the company,
the company is saved the worry and cost of designing the product. In addition, the supplier
will not have to acquire new equipments and tools to be able to manufacture the product,
and since he already has some experience in manufacturing the component, the cost of
manufacturing would be lower than what it would be if he manufactured a completely new
component.
Customer involvement: Marketers take feedback from customers, and pass them on to
designers – it is too passive a approach to be successful. Also, distortions take place as
customers’ needs travel to designers through marketers. To make a successful product,
designers have to understand customers’ needs in all its glorious sophistication and nuances,
and for this to happen, designers have to be in first-hand contact with customers. But the
idea has to move beyond designers asking their needs – representative and interested
customers have to be part of the design team. They have to be part of the deliberations when
designers make tradeoff between different levels of functionality, because highest levels
cannot be provided for all functionalities – customers will make the best tradeoff because
they know the precise level of each functionality that they want in the product.
MODULAR DESIGN
A product is separated into modules or subsystems. For example, a pen consists of three modules –
body, cover, and refill. The modules are interchangeable. For example, if a company makes five
varieties each of body, cover, and refill, any of the five covers can fit with any of the five bodies, and
any one of the five riffles can fit into any of the five bodies. A total of 5 3 = 125 varieties of pen can be
made by designing and manufacturing 5 x 3 = 15 components. If 125 varieties of pen had to be made
in the traditional way, 125 varieties each of body, cover, and refill have to be designed and
manufactured – 125 x 3 = 375 components have to be designed and manufactured. Modular design
offers special type of customization – a customized final product is made of standard components. A
customer can select any one type of cover, body, and refill, and get a pen of his choice assembled.
Modular design will work in products which can be made of separate and interchangeable
components, and when range of performance and price of a component is significant. Modular design
will work if the price of the refill, body, and cover ranges from Rs. 10 to 100 – customers will express
themselves by their choice of modules. A sincere student may buy a refill of Rs. 100, but he may buy
cover and body of Rs. 10 each. In contrast, someone who wants to show off will buy cover and body
of Rs. 100 each. But it is unlikely to work if price ranges merely from Rs. 10 to Rs. 20 – there would
merely be two or three segments.
In sum:
The products are discrete such as automobiles, and all the products go through all the workstations in
the same sequence, and discrete tasks like welding are carried out at each workstation on each
product as it passes through each workstation. An assembly line is a typical example of a line flow
process.
Line Flow
There is linear sequence of operations to make the product. Sequence of machines in a line
flow depends on the sequence of operations that have to be carried on the product.
Product must be standardized. Line flow produces narrow range of standardized products.
Line flow is inflexible beyond the narrow range of products that it is designed to produce.
Product design is stable – a new line has to be installed for producing a new design.
Individual tasks are closely coupled, and hence should be balanced so that one task does not
delay the next task – task times of individual workstations should be as close to each other as
possible. If task times vary, products wait at some workstations to get processed while some
workstations remain idle – efficiency of a line flow is highest when all workstations have the
same task time.
Line processes are very efficient. Labor is mostly substituted by machinery. Total task is
divided into large number of small tasks. Specialized equipments are employed to carry out
these tasks. A worker carries out the task assigned to him repeatedly, and hence becomes
very proficient at carrying it out.
Workers perform a narrow range of tasks on only a few product designs. Skill requirements
are not high, and workers can manage with minimal training and supervision.
Specialized equipments are used. These equipments are set up once to perform a specific task
for a long period of time on one product. To change over these machines to a new product
design requires great expense and long down times as the machines have to be set up to
produce the new product design.
Large volume of output has to be maintained to recover the cost of specialized equipments.
Therefore, standard products are manufactured on line flow, because their demand is
relatively stable.
Line processes are inflexible due to sequential organization of equipments and work force,
though some flexibility is being introduced by computerized control of equipments and
reduction of changeover / set up times.
Mass production refers to assembly line type of process flow as is used in automobile industry,
whereas Continuous production refers to process industry as is used in the chemical industry.
Equipments are specialized since overall production process is decomposed into narrow
tasks, and specialized equipments can perform these tasks very efficiently. Investment in
specialized equipments are justified because they perform their assigned tasks on large
number of products.
Jobs are specialized since each worker performs only a narrow range of tasks. Workers
learn fast, and they achieve peak productivity much earlier than if they had to perform a
wide range of tasks – this is the benefit of performing a task repeatedly. Less skilled
workers can carry out these tasks, and hence labor cost is low. Workers can be easily cross-
trained to perform tasks of different workstations – cross-trained workers bring in
flexibility in labor deployment which can be used to handle absenteeism, and it eases the
boredom that is caused by a worker doing a narrow range of tasks repeatedly.
Material handling is efficient as all products flow in the same direction and in the same
sequence. Fixed material handling equipments like conveyers, pipes, and gravity slides are
used.
Work-in-process inventory is small – one product each is processed at each workstation
all the time. More than one product is kept at each workstation only when it is expected
that defective products would be produced or that equipments can breakdown. It is also
known as to how long a product will have to be wait to be processed at each workstation,
and for how long will each workstation be idle – waiting time and idle time are fixed for
each workstation.
There is efficient utilization of space as work-in-process inventory is small, and also
because fixed material handling equipments like conveyers and pipes do not take much
space – the wide aisles that are required to run forklifts and other mobile material
handling equipments are not required.
Quality conformance is easy to achieve because (i) workers become skilled at performing
the narrow range of tasks assigned to them – doing a task repeatedly improves workers’
skills (iii) workers understand the small range of quality requirements (ii) there are less
changeovers and hence less frequent set ups – bad quality is often the result of wrong set
up.
Scheduling is simple, because products have to be scheduled only on the first workstation
– once a product is placed on the first workstation, it moves from one workstation to the
next workstation automatically.
Cost is easy to monitor because all products are processed in the same way, and hence
consume the same amount of resources.
Line flow process is inflexible – only products that require the same processing in the same
sequence can be produced in a line flow. If a product requires different processing or the same
processing in different sequence, the physical configuration of the line flow has to be changed
– new equipments have to be installed or their arrangement has to changed. Once a line flow
has been set, it is expensive and time-consuming to change its physical configuration.
Changeover may take months even when new models of the same product has to be produced
– for example, automobile industry.
Line flow process is inflexible in terms of volume – in continuous flow process, processing
must occur at a specific rate for the product to be made correctly. Line flow process that
produces discrete products can run at lower speeds, and hence can produce smaller volumes,
but since no other product can be produced, per unit cost of production is very high when a
line flow produces below its capacity.
It is expensive to install a line flow process – specialized equipments are expensive. It takes
time to decompose the overall production process into tasks and allocate the tasks to
workstations – it should also be ensured that workstations have balanced workloads.
Work on a line flow is boring – workers have to perform narrow range of tasks repeatedly.
Bored workers make more mistakes which leads to low quality and low productivity. Cross-
training and job rotation allow workers to do different jobs, and hence ease some tediousness.
Line flow is extremely vulnerable to unplanned stoppages – if one equipment breaks down,
the line flow is stopped, and if a workstation produces a defective product, work stops on
downstream workstations till a fresh product arrives.
A company installs a job shop process when it produces products that require different types of
processing, and / or when products are processed in different sequences – For example, product X
requires to be processed on equipments A, B, and C in the sequence B – A – C while product Y requires
to be processed on equipments A, C, and D in the sequence D – A – C. Hospital is a classical example
of a job shop process and so is a tool room of an automobile company. If such a company installs
equipments as a flow process configured to produce one product, it would be expensive and
cumbersome to change the arrangement of equipments each time a different product has to be
produced. Also, specialized equipments of a line flow process configured to make one product cannot
be used to make another product.
Equipments performing similar tasks are placed in one department – for example, all welding
machines are placed together in welding department. Equipments are general-purpose. Spatial layout
and material handling equipments are so designed that products can move throughout the system in
various ways. Therefore, a job shop is divided into departments, and one department carries out one
type of task such as welding, painting, or scanning. Products move among the departments in any
sequence – depending on the tasks that have to be performed on them, products go to some
departments and skip the rest – for example, product X goes to departments B, C, and D while product
Y goes to departments A and D. A job shop is flexible – it can produce small quantities of large variety
of products.
Job shop is used when small batches of many different products are to be produced – for
example, 10 units of product A + 5 units of product B + 15 units of product C, and so on.
Each product is custom designed.
Each product requires an unique set of processing steps – for example, product A: X – Y – Z,
product B: Y – Z, Product C: Z – X – Y
Large number of diverse processes have to be performed – for example, welding, painting,
heat treatment, and so on.
Production is done in batches at intermittent intervals – for example, 10 units of product X is
produced, and then 15 units of product Y is produced.
All equipments that carry out a process are put at one place – for example, all furnaces are in
Heat Treatment shop.
A product will flow only to those work centers on which it has to be processed, and will skip
the rest – for example, product A will go department X and then to department Y (will skip
department Z), whereas product B will go to department Z and then to department X (will skip
department Y).
General purpose equipments and highly skilled labor are used, which enable flexibility in
processing different products. These equipments can be changed over rapidly to carry out
new operations on different product designs. Workers must adapt quickly to the multitude of
operations to be performed on each unique batch of products being produced. These workers
must be highly skilled, and require intensive job instructions and technical supervision.
The materials-handling equipment generally consists of forklift trucks and other mobile
vehicles that allow for the variety of paths followed through the facility by the products
produced.
The jumbled flow pattern and product variety leads to problems in managing inventory,
scheduling, and quality.
Proper scheduling is required to ensure equal loading on equipments – it often happens that
while some equipments have long waiting lines, others run idle.
There is job interference when different products require the same equipment / labor at the
same time, which results in high in-process inventory – products wait on workstations to get
processed. Completion time increases, which leads to inefficiency.
A product may wait on a workstation as some other product may be being processed. In fact,
some of the equipments may have long queues of products waiting to be processed. In
contrast, some equipments may be idle – it may not have any product to process. And it is not
that same equipments are always overloaded or idle. Depending on the type of products that
arrive to be processed, a equipment may be idle now, but may be overloaded a few hours
later. But, when an equipment is consistently overloaded, it means that another equipment
has to be installed, and if an equipment is idle for long periods of time, it means that its
capacity is higher than required.
Throughput time = Process time + Waiting time
Throughput efficiency = Process time / ( Process time + Waiting time)
Throughput efficiency of job shop is rarely more than 40 per cent, whereas throughput
efficiency can be 90 per cent for line flow.
Job shop is justified when product lacks standardization or its volume is low. Job shop is used
in early life cycle of all products, for customized products, and for products with low demand.
The startup cost of producing a new product in a small quantity is much lower in a job shop
process than in a line flow process, because many other products are also being produced in
the same job shop – if the new product is successful and is required to be produced in large
numbers, a line flow process may be installed to produce it.
Job shop / Intermittent flow is also called Process Layout as similar equipments and work skills
are grouped together. Line flow is also called Product Layout as equipments and labor skills
are arranged in sequence according to the way the product is made.
Job shop is extremely flexible – products can move in any sequence from one department to
another, and they can have different task times in different departments. For example, a
company has four work centers: A, B, C, and D. Product X is processed at work center C for 10
minutes first, is then processed at work center A for 12 minutes, and is finally processed at
work center D for 8 minutes. Product Y is processed at work center B for 20 minutes first, is
then processed at work center C for 14 minutes, is then processed at work center A for 10
minutes, and is finally processed at work center D for 16 minutes.
A job shop can handle different batch sizes – product X is produced in batch size of 20, product
Y is produced in batch size of 300, while product Z may be a customized product.
The breakdown of one equipment need not halt an entire process, because work can be
transferred to other equipments in the department.
It is less expensive to install a job shop process – general purpose equipments are much less
expensive than specialized equipments.
Workers perform a variety of tasks, and hence are less likely to get bored.
Scheduling equipments is a huge challenge. At any one point in time, there may be large
number of products with different processing requirements and different dispatch / due dates
waiting to be processed on equipments, some of which may be idle and some of which may
be heavily loaded. Scheduling a job shop is difficult because (i) each product has unique
processing requirements in unique sequence (ii) lot sizes vary from a few units to hundreds of
units (iii) processing time at every equipment is unique for every product (iv) dispatch / due
dates are different for different products (vi) it is difficult to forecast what products would be
required to be processed in any future time period.
General purpose equipments are less efficient – specialized equipments can process more
material than general purpose equipments in a given time.
Only skilled workers can set up and operate general purpose equipments. Since different
products are made on an equipment, set ups are more frequent – wrong set up will result in
bad quality. Therefore, jobs shops spend considerably on hiring and training of skilled workers.
Flexible material handling equipments like forklifts and hand trucks are used, but they are less
efficient than fixed location material handling equipments like conveyers and gravity slides.
Work-in-progress inventory is very large since products wait to be processed when
equipments are engaged in processing another product – 100 units of product X may arrive
on work center when it has just started processing a batch of 400 units of product y, and hence
product X cannot be processed till all the 400 units of product Y have been processed. Also,
since products do not move continuously from one work center to another, work-in-process
inventory is maintained at work centers so that they do not run idle when the work centers
that feed them are being set up. Flow of work to a workstation cannot be predicted since it
depends on the process requirements of the products that are entering the job shop,
therefore in-process inventory is maintained at each workstation so that each workstation has
enough work.
Quality conformance is difficult because (i) there are more product changeovers, and
defective products are most likely to be produced just after a new set up (ii) workers work on
different products, and hence they do not become experts in processing any particular
product (iii) workers work on many products whose quality requirements are different, and
hence workers take time to understand them.
It is difficult to find out the cost of each product as large number of products with different
processing requirements are being produced – hence, it is difficult to find out the profitability
of each product.
Batch flow process is chosen for standard products with volumes insufficient to dedicate line flows to
them.
Batch in line flow: Flexibility is introduced in the line flow as more than one product is
produced, but the line has to be set up every time a different product has to be produced, and
therefore there is loss of production time. Therefore, the line flow becomes more flexible, but
less efficient. The line flow retains its other advantages and disadvantages.
Batch in a job shop flow: No set up time is incurred when products of a batch are being
processed on a work center, and therefore the job shop has more available production time
– larger is the batch size, less is the per unit set up time, and hence larger is the available
production time. If the job shop processes large number of large batches, work centers may
not be left with enough capacity to process the large number of different products that the
job shop is meant to process. Therefore, the job shop becomes more efficient, but less flexible.
The job shop retains its other advantages and disadvantages.
CELLULAR PROCESS
Cellular process is hybrid of job shop and line flow. The basic production environment remains that of
a job shop, but a few specific equipments are taken out from their respective departments / work
centers and are arranged in a line flow. For example, a job shop has four work centers – A, B, C, and
D. The job shop produces large number of products, and it is found that a large number of these
products are processed on work centers B and D in the sequence D – B, and therefore one equipment
each of B and D is taken out from their respective work centers and are arranged in the line flow D –
B. The products which were being processed on work centers B and D in the sequence D – B now come
to the line flow D – B when it is their turn to go to work center D. These small line flows are called
cells.
A cellular process is as flexible as job shop because it is still a job shop, and other than the ones in the
small line flows, the equipments are still grouped together in departments / work centers. A cellular
process is more efficient than a job shop, because once a product comes to the line flow, less time is
incurred in set ups and waiting than if it had gone to work centers.
A company divides its products into groups of products that require to be processed on a set of
equipments in the same sequence, and a small line flow or cell is created for each group of products.
For example, one group of products requires to be processed on equipments B, D, A in the sequence
B – D – A, and hence cell B – D – A is created for this group of product while another group of products
requires to be processed on equipments C and A in the sequence C – A, and hence cell C – A is created
for this group of products. It is important to always bear in mind that the basic production
environment remains that of a job shop, and hence after being processed in a cell, a group of products
may be send to another worker center to get processed, while another group of product may emerge
as finished products from its cell, while some products may not visit any cell at all.
Set up times are reduced because products that are processed at a cell are similar, and hence
less set up times are incurred when changeovers occur between products of the same group.
Material handling is reduced, and materials travel lesser distances as equipments of a cell are
placed close to each other.
Throughput time is reduced – products move out faster from the job shop because set up
times, waiting times, and travel times are less in a cell.
In-process inventory is less because the cell operates as a line flow, and a product moves from
one equipment to another in a single unit. Also, because set up times are smaller, large in-
process inventories do not have to kept at equipments to keep them running while their
feeder equipments are being set up.
Less space is required as equipments of a cell are placed together, and also because in-process
inventory is less.
Equipments are mostly pulled out of departments / work centers and arranged in cell, and
hence new equipments do not have to be purchased to create cells. Utilization of equipments
improves because less set up times are incurred, and hence the same set of equipments is
able to produce larger number of products. Requirements of fixtures and tools are reduced
since similar products are processed in a cell.
Quality conformance is easier as similar products are processed in a cell, and hence their set
ups are similar – and hence there are fewer wrong set ups, and hence fewer defective
products are produced due to wrong set ups.
It is not easy to identify groups of products which are required to be processed on the same
set of equipments in the same sequence – a job shop may handle thousands of products, and
the process requirements of all the products have to be analyzed to arrive at such groups of
products.
Equipments have to be withdrawn from departments / work centers and space has to be made
available for installing cells.
Cells operate as line flows, and hence separate managers must be appointed to manage cells
– managers who manage job shops have different skills and attitude from managers who
manage line flows.
PROJECT PROCESS
A company may produce unique products such as flyovers or software programs. Products that cannot
be physically moved once completed use a project process. Resources like materials, equipments, and
people are brought to the site where the product is to be built. These resources are allocated for the
duration of the project, and are reallocated once the project is over – resources are allocated to
complete specific tasks of the project, and are reallocated once those tasks have been completed. The
company uses similar equipments and skills for all the products that it makes, but each product is
made in a particular way – skills and equipments are combined in unique ways to make a product.
Coordination is a huge challenge since large number of skills and equipments may be required to
produce a product and a company may be simultaneously working on large number of products –
required skills and equipments must be made available to a project when it is required, and any delay
in providing them will result in the project getting delayed. Another challenge is to complete a project
on time and within budget.
Project process is dictated by the product itself – project process is used when each product is unique
and requires a unique combination of activities.
It deals with one-of-a-kind products that are tailored to unique requirements of each
customer. It is used to produce an unique product such as a building or a movie.
Each unit of product is produced as a single item. Products are not standardized.
There is no product flow, but there is sequence of operations. All operations should be
sequenced to contribute to the completion of the project.
There is precedence among tasks required to complete the project. It is important to plan,
sequence, and control the individual tasks so that the project is completed on time.
Since each project is unique, it involves lot of creativity. Hence, it is difficult to define how a
project will ultimately turn out to be.
There is less automation, but general purpose machines are used.
Flexibility in equipment capabilities, human skills, and procedures are required.
Project process cannot exploit economies of scale and productivity gains from doing a process
repeatedly.
Flexible manufacturing system (FMS) is an automated job shop process – it is a self-contained system
of programmable computer-controlled machines, a tool changing system, and a material handling
system. The simplest flexible manufacturing system is a CNC (computer numerically controlled)
machine with an automatic tool changer, a rack to hold tools, and an automatic loading and unloading
system. The most popular flexible manufacturing systems contain (i) two or more CNC machines that
can change tools automatically (ii) a tool storage and changing system (iii) automatic loading system
to feed materials into the CNC machines (iv) a material handling system that moves materials between
the CNC machines, and to and fro from the system (v) an unloading and stacking system (vi) a central
computer that controls the movement of materials, the changing of tools, and the operation of the
CNC machines. Many flexible manufacturing systems can be connected to create a production system
that can produce entire products.
Flexible manufacturing system is a job shop process, but the delays that happen in a job shop process
are avoided – set ups are done rapidly, materials are loaded rapidly into the machines, materials are
moved rapidly between the machines, and materials are rapidly unloaded from the machines. For
example, a flexible manufacturing system has three CNC machines A, B, and C, and three products X,
Y, and Z have to be produced and their processing requirements are as follows: Product X: B – C – A –
C – B, Product Y: A – B – A – C – B, Product Z: C – B – A – C. The processing requirements (with their
sequence) of all the products are fed into the central computer which controls all the sub-systems like
tool changing system and material handling system. When product X arrives at machine B, its set up
has already been done, and while it is being processed at machine B, machine C is being set up for it .
And as soon machine B has finished processing product X, it is set up for product Y which would be
the next product to arrive on machine B. Similarly, product Z arrives at machine C which is already set
up for it. Products wait at machines as they do in a typical job system – either product Y or product Z
may wait at machine B. It would have been wise to use batch flow in job shop to handle such variety
of products, but a flexible manufacturing system can produce in unit of one because the set ups are
done rapidly and materials are moved rapidly between the machines – product A is produced and then
product C is produced and then product A is produced and so on.
A product is of superior quality when the following two conditions are met simultaneously: (i) Design
of the product is such that it provides all the benefits and has all the features that customers desire –
Design quality (ii) The design is faithfully and consistently converted into products – Conformance
quality. It may happen that designers of a company may do an excellent job of identifying customers’
needs and coming up with a design that will meet customers’ needs, but if production department is
not able to produce products according to the design, customers will reject the product – Inferior
quality due to poor conformance quality. Similarly, if production department is able to produce
products according to the design, but the design does not provide benefits and does not have features
that customers desire, customers will reject the product – Inferior quality due to poor design quality.
DESIGN QUALITY
Design quality is about incorporating customers’ needs into products faithfully. A designer can include
the benefits and features that customers desire in the product design only when he understands
customers needs intimately. A designer can use following three methods to understand customers’
needs and incorporate them into the product being designed:
Customers are included in design team: Designers survey customers to understand their
needs, but customers’ needs are often sidelined when difficult trade-offs have to be made –
for example, a feature desired by customers is not included in the design because it will
increase the cost of the product. But, when customers are part of the design team, they do
not allow designers to make such trade-offs. When customers and designers interact during
the design process, they are able to understand each other’s positions, and suitable
compromises are found – for example, the designer will be willing to include the aforesaid
feature, but will the customer be willing to bear the extra cost?
Prototyping: Even when designers have done all that they could to ensure that customers’
needs are incorporated into the product being designed, they would not know if they have
been able to do so until customers use the product. Designers produce a few units of the
product before going for full-scale production, and test them for its functionality and looks –
Do the products provide all the benefits and have all the features that were planned, and do
they look as was thought they would look? The product is redesigned if the designers are not
satisfied with the functioning and looks of the products. Customers are also asked to use these
products, and the product is redesigned if their feedback is negative. This process of producing
few products and testing them for functionality and looks is called prototyping.
CONFORMANCE QUALITY
Once customers’ needs have been incorporated in the product design, it must be faithfully converted
into products – the products should provide the benefits and have the features that are incorporated
in the product design. Though production process is primarily responsible for conformance quality,
the way the product has been designed also determines if the product design will be faithfully
converted into products.
Designers can so design a product that it is easily producible – designers should always keep in mind
the producibility of the product being designed while they are designing the product. Following
product design principles improve conformance quality:
Minimum number of components: When a product can be made from fewer components,
there are lesser number of production processes to install and monitor – company can buy
the best equipments for its production processes and hire skilled operators to run them.
Lesser number of components have to be inspected, and hence their inspection can be made
more stringent – defective components are not allowed to move to the next production stage.
Common components: Common components are components that the company uses to
produce its products. When a product can be made from components that are already being
used to make other products of the company, their production processes are already well
established, and hence chances of producing defective components are lesser. Also, since
common components are used in many products, they are produced in large volumes, and
hence the company installs high-capacity equipments to produce them. High-capacity
equipments are generally imbued with latest technologies, and hence they are less likely to
produce defective products – they would also make components with finer tolerances.
Standard components: Standard components are components that are used by large number
of companies to produce their own products. The company purchases standard components
from manufacturers who specialize in production of these components. Manufacturers of
standard components produce in large volumes, and hence they use high-capacity
equipments to produce them. High-capacity equipments are imbued with latest technologies,
and hence they can produce components with finer tolerances at low cost.
Assembly process: External dimensions of components should be such that they can be easily
put together to make a product. When designers create prototypes, they should also pay
attention to whether the components are being easily put together – external dimensions of
components should be changed if components are not easily put together.
Reasonable tolerance / design range: Narrower tolerances result in better quality, especially
when large number of components have to be put together to get a functional output – it may
happen that all components are within their specified ranges, but too many components are
near the end of their specified ranges, and the product does not perform due to what is called
‘tolerance stack up’. Therefore, narrower the design range, lower is the probability of
producing a defective product due to ‘tolerance stack up’. But, it is expensive to produce
components in narrower tolerances – equipments that produce in narrower tolerances
employ sophisticated technologies, and hence are expensive. Therefore, if a designer
prescribes narrower design range / tolerance, either expensive equipments have to be
purchased or large number of defective components would be produced. Therefore,
designers should not prescribe tolerances narrower than those needed for the components
to work well together – cost of production is raised, but customers are not more satisfied if
unreasonably narrower tolerances are prescribed.
1. Sporadic problems – Sporadic problems are short term and cause sudden change in quality.
These are identified and controlled by techniques like statistical process control.
2. Chronic problems – Chronic problems are long term and cause poor quality continuously.
They require new and scientific solutions.
Statistical process control is used to measure the performance of a process. Statistical process control
uses control charts to identify when a production process has gone out of control and needs corrective
action. A process is said to be operating in statistical control when the only source of variation is
natural or random. The process must first be brought into statistical control by detecting and
eliminating assignable or non-random causes of variation. Then its performance is predictable, and its
ability to meet design specifications can be assessed. Control charts provide statistical signals when
assignable causes of variation are present.
VARIATION IN PROCESSES
Types of variations
A production process is designed to produce a specific dimension. For example, a lathe machine is
designed to produce rods of 10 cm, but the length of rods will vary from its target value of 10 cm.
Production process face two kinds of quality conformance problems:
1. Natural or Random variation – Random variation is related to the process, and there is no
specific cause for it. All processes exhibit some random variation, and this variation is due to
factors like temperature variation, machine vibration, and randomness in human actions. The
process is under control, and this variation can never be reduced to zero. It can be reduced by
increasing the technical sophistication of equipments – high-technology equipments will have
lower random variations than low-technology equipments.
It is important to be able to distinguish between these two variations – both show up as deviations
from the target value. For example, a rod of 10.3 cm is produced when the target value of length is
10.0 cm. The deviation of 0.3 cm can be either random variation or assignable variation – if the
deviation is random variation, then nothing can be done about it, and the machine should continue to
run, but if the deviation is assignable variation, then the machine should be stopped, and repaired. If
a company is not able to distinguish between the two kinds of variations, it will stop a process that is
under control, and it will continue running a process that has gone out of control.
Statistical process control recognizes that some variation from the target value is inherent in a
production process – this variation is less when more technologically sophisticated equipments are
used, but some variation will always be there. Statistical process control uses control charts that helps
to determine whether variations from the target value are due to natural randomness or due to
process being out of control. Control charts identify an out-of-control situation before defective
products starts getting produced.
Control charts provide a signal when assignable causes of variation are present.
1. Take a sample of around 20 units, and measure a characteristic like radius. Plot frequency and
radius.
2. Take around 10 such samples at different times and plot the curves.
3. If all curves are normally distributed, variation is only natural – distribution is stable over time
and is predictable. But, if there is any curve which is not a normal distribution curve, then
there are some assignable variations, and investigations should be carried out to find their
cause. Unexpected distributions that vary by mean, standard variation, and shape signal that
assignable causes of variation are present.
4. In both random and assignable variation, some radius may be out of lower and higher design
limits, which additionally means that the process is not capable of performing within control
limits.
Process capability study is carried out to find if the process is capable of producing within design limits.
5. Value of K is decided by management. It has to decide whether its applications require use of
wider or narrower control limits. Generally it is kept at 3. Therefore, substituting K = 3 in UCL
and LCL,
For all products to fall within UCL and LCL, process range or UCL – LCL ≥ 6σ or
(UCL – LCL) / 6σ ≥ 1
Therefore, for six sigma quality level, (Upper design limit – Lower design limit = Design range)
≥ (Average + 6σ) – (Average – 6σ) ≥ 12σ
If K is kept small (K = 1), control range is small. The process may seem out of control even
when it might be meeting functional requirements. This is called Producer’s risk or Type I
error.
If K is kept large (K ≥ 3), control range is large. The process may seem in control even though
it might not be meeting functional requirements. This is called Consumer’s risk or Type II error.
Therefore, process range is high when K = 6, and product should be performing its functions
at this high range, which can happen only under two circumstances:
Innovations in design enable the product to function even when design range is high
– Design range ≥ 12σ.
σ is reduced significantly from K = 3 levels so that 12σ is still an acceptable range within
which the product performs. Control of the process is strong – variability is low. Since
only chronic problems exist, lowering σ means that process variability is less. Process
variability will be less when technological capabilities of the equipment will be
upgraded.
Therefore, six sigma quality level can be achieved either by making design innovations or by reducing
σ.
It is important to keep in mind that products’ performance is better when range is as low as possible
so that even when components of extreme specifications come together, performance is still
adequate. For example, consider the body of a pen and its cover. The design specification of the body
is 9 mm while the design specification of the cover is 10 mm. Two possible ranges for body are 8.9 –
9.1 and 8.7 – 9.3. Similarly, two possible ranges for cover are 9.9 – 10.1 and 9.7 and 10.3. If the larger
range is considered, and body of dimension 8.7 and cover of 10.3 are matched, the fit would be loose.
Similarly, if the body dimension of 9.3 is considered and cover dimension of 9.7 considered, the fit
would be tight. If the smaller range is considered, and body dimension 8.9 and cover dimension of
10.1 are matched, the fit would be better than the earlier case. Similarly, if the body dimension of 9.1
is matched with cover dimension of 9.9, the fit would be much better than the earlier case. And this
is simple product with just two components. When many components have to work together, it is very
important that range of each component is small so that even when components of extreme
specifications come together, the performance of the product is still adequate.
7. Once UCL and LCL established, a sample is taken periodically (say, 5 units every hour), and its
mean is plotted on the X-chart. If the mean is within UCL and LCL, the process is in control,
and is allowed to run. But, if a mean of a particular sample is above UCL or below LCL, another
sample is taken immediately, and if it is again outside the limits of UCL or LCL, the process is
stopped, and investigation is carried out. The idea is that even when k = 3, only 0.3 (100 –
99.7) per cent of products should lie outside UCL and LCL, and when two consecutive samples
are outside LCL and UCL, there is something wrong with the process.
PROCESS CAPABILITY
If natural variation or process standard deviation or σ is less compared to design range / tolerance,
defect rate would be low – process capability is high. If natural variation or σ is high compared to
design range, defect rate would be high – process capability is low. For example, if σ of a process is
0.1, the process capability will be higher if the design range is ± 0.4 than if the design range is ± 0.2.
An expression for process capability is Design range / σ. If a process has high process capability, it will
not produce defective products even if process mean moves out of control by 1-2 σ. If process
capability of a process is low, it will produce defective products even when the process is in control,
and defect rate will increase dramatically if the process goes out-of-control even slightly. Therefore, if
a process has high process capability, process mean can be allowed to move away from the target
value without producing defective products. And when the design range / tolerance is ± 6 σ (12 σ),
only 3.4 products in a million products would be defective – six sigma quality level.
Process capability ratio = Design range / Process range. Process range is 2 n σ, and since n = 3 in most
cases, process capability ratio = Design range / 6 σ. Following conditions emerge:
If process capability ratio is greater than 1, the process is capable of producing within design
specifications.
If process capability ratio is less than 1, the process is not capable of producing within design
specifications.
Another related term is process capability index, which indicates if the process mean has shifted away
from the design target, and in which direction it has shifted.
Process capability index = minimum of (Average – Lower specification) / 3 σ and (Upper specification
– Average) / 3 σ. Following conditions can emerge:
If process capability index is greater than 1, the process will produce within the design range
/ tolerance.
If process capability index is less than 1, the process mean has either moved to upper design
specification or lower design specification, and defective products would be produced.
σ can be found out through controlled experiments, but a simpler way to calculate σ is by using range
of an initial set of samples – at least 20 samples must be used. A minimum of 20 samples are taken,
and range of each sample is calculated – range is the difference between the largest and lowest value
in the sample. The average of ranges is calculated and is divided by d2 to get the value of σ.
ACCEPTANCE SAMPLING
Acceptance sampling is a statistical quality control technique used to decide to accept or reject a batch
of input / output by examining the quality of random samples of the batch.
100 per cent inspection is time consuming and costly. 100 per cent inspection might be essential in
industries with life-and-death consequences like aerospace, but here too, the system must move from
human inspection towards automated methods. In case of destructive testing, acceptance sampling is
the only acceptable alternative.
Accepting sampling involves taking random samples of batches of products and measuring
them against predetermined standards. The quality of the samples is used to judge the quality
of all the products in the batch.
Sampling plan specifies the number of units to sample and the number of sample units that
must conform to specifications if the shipment is to be accepted.
The specification must contain a statement of what constitutes non-conformance.
How is it done: n units are taken from the batch and quality of each sampled unit is
determined. If C or fewer units are non conforming, the entire shipment is accepted. And if
more than C units are non conforming, the whole batch is rejected or subjected to 100 per
cent inspection. C is called acceptance number. This plan is called single sampling plan.
Acceptable Quality level (AQL) is the quantity or proportion of non-conformance which is
allowed if a batch of product is to be considered satisfactory. It is the poorest level of quality
that will be acceptable – lots that have this level of quality will be accepted. For example, if an
acceptable quality level is 20 defects in a lot of 1,000, then AQL is 100 x 20 / 1,000 = 2 per cent
defectives. AQL will depend on the end use of the product. There are critical, major, and minor
defects which demand different quantification if customer satisfaction is to be ensured – AQL
should be low for critical defects.
Lot tolerance percent defective (LTPD) is the quality level of lot that is considered bad – lots
with this level of quality will be rejected. For example, if unacceptable quality level is 70
defects in a lot of 1,000, then LPTD is 100 x 70 / 1,000 = 7 per cent defective.
Double sampling – Two samples are taken and cumulative evidence from both samples lead
to either an acceptance or rejection.
Multiple sampling – Many samples are taken and cumulative evidence of all samples lead to
either an acceptance or rejection.
Sampling errors
The batch is accepted or rejected based on quality of samples – if decision is to accept, the whole
batch is accepted, and if decision is to reject, the whole batch is rejected. It does not give guarantee
of actual quality of the batch. The producer wants to avoid the mistake of having a good lot rejected,
and the customer wants to avoid the mistake of accepting a bad lot.
1. Type I – α – Producer’s risk: It is the probability that a good lot will be rejected. A batch of
good quality can be rejected if disproportionately large number of defective units from the
batch is selected in the sample – the sample has higher proportion of defects than the batch
as a whole.
2. Type II – β – Consumer’s risk: It is the probability that a bad lot will be accepted. A batch of
bad quality can be accepted if disproportionately large number of good units from batch is
selected in the sample – the sample has lower proportion of defects than the batch as a whole.
Operating characteristics curve describes how well an acceptance plan discriminates between good
and bad lots. A curve pertains to specific plan – a combination of n and C.
1. Ideally, all batches which have defects less than C should be accepted, and all batches which
have defects greater than C should be rejected. In practice, no sampling plan will offer such
discrimination.
2. Operating characteristic curves reveal how sampling plans discriminate among lots.
3. If a lot is of high quality (low percentage of defective items), a good sampling plan yields a
high probability of accepting the lot. If a lot is of poor quality (high percentage of defective
items), a good sampling plan yields a low probability of accepting the lot.
4. Operating curves show relationship between probability of accepting lot on Y axis and quality
of lot (percentage of defective items) on X axis. A sampling plan specifying an unique pair of n
and C has an unique operating characteristic curve – Steeper is the curve better is the sampling
plan. These curves are developed by using Poisson distribution.
5. As n increases, the sampling plan becomes more discriminatory – the operating characteristic
curve is more rectangular. Therefore, larger is the value of n, larger is the probability of
accepting good quality lots and lower is the probability of accepting bad quality lots. But
inspection cost is increased.
6. By decreasing C, the inspection is made more stringent. Low value of C does not guarantee
defective free batches, but many bad lots will be rejected which might have been accepted if
value of C was higher.
7. Choosing the value of n and C is a management decision. High value of n and low value of C
show high quality consideration.
COST OF QUALITY
Most companies do not know what it costs them to produce products of inferior quality. Quality cost
audit identifies all costs which are incurred to produce products of superior quality or costs that results
from producing products of inferior quality. It is important that a company carries out quality cost
audit so that it learns how expensive can it be to produce products of inferior quality – most companies
need compelling evidence to install quality systems that enable them to produce products of superior
quality.
Traditional cost of quality was costs incurred in production of defective products and in their rework.
These costs include direct material and labor costs and overheads during production and rework. An
in-depth analysis of costs of quality often reveals that companies are actually spending several times
this amount on activities related to quality.
Quality costs are all the costs or losses of revenue that are linked to prevention of production of
defective products. They are also linked to producing, detecting, rectifying, and scrapping defective
products. There are four categories of quality costs: (i) External failure cost (ii) Internal failure cost (iii)
Appraisal cost (iv) Prevention cost. Failure costs result from producing defective products. Failure cost
is external failure cost if it is discovered by customers. Failure cost is internal failure cost if it is
discovered before it is delivered to customers. Appraisals costs are the costs of inspecting the
condition of raw materials, work-in-progress, and final products, and identifying the defectives ones
among them. Prevention costs are incurred to minimize appraisal and failure costs – focus is on
installing better equipments and setting better production processes.
External failure costs – Product fails after the product is delivered to the customer.
Internal failure costs – Product fails before the product is delivered to the customer.
The cost of determining as to what activities would be undertaken when a defective product
is discovered: (i) Rework the defective product. (ii) Discard the defective product (iii) Adjust
the equipment which produced the defective product.
Scrap cost: It is the material and processing cost of a defective product that has to be
scrapped.
Rework cost: It is the cost of processing a defective product to make it reusable.
Retesting cost: It is the cost of retesting the defective products that have been reprocessed.
Yield loss: It is the cost of any extra material that is used in producing a product.
Downtime cost: Equipments run idle when production is stopped to solve a quality problem.
The cost of safety stock that is held to at all production stages to substitute for defective
products that might be produced.
Appraisal costs – It is cost of determining the quality status of raw materials, in-process materials, and
final products.
Cost of inspection of incoming material: This is the cost of inspecting components and
materials from suppliers.
Cost of inspection of in-process material: Inspection is carried out at production stages to
determine if the output of each stage conforms to its design specification.
Cost of inspection of final products.
Cost of purchasing and maintaining testing equipments.
Salary and other benefits of inspectors.
Prevention cost – It is the cost of actions that are taken to reduce defect rates.
Planning: Costs are incurred to ensure that tasks are performed correctly. These include costs
to install correct procedures for performing tasks.
Product design: Product design is reviewed. All product designs should be evaluated on two
counts: (i) Does the product design include benefits and features that are desired by
customers. (ii) Is the product design easily producible?
Process design: Production process is reviewed. Can the installed process convert the product
design into products faithfully and consistently, or should new processes be installed?
Process control: Equipments are continually monitored to ensure that products are being
produced within the range that has been specified by the product designers – for example,
within 10 ± 0.1 cm. Statistical control charts are used to monitor production processes – they
are able to indicate if a process is behaving normally and will continue to produce products
within the design range or if it is behaving abnormally and will soon start producing defective
products.
Preventive maintenance of equipments is carried out.
Job design: Jobs are so designed that workers get to learn the quality status of the output they
have produced as early as possible. Workers are regularly trained.
Suppliers are trained.
Data collection and analysis: Data related to defects, downtime, and other quality problems
are collected at every production stage. This data is analyzed to arrive at solutions to quality
related problems.
Root causes of manufacturing problems are solved.
Quality improvement programs: The company runs quality improvement programs like quality
circles, suggestion schemes, and other defect reduction programs.
INVENTORY MANAGEMENT
Inventory is stock of material which is used to facilitate production or to satisfy customer demand.
Therefore, inventory is the product at some point in conversion and distribution process. There are
three types of inventory: Raw materials, Work-in-process, and Finished products. Raw material and
work-in-process facilitate production, and finished product serves customer demand. Inventory of
finished product is required because rate of demand differs from rate of supply.
Cost of material forms a substantial part of the total cost of a product. Material is frequently stored
either raw, partly finished or completely finished, and cost of storing such materials can be high. Stock
should be considered as an investment from which some return should be expected – thus if target
rate of return is 10 per cent, the stock will cost at least 10 per cent of the average value of stock held.
If stock is purchased from borrowed money, then cost of capital should also be added to the cost.
Therefore, there is urgent need to manage inventory well.
Companies are trying to reduce inventory. Inventory ties up capital which can be used for other
purposes, and they incur expenditures of storage, insurance, and deterioration. Excess inventory is
often a symptom of deeper malaises such as poor quality, poor maintenance of machines, slow set
up, poor scheduling, and unreliable suppliers – when inventory is reduced these problems come to
the surface one by one, and management can take actions to solve them. For example, a company
may be keeping work-in-progress inventory because it expects some of the components to be
defective – the company will not need to keep inventory when it has improved its quality system to
the extent that it does not expect any component to be defective.
It is not to be believed that inventory serves no purpose at all – when production is done in lots, the
set up cost is distributed among the components in the lot, and when transportation is done in lots,
the transportation cost is distributed among the components in the lot. Therefore, inventory enables
efficient production. Inventory also enables a company to serve customers more promptly, and serve
unpredictable demand more efficiently.
INVENTORY COSTS
Since all components of holding cost is directly proportional to the amount of inventory being
held, holding cost is generally given as ‘holding cost per unit of product per unit time’. It is
normally given as a fraction of the product’s price – for example, holding cost may be 10 per
cent for potato and 20 per cent for tomato, and if the price of both potato and tomato is Rs.
20 per kg, holding cost of potato is 10 / 100 x 20 = Rs. 2 per kg per year, and holding cost of
tomato is 20 / 100 x 20 = Rs. 4 per kg per year.
Stockout cost: A stockout situation arises when the company does not have products in
inventory to serve customers. Cost of stockout is the loss of profit that the company would
have earned if it had productS in inventory to sell to customers. It also includes lost future
profits if the customers feel so aggrieved that they stop buying the company’s products. But,
if it is a strong brand, some of its customers might be willing to wait to be served or the
company can offer some incentives like lower prices to the customers to make them willing
to wait to be served – such a situation is called Backordering, and it can actually be a good
situation to be in, because a backorder is a confirmed demand.
A company can proactively convert some of its demand into backorders by asking customers
to place orders in advance in return of some price concessions so that sigma or variability
influences lesser part of its total demand, and hence it would need to keep lesser safety stock
– for example, a car company may offer a price reduction of Rs. 5,000 to customers who book
their cars one month in advance. It is important to keep in mind that backorder is type of
demand in which customers want to be served in the current period but are willing to wait
either because they do not want to buy any other product or because there is some financial
incentive to do so. Backordering is different from reservation in which the customers want to
be served in a future time period but book company’s capacity in current period as they do
not want to take the risk of not being served due to company’s lack of capacity – for example,
railway reservation.
Another type of stockout situation occurs if a company does not have raw materials or work
in progress inventory to run its production, and associated costs in this type of stockout
situation are idle machines and workers, and lost production.
Hidden costs: Companies keep inventories because (i) they expect some of the products to
be defective (ii) they expect some of their machines to break down (iii) they expect their
suppliers to default on their delivery schedules (iv) they expect their forecasting to be wrong
(v) they expect their lead times to be longer than what they are supposed to be. Therefore,
companies keep inventory to compensate for fundamental shortcomings in their operations,
and that if these shortcomings were removed, they would not have to keep these inventories
– worse, these inventories help in hiding these problems and companies do not get to
identifying and removing these shortcomings. The solution is to slowly reduce the level of
inventory and solve the problems as they surface – for example, current level of inventory is
100 units, and when it is reduced to 80, it is discovered that the planned number of output
could not be produced because two machines broke down, and company develops a
preventive maintenance schedule to prevent such breakdowns.
PURPOSE OF INVENTORY
Inventory can be useful – a company should not try to eliminate all inventory, and it should be able to
identify inventory which is useful, and should do away with only the unnecessary inventory.
Decoupling the stages of production: Inventory uncouples various stages of operation. Raw
materials uncouple a manufacturer from its suppliers. When a manufacturer keeps inventory
of raw materials, its operation can run uninterrupted even when its supplier is not able to
supply for some time. Inventory of raw material also helps when the supplier’s lead time is
uncertain. Work-in-process inventory uncouples various stages of manufacturing. If a
company does not keep work-in-process inventory, and a machine breaks down, all the
machines that follow the ‘machine under break down’ will run idle because they will not get
material from their preceding machines. But, if work-in-process inventory is kept at all
machines, then all the machines will run for at least some time even if a machine breaks down.
Similarly, if a machine produces a defective product, the machines that follow will produce
one product less if the machines do not keep work-in-process inventory. But, if work-in-
process inventory is kept at all machines, they can produce the required number of products
even if some defective products are produced.
Inventory of raw material and work-in-process insulate a stage of production from its
preceding stage, and hence work continues to be done even when a supplier does not supply
on time , or a machine breaks down, or when a defective product is produced. It also implies
that a manufacturer need not keep inventory if it takes care to enroll reliable suppliers, has a
robust preventive maintenance program, and has installed an effective quality management
program. Therefore, it can be inferred that a manufacturer keeps raw material and in-process
inventory to compensate for unreliable suppliers, and a production system that is likely to
break down, and also produce defective products.
Set up cost and transportation cost are spread over larger number of units: Inventory allows
economic production and transportation. Most machines produce many products. A machine
has to be reconfigured or setup if it has to produce a product different from the one it is
producing currently. Therefore, a machine incurs setup time each time it has to produce a
product different from the one it is producing now. For example, a machine processes
products A, B, and C. It takes 10 minutes to setup the machine to produce B when it is
producing A, 15 minutes to setup the machine to produce C when it is producing B, 5 minutes
to setup the machine to produce A when it is producing C. It takes 5, 7, and 9 minutes to
produce A, B, and C respectively. If the company decides to produce one unit each of A, B, and
C in the sequence A-B-C, the average production time of one piece would be 5 + 10 (set up) +
7 + 15 (set up) + 9 / 3 = 15.3 minutes, but if it decides to produce 10 pieces each of A, B, and
C in the sequence A-B-C, the average production time of one piece would be 5 x 10 + 10 (set
up) + 7 x 10 + 15 (set up) + 9 x 10 / 30 = 7.8 minutes. The average production time would
decrease as the production lot size increases because the setup time is spread over larger
number of units. But as production lot size increases, the average inventory of each product
increases – the average inventory is half of lot size. Companies have to reduce setup times of
their machines if they have to reduce their production lot sizes.
The per unit cost of transportation is low when products are transported in bulk. For example,
a truck has a capacity of 1000 units of product A, and it takes Rs. 2,000 to transport from
location X to location Y. If a company moves 100 pieces, then per unit cost of transportation
would be Rs. 2000 / 100 = Rs. 20, but if it moves 1000 pieces, then per unit cost of
transportation would be Rs. 2000 / 1000 = 2, but the inventory at Y increases. If a company
wants low per unit cost of transportation, and also keep its inventory low, it should use third
party to transport its products. The third party will fill its truck with products of large number
of clients, and hence achieve low per unit cost of transportation as well as low inventory levels
for its clients. Companies that follow Just in-Time principles insist that their suppliers locate
themselves close to their manufacturing premise so that the cost of transportation is low even
when suppliers are supplying in smaller lot sizes.
Takes care of predictable rise in demand: Companies keep inventory to take care of
anticipated changes in demand and supply. For example, an automobile company will keep
higher inventory during Diwali as it understands that large number of customers would have
postponed their purchase to Diwali. Similarly, if a manufacturer anticipates a shortage of a
particular raw material or component, it may stock the particular raw material or component
so that its production does not get adversely affected. These are speculative inventories.
Takes care of operations during transit time: Companies keep inventory to take care of transit
between locations and between stages. For example, if takes 10 hours to get a component
from a supplier, the company will need to have enough inventory to run its production unit
for 10 hours when it orders the component, but if it took only 5 hours to get the component
from its supplier, it would need enough inventory to run its production unit for 5 hours.
Therefore, the larger the transit time of a component, more inventory of it have to be kept to
ensure uninterrupted operation.
TYPES OF INVENTORY
Inventories are held for different purposes, and inventory can be of three types depending on why
they are being held.
1. Cycle inventory: A company produces in lots because its set up cost is significant – when it
produces in lots, the set up cost is spread over all the units in the lot, and hence per unit set
up cost is low. Similarly, when a company transports in lots, the transportation cost is spread
over larger number of units, and hence per unit transportation cost is low. Such inventories
which enable companies to decrease their cost of production / distribution are called cycle
inventories – they enhance economic efficiency.
Average cycle inventory = Lot size / 2
It also means that half of the lot remains unused throughout the year.
2. Safety inventory: Demand is unpredictable, and hence inventory is kept to serve demand that
is above the average demand of previous periods. Hence, safety stock is related to sigma /
variability of demand – higher the sigma of demand, higher should be the safety stock.
Safety inventory = Z σ (t)
Z is safety factor which is related to service level. Service level is the percentage of customers
that a company may want to serve. Though a company may keep any service level that it may
want to, a suitable expression for service level is (Price – Cost) / (Price – Salvage value). Salvage
value is the price fetched after a certain period of time – salvage value of biscuits is higher
than the salvage value of bread. Hence, a company should keep higher service level if its profit
per unit (Price – Cost) is high, and also if its salvage value is high. A company can also try to
increase its salvage value – a retailer of milk can have contract with paneer makers to supply
them milk after it has deteriorated. It is also important to remember that as service level
moves closer to 100 per cent, the safety stock needed to increase service level by 1 per cent
is much more than the safety stock needed to increase service level at lower values of service
level – for example, it may need 10 units of safety stock to increase service level from 80 per
cent to 81 per cent, but may require 100 units of safety stock to increase service level from 95
per cent to 96 per cent, and 1000 units of safety stock to increase service level from 98 per
cent to 99 per cent. This happens because as service level increases close to 100 per cent,
safety stock has to be kept to care of the few unusual hikes in demand. Once a company has
determined its service level, it can find the corresponding safety factor(Z).
𝐶𝑜𝑠𝑡 𝑜𝑓 𝑢𝑛𝑑𝑒𝑟𝑠𝑡𝑜𝑐𝑘𝑖𝑛𝑔 𝐶𝑢
Service level is also equal to 𝐶𝑜𝑠𝑡 𝑜𝑓 𝑢𝑛𝑑𝑒𝑡𝑠𝑡𝑜𝑐𝑘𝑖𝑛𝑔+𝐶𝑜𝑠𝑡 𝑜𝑓 𝑜𝑣𝑒𝑟𝑠𝑡𝑜𝑐𝑘𝑖𝑛𝑔 = 𝐶𝑢+𝐶𝑜
A service factor table has been given below or one can use Excel function NORMSINV to
convert service level percentage to safety factor (Z).
Table 1: Safety factor
Reduction in safety inventory: If forecasting is done close to the selling period, chances of it
turning out to be closer to actual sale is higher than if forecasting is done away from the selling
season – more changes are likely to happen in the intervening period. But, forecasting can be
done closer to the selling period only if lead time is small – for example, if lead time is six
months, forecasting has to be done six month before the selling season, but if lead time is only
one month, forecasting can be done just a month before the selling season, and this forecast
would be more accurate. Also, if lead time is small, multiple forecasting can be done for one
selling season, and products can be produced and dispatched in smaller batches – for
example, if selling season is of four months, and lead time is of one month, then forecasting
can be done for each month of the selling season, and since each forecast will be done closer
to its selling period, it will be more accurate. Hence, safety inventory can be reduced by
reducing lead time.
Safety inventory can also be reduced by motivating customers to place advance orders by
offering them incentives. For example, a car company offers an incentive of Rs. 10,000 if a
customer places an order one month in advance, and if say, 20 per cent of customers place
orders in advance, uncertainty / sigma impacts only 80 per cent of the demand.
Reduction in speculative inventory: If the production rate is equal to demand rate of high-
demand period, speculative inventory is not required, but there would high idle capacity
during low-demand periods. Hence, increasing production rate may not be a viable strategy.
A more viable strategy would be to shift demand from high-demand periods to low-demand
periods by offering lower prices during low-demand periods, and when such a shift occurs
demand in various time periods are more equal to each other.
4. Pipeline inventory: It is the inventory in-transit such as the one in a truck moving from a
supplier to its buyer.
Pipeline inventory = Average demand x Lead time (Time taken to move inventory from
supplier to buyer)
Reduction in pipeline inventory: Reduce lead time by using faster means of transport or by
selecting customers who are close by.
There is cost in placing an order. This cost is spread over the number of units in the batch. Larger the
batch size, fewer the number of batches required per year and lower is the ordering cost. With larger
batch size, average inventory levels are higher and stockholding costs increase – Average inventory
level = Batch size / 2. The optimum batch quantity is where the total cost is minimum and this quantity
is known as Economic order quantity or Economic batch size – sum of ordering cost and holding cost
is minimum at economic order quantity. Total cost (ordering cost + holding cost) is considered because
ordering cost and holding cost move in opposite directions with regards to batch size – ordering cost
decreases when batch size increases, and holding cost increases when batch size increases and vice-
versa.
A company has to decide its order size. For example, a company’s annual demand for a product may
be 1000. It can place one order of 1000, or it can place 5 orders of 200 each, or it can place 10 orders
of 100 each. There is tradeoff between ordering frequency and inventory level. Small order size will
lead to frequent reorders, but inventory level will be lower. Large order size will lead to less frequent
orders, but inventory level will be higher.
It is important to remember that half of order size remains in inventory permanently. If a company
orders Q amount, and keeps ordering Q amount after T time when Q has been consumed, it carries
an average inventory of Q / 2 throughout the year. For example, a company has an order size of 10
units and one item is consumed every day. Therefore, it receives an order of 10 units every 11 th day.
On the 1st day, the company has inventory of 10 items, on the 2nd day, it has inventory of 9 items, and
so on. It has zero inventory on the 11th day. The average inventory that the company holds = 10 + 9 +
8 + 7 + 6 + 5 + 4 + 3 + 2 + 1 + 0 / 11 = 5 = 10 / 2 = Q /2. The average inventory is the inventory that is
held throughout the year.
Therefore, it is important that steps are taken to reduce order size, since it directly results in reduction
of average inventory levels. Holding cost is directly proportional to average inventory – Holding cost =
Holding cost per unit per year x Average inventory ( Q / 2 ). Holding cost is normally given as some
percentage of cost of product – for example, holding cost of tomato may be 15 per cent of its price,
whereas holding cost of potato may be 5 per cent of its price, since tomato requires a more controlled
environment than potato.
Ordering cost is cost of all the orders that a company places. Therefore, ordering cost = ordering cost
of one order x number of orders. Number of orders = Annual demand / Order size. The number of
orders will be more if order size is small, but the average inventory will be lower. Conversely, the
number of orders will be less if order size is large, but the average inventory will be higher. Since
ordering cost is directly proportional to number of orders and holding cost is directly proportional to
average inventory, ordering cost is inversely proportional to holding cost. For example, if annual
demand is 1000, and order size is 10, the number of orders will be 1000 / 10 = 100, and the average
inventory will be 10 / 2 = 5, but if the order is 100, the number of orders will be 1000 / 100 = 10, and
the average inventory will be 100 / 2 = 50.
Since ordering cost and holding cost move in opposite directions with regards to order size, ordering
cost decreases and holding cost increases when order size is large, whereas ordering cost increases
and holding cost decreases when order size is small. Therefore, the best order size is one where the
sum of ordering cost and holding cost is minimum.
To arrive at an expression for the best order size or economic order quantity (EOQ) , let
Number of orders = D / Q
Ordering cost = S x D / Q
When Q is the best order size or economic order quantity, Total cost is minimum. To find a value the
Q, expression of Total cost is differentiated with respect to Q, and equated to zero.
d(Total cost) / dQ = - S x D / Q2 + I x C / 2 = 0
S x D / Q2 = I x C / 2
Q2 = 2 x S x D / I x C
Q=√2xSxD/IxC
Orders are placed when the level of stock drops to a previously determined level called Reorder level
or ROL. An order is placed for a fixed quantity called Reorder quantity or ROQ. ROL is based on demand
rate and procurement lead time, so that the order arrives before stockout or before a minimum level
is reached. ROQ is based on ordering cost, holding cost and quantity discount. Continuous review
system has fixed order quantities but variable time intervals between orders.
The stock position is continually monitored, and when the stock position drops to Reorder point (ROL),
a fixed quantity (ROQ) is ordered. The order arrives after its lead time (L), and cycle of consumption,
reorder, and order receipt is repeated.
Reorder level (ROL) is set such that there is enough stock to serve the demand that arises during lead
time. Normal distribution of demand is assumed during lead time, and the inventory is exposed to
stockout during lead time – order has been placed, but stock is yet to arrive.
Reorder level (ROL) = Average demand over lead time + Safety stock
Safety stock = Safety factor (Z) x standard deviation of demand over lead time (σ) = Z x σ (L)
For example, a company has an annual demand of 60,000, and its supplier’s lead time is 5 days. The σ
of its daily demand is 160, and its desired service level is 90 per cent. Its ordering cost is Rs. 20, and its
holding cost is 15 per cent. The cost of the product is Rs. 10.
Order size (ROQ) = √2.D.S / I.C = √2 X 60,000 X 20 / 0.15 X 10 = 1131 (Economic order quantity)
Average daily demand = d = 60,000 / 365 = 164
Variance (σ2) is additive, hence σ2 over lead time ( 5 days) is 5 times daily variance = 5 x 1602 = 128,000
For 90 per cent service, the corresponding value of z = 1.3 (From table 1)
Reorder level = Average demand x Lead time + z x σ (lead time) = 164 x 5 + 1.3 x 358 = 820 + 465 =
1285
Therefore, an order of 1131 is placed every time the stock position drops to 1285.
It may also happen that lead time itself may vary with a mean of L and a standard deviation of σ L. In
such cases, σ during lead time is calculated by the following formula:
For example, daily demand of a product is 1000 (D), and σ of demand is 100 per day. Its average lead
time is 0.5 day with σ of lead time being 0.1 day. Therefore, σ during lead time = √ 0.5 x 100 x 100 +
1000 x 1000 x 0.1 x 0.1 = 105
Continuous review system can be followed when current inventory position is known – ROQ is ordered
when inventory level drops to ROL. With advances in computer technology, it is now possible to know
the inventory position of any number of items any time, and hence continuous review system is
becoming popular. Since the company always knows its inventory position, it can place an order at
such a time that new stock arrives before stockout can occur.
Two-bin system: It is the visual system version of Q system – a visual system allows employees to
place orders when inventory visibly reaches a certain level. In two-bin system, inventory of a product
is kept in two bins. Inventory is first withdrawn from one bin. When the first bin is empty, an order is
placed. Inventory is withdrawn from the second bin till replenishment arrives. When the
replenishment arrives, the second bin is filled first, and the remaining units are put in the first bin.
Re-ordering is done at fixed intervals of time after assessing the usage rate. The quantity ordered is
calculated to bring the stock to some predetermined maximum level. Suppliers know well in advance
when orders are going to be received. It also allows purchasing department to plan its own work.
Periodic review system has variable order quantities but fixed time intervals between orders.
Most companies have been using periodic review system because it was prohibitively expensive to be
update about inventory positions of thousands of items. The company counts the inventory for a
product and places an order at fixed time intervals – every Monday or 1st of a month.
The stock position is reviewed at fixed intervals (p), say every Monday or 1st of every month, and an
amount equal to Target inventory minus the stock position is ordered after each review.
P is the time between orders, and if d is the daily demand or usage rate, then Q or the order size or
the economic order quantity = P x d
Target inventory (T) is set high enough to cover demand during the review period (P) and lead time
(L). Such high inventory is required because new stock will not be ordered again until the next review
period (say next Monday or 1ST of next month), and the ordered stock will arrive only after the lead
time (L) is elapsed – therefore, target inventory (T) should be enough to cover demand during the
time period ( P + L)
Target inventory = T = Average demand x (Review period (P) + Lead time (L) ) + Safety stock
T = d x (P + L) + Z x σ (P + L )
Taking the above example (in continuous review system) where Q and d were calculated to be 1131
and 164 respectively
Therefore, the stock position is reviewed every 7 days, and order is placed to reach the target level of
3579.
It is worthwhile to note that the periodic review system needs a safety stock of 1611 units to achieve
a service level of 90 per cent, while the continuous review system just needs safety stock of 465 units
(refer example of Q system) to achieve the same service level. This happens because the periodic
review system must provide safety stock for (P + L ) time period, whereas the continuous review
system needs safety stock for L time period.
Single-bin system: It is the visual system version of P system. Inventory of a product is kept in a single
bin, and inventory is withdrawn from this bin. After a fixed period of time, say every Monday, the bin
is filled to the top with fresh replenishment.
Both the systems are used to manage independent demand, where σ plays a significant role. However,
there are situations in which one system is better than the other.
In periodic review system, inventory is exposed to vagaries of demand (σ ) over the entire time
period of review period and lead time, and hence stock out can happen at any time. Therefore,
to prevent stock out, safety stock is kept for the entire time period of review period and lead
time. In continuous review system, inventory is exposed to vagaries of demand (σ ) only during
lead time, and hence stock out can occur only during lead time. Therefore, safety stock has to
be kept only for lead time.
The periodic review system is used when orders have to be placed at specified intervals, say
every Monday or 1st of every month.
The periodic review system is used when multiple products are ordered from the same
supplier and delivered in the same shipment. The supplier achieves economies of scale of
transportation over multiple products.
The periodic review system is used for inexpensive products whose inventory status is not
maintained – inventory is not counted after every withdrawal. For example, bins containing
nuts or jars containing toffee are filled to top (Target inventory) at fixed intervals of time
(Monday).
The periodic review system should not be used for expensive products, since keeping high
amount of safety stock of expensive products can tie lot of capital in inventory. Therefore,
continuous review system should be used for expensive products.
The periodic review system has been popular because inventory has to be continually counted
to be able to implement continuous review system. Computers have simplified the process of
continual counting – the stock position is changed after every withdrawal, and hence it is
easier to implement continuous review system now.
Long range production plan: The company forecasts how many units of its product would it
be able to sell in coming years, and installs capacity to meet its forecast – for example, it will
sell 5 million units per year over the next five years. Long range production plan is very risky –
if actual demand does not match the forecast in coming years, the company is burdened with
extra capacity, and its cost of production increases, and if actual demand is more than the
forecast, the company foregoes profits that it could have earned if it had sufficient capacity.
Forecasting is a very important element of long range production planning.
Intermediate range plan: The company forecasts how may units of its product would it be
able to sell in the next few months – for example, it will sell 50,000 units per month over the
next few months. The company readies and tweaks its already installed capacity to meet its
forecast. This planning process is called aggregate planning – for example, a company uses
overtime to meet the extra demand or hires temporary workers to run an extra shift or does
not run second shift.
Short range plan: The company forecasts how many units of each variety of its product would
it be able to sell in the next few days or weeks, and the company makes a plan to produce
them. This is called master production schedule – for example, 100 units of X and 150 units of
Y would be produced in second shift on Thursday. Resources like machines and workers are
allocated to products – for example, a company decides to produce 50 units of X on machine
A. This process is called scheduling.
Purpose of aggregate plan is to specify the optimal combination of production rate, work force level,
and inventory at hand so that demand is met. Production rate is number of units produced per worker
per unit of time – for example, one worker produces 100 units per hour. Workforce level is the number
of workers needed to produce the required number of units: Number of units produced = Production
rate x Workforce level. Inventory on hand is unused inventory carried over from the previous period.
A typical aggregate planning problem reads like this: Given that the demand for product X is 100 units
per month over the next six months, determine the production level, inventory level, and workforce
level for each month so that cost incurred over the six months is minimum.
Aggregate planning determines the capacity that the company will need to serve demand in
the next 6 – 12 months. Aggregate plans are prepared for product lines, rather than individual
products.
Aggregate planning matches supply with demand over the medium time range up to
approximately 6-12 months into the future. It is done for a single product or a few similar
products that require similar components and processing.
A meaningful and common measurement of output has to be decided. Outputs which share
common production processes or consume similar resources can be put under one category
– for example, an automobile company will produce 1,000 units of small cars and 500 units of
sedans per month over the next six months. The units of output are small cars and sedans,
though the automobile company produces many brands of small cars and sedans.
An aggregate plan sets an overall output level to deal with demand which is inherently
uncertain. It also specifies the mix of resource inputs that would be used to achieve the
planned output level – for example, an extra shift is run or temporary workers are hired or
workers are fired.
During aggregate planning, capacity of facility is considered fixed and cannot be expanded –
for example, if installed capacity is 100 units per shift, more than 100 units cannot be
produced during a shift. Therefore, aggregate planning uses installed capacity to serve
demand in the most profitable way – for example, a company runs an extra shift or works on
Sunday.
Aggregate planning also tries to influence demand by manipulating price, advertising, and
product mix – for example, a company reduces price to increase demand if demand is
significantly less than installed capacity, or advertises profusely during low-demand periods
to shift demand from high-demand periods to low-demand periods, as do companies that sell
air conditioners and refrigerators. Therefore, operations and marketing work together to
serve demand in the aggregate planning stage.
In services, capacity is often limited by space – for example, number of rooms in a hotel
determines its capacity, and number of seats in a plane determines its capacity. Time also
affects capacity – for example, the number of customers who can be served dinner in a
restaurant is determined by the number of seats as well as the number of hours dinner is
served.
A production system works best when its rate of production is constant, and it is always expensive to
change rate of production. Product companies hold inventory to meet demand, and keep its rate of
production constant, but holding inventory is costly when the product is expensive or perishable, or
when high levels of fashion and / or technology are embedded in it. Service cannot be inventoried,
and hence if rate of demand is very different from rate of production, it either results in idle capacity
or unserved customers. Therefore, most companies should always explore the possibility of making
rate of demand stable – demand should not vary much from one time period to another.
Pricing – A company uses differential pricing to reduce peak demand or increase demand during low-
demand period – higher price is charged when demand is high, and lower price is charged when
demand is low. The idea is to shift demand from high-demand period to low-demand period so that
demand is more even throughout the time period.
Advertising – A company advertises profusely during low-demand period to stimulate demand, and
also to shift some demand from high-demand period to low-demand period.
Backlog – When a company cannot meet the demand of all its customers, it asks its customers to wait
for their orders to be delivered at a later date. Backlogging shifts demand to a future time period, but
customers may not be willing to wait, and may buy some other brand. Backlogging can be used only
by a strong brand. Backlog is confirmed demand.
Reservation – A company motivates its customers to reserve capacity – customers promise to buy in
some future time period. The company is able to know demand in advance, and hence it is able to
arrange capacity to meet demand.
The company changes its rate of production to make it equal to rate of demand. It is important to
keep in mind that a company cannot go beyond its design capacity to match rate of demand.
A company hires new employees when the demand is high, and it fires its employees when the
demand is low. This is a good strategy when the production process is mostly manual and output is
almost directly proportional to the number of employees. But, when production is mostly done by
machines, the relationship between output and the number of employees is no longer linear – output
is largely determined by capacity of machines. Therefore, for most manufacturing facilities, hiring and
firing is not a good strategy since most of them employ machines to produce their products. This
strategy affects labor relations and worker morale, which affect productivity. This strategy is becoming
useful in knowledge industry where output is again directly proportional to number of employees –
for example, software companies, law firms.
Procedure:
Determine the productivity of each employee – for example, one employee produces 10 units
of the product per hour.
Calculate the number of employees needed to produce the output – if rate of demand is 100
units per hour, the company requires 100 / 10 = 10 employees.
Hire new employees or layoff – if the company has 5 employees on its rolls, it hires 10 – 5 = 5
employees, but if it has 15 employees on its rolls, it fires 15 – 10 = 5 employees.
Hiring and layoff costs are incurred – it costs Rs. 10,000 to hire an employee, and hence the
company incurs 5 x 10,000 = Rs. 50,000 as hiring cost, and its costs Rs. 15,000 to fire an
employee, and hence the company incurs 5 x 15,000 = Rs. 75,000 as firing cost.
Problems:
Direct and indirect costs of training are incurred: New employees may have to be trained for
long periods if difficult-to-acquire skills are required to do tasks that would be assigned to
them. New employees take time to reach the desired productivity levels, and hence there is
loss of output. Also, new employees are likely to make mistakes while performing their tasks,
and hence are likely to produce defective products.
Employees of required skill may not be available: Employees with difficult-to-acquire skills
are mostly in short supply, and it is especially difficult to hire them during economic prosperity
when all companies of an industry are expanding production, and hence are hiring.
Time is required for hiring and training: It always takes time to hire employees, and then train
them, and in the meantime, customers cross over to competitors and are lost forever.
Competent employees with multiple jobs may not join the company: Employees do not like
companies which do not provide stable employment, and hence if a company develops a
reputation of firing employees when they are not required, competent employees may not
join the company.
This strategy may not work in unionized environment: Union will not allow the management
to fire employees.
Keep work force constant, but vary its utilization – use overtime and undertime to meet demand
A company uses overtime and undertime to match supply with demand. It is an useful strategy when
change in demand is temporary. A company uses overtime when its demand is more than its capacity.
For example, a manufacturer may have capacity to produce 140 tractors in two shifts of 8 hours each
– it is able to produce one tractor in (8 + 8) x 60 / 140 = 6.9 minutes. If the company forecasts that its
demand of tractors will be more than 140, it can run overtime to produce the required number of
tractors because the company has 8 hours of idle capacity in third shift. But it can only produce 8 x 60
/ 6.9 = 70 tractors during overtime – if its forecasted demand exceeds 140 + 70 = 210 tractors, it cannot
satisfy demand by producing during overtime. The advantage of overtime is that a company can meet
a fluctuating demand without increasing its capacity, but a company should not try to meet permanent
increase in demand by overtime – it should increase its capacity to meet permanent increase in
demand. If a company uses too much overtime, its machines will deteriorate faster because it will not
have enough time to carry out preventive maintenance, and its people will face burnout. Workers are
interested in overtime, because wage during overtime is significantly higher than in normal working
hours.
A company uses undertime when its demand is less than its capacity – its capacity remains
underutilized. A company does not like undertime, and the issue really is to make it less painful for
the company. For example, if a manufacturer has capacity to produce 140 tractors in two shifts of 8
hours each, but if its demand is only for 70 tractors, it can follow one of the following options: (i) it
can produce 70 tractors in one shift and close the plant in second shift (ii) it can produce 35 tractors
in first four hours of each shift (iii) it produces 35 tractors in each shift, and workers are allowed to
work at a slower pace (iv) it produces the required number of tractors in 15 days and closes the plant
for the rest of the month. The fourth option works out to be better than the first three methods as (i)
workers do not drop their pace, and hence there is no decline in productivity when demand picks up
(ii) since the plant is shut during 15 days, it does not incur any cost in running the plant for 15 days (iii)
it can plan an extensive preventive maintenance schedule during the 15 days that the plant is shut (iii)
workers come back fresh and energized after the break. It is important to remember that though
workers are not fully utilized during undertime, they are supposed to be paid their full wages.
To sum up:
Companies in certain industries face predictable surges in demand during parts of the day / week /
month / year. For example, a restaurant has high demand for four hours in the evening. Therefore, it
keeps a base level of employees and hires part time employees who work four hours in the evening.
It is able to meet peak demand without having to keep large number of employees throughout its
working hours. It is important that a company pays higher wages to part time employees so that it is
able to attract competent part time employees. It is important to ensure that the part time employees
are able to provide a high level of service, and hence the company should standardize its processes
and provide adequate training to its part time employees. It is also important that it is able to retain
its part time employees so that they do not have to be trained very frequently. Companies in certain
industries face predictable surges in demand during parts of the year. Most holiday destinations
attract tourists during only some part of the year, and hence it would not be wise to keep the same
number of employees throughout the year. A hotel in such a place can hire temporary employees
during its peak period.
Demand for most products fluctuate, but a company is always better off producing at a constant rate.
A company continues to produce at a constant rate even if its demand is less, and it stocks the product
that it is not able to sell. It uses this inventory to serve customers when its demand becomes more
than its rate of production. Therefore, inventory decouples demand from capacity, and is a safe
aggregate planning strategy for most products. But this strategy will not be effective if there are large
and predictable swings in demand between two time periods, and hence inventory has to be stored
for longer time periods which is what happens when the demand of a product is seasonal in nature.
Products like air conditioners and refrigerators are examples of products whose demand is seasonal.
Holding inventory cannot also become dangerous if demand varies wildly and unpredictably from one
time period to another – there would be stockout when the demand is high, and there would be idle
inventory when the demand is low. Holding inventory can become dangerous if a product has high
levels of technology and fashion embedded in it – the inventory can become obsolete by the time the
selling season arrives. Inventory also cannot be held if the product is perishable. This is a good strategy
for products whose demand does not change significantly from one time period to another, and
products which are not perishable, and products which will not become obsolete due to changes in
fashion and technology that are embedded in them.
To sum up:
Production is carried out at a constant rate, and hence equipment utilization is high.
Production cost is low.
Workforce is constant, and there is no overtime and idle time. Employee morale is high.
Inventory accumulates during low-demand periods, which is used to serve customers during
high-demand periods.
Inventory ties up working capital, takes space for storage, and there is risk of damage and
obsolescence.
Subcontracting
A company may use capacity of another company if its demand outstrips its capacity – it contracts
with another company to manufacture a part of its demand. This can be a good strategy when the
industry in which the company operates has overcapacity, but it will generally not be applicable in an
industry whose capacity is less than its demand, because it will be difficult to find companies with
spare capacity. The company also has to ensure that its subcontractor is able to produce quality
products – it is always better to forgo demand than serve customers with inferior products. It is also
important to remember that subcontracting is different from outsourcing – when a company
outsources to a company, it procures all its supplies, and does not manufacture at all, but when a
company subcontracts, it manufactures major part of its demand and uses the subcontractor to supply
only a part of its demand.
Companies of an industry may make arrangements among themselves to shift customers to each
other when demand of any one of them outstrips its supply. Airlines shift passengers if they are
already full, and hospitals and restaurants are known to recommend customers to competitors if they
do not have capacity to serve them. If played fairly and with some amount of generosity, this can be
a good strategy as the industry as a whole can serve larger number of customers with less total
capacity – the industry becomes more profitable.
BASIC STRATEGIES
A company produces at a constant rate which is defined by its capacity, and hence it is able
to utilize its capacity well. It serves most of its demand at its lowest possible cost. It uses
demand modifying strategies like differential pricing to shift demand from high-demand
periods to low-demand periods (to reduce the peaks and troughs in demand), and then uses
supply strategies like inventory and subcontracting to match supply with demand. Inventory
costs are high, but there are no costs of overtime, hiring, and layoff. This strategy is common
in industries where fixed costs dominate the cost structure of the manufacturing
infrastructure – utilization of the manufacture infrastructure is vital, and has direct bearings
on per unit manufacturing cost.
A company is able to produce to meet fluctuating demand, which essentially means that the
company keeps extra capacity – production rate is kept equal to demand rate. This can be a
good strategy in industries where incremental costs of installing larger capacities are small.
For example, if the cost of installing an assembly plant for 1500 cars is just 110 per cent the
cost of installing an assembly plant for 1000 cars, the company would be wise to install an
assembly plant for 1500 cars. Unfortunately, in most traditional manufacturing setups, costs
are almost directly proportional to capacity – the cost of installing an assembly plant for 1500
cars would be around 150 percent of the cost of installing an assembly plant for 1000 cars.
Keeping extra capacity in such industries can be suicidal since per unit manufacturing cost
would be high, as its utilization would be low. But, in new industries like data storage and
processing, the incremental costs of installing larger capacities are small, and companies in
such industries can afford to keep extra capacity. In industries, where the output is directly
proportional to number of employees, a company can hire and employees to match supply
with demand, but in most traditional manufacturing setups, where manufacturing is done by
machines, output is no more directly proportional to number of employees. But, in people-
based industries like law firms and software companies, where output is directly proportional
to number of employees, a company can hire and fire to match supply with demand. Inventory
costs are low, but costs of overtime, hiring, and firing are high.
Mix strategy
Demand rate is relatively constant – it does not change significantly from one time period to
another. The company keeps some extra capacity – say, 20 per cent of average demand. It
mostly produces at a constant rate, and uses inventory to serve any moderate increase in
demand. But, it is also willing to produce during overtime. And if output is directly
proportional to number of employees on its rolls, it is also willing to hire and fire to match its
production rate with demand rate.
The aggregate plan is converted into a master production schedule. Master production schedule
specifies which product would be produced when and in how many units. A master production
schedule is generally developed for a week – for example, 500 units of X and 700 unit of Y would be
produced in the first week of June. Master production schedule generally covers a 6- to 12- month
time horizon.
Demand of the individual product in various time periods is taken into consideration – for
example, of the 1,500 small cars that is expected to be sold next month, 800 would be brand
X, 400 would be brand Y, and 300 would be brand Z.
Master production schedule translates aggregate plan into production schedules for
individual products – for example, of the 1,500 small cars that would be produced next month,
800 would be brand X, 400 would be brand Y, and 300 would be brand Z.
Master production schedules specifies as to which product and how many of them will be
manufactured in which week, on which day, and in which shift – for example, brand X would
be produced during from 1st to 18th of the month, brand Y would be manufactured from 19th
to 25th of the month, and brand Z would be manufactured from 26th to 30th of the month.
Capacity and availability of machines is taken into consideration – for example, 100 units of A
have to be processed on machine M in shift 1 on Monday, but it can process only 80 units of
A.
Availability of unique components needed for a particular product is considered – for example,
component N is required to assemble brand Y.
Time fences in Master Production schedules
Master production schedule gives the quantity of end products that would be produced each week in
next 24 – 48 weeks. For example, 100 units of brand X and 200 units of brand Y would be produced in
week 1, and 150 units of brand X and 150 units of brand Y would be produced in week 2, and so on till
the next 24 – 48 weeks. The first few weeks of the master production schedule cannot be changed
except under extraordinary circumstances because it would be expensive to reverse the plans to
purchase materials and produce the components that go into the products. Production schedules for
later weeks can be changed more easily.
Master production schedule is updated weekly – after one week has passed, one week is removed
from the front end of the Master production schedule and one week is added to its back end, and
production schedule for all the weeks are estimated afresh. Production schedule of later weeks are
more likely to be changed as they are mostly forecasts rather than confirmed orders. Production
schedule of early weeks are less likely to change as they are mostly confirmed orders. Since production
schedule of all weeks are updated every week, they more accurately reflect actual demand as they
move towards the front end of the master production schedule.
A more comprehensive view of just-in-production system is to transform the way product is designed,
work is delegated, and responsibility is assigned – it seeks to eliminate all forms of waste at all stages
of the value chain. It specifically seeks to eliminate the reasons for holding inventory – long setup time,
unreliable delivery from suppliers, breakdown of machines, and defective products. Inventory covers
up for a faulty production system – inventory protects the company from shortages and work
stoppages due to long setup times, unreliable deliveries of materials, machine breakdowns, and
defective products. Hence, it strives to (i) reduce setup time (ii) ensure reliable delivery of materials
from suppliers (iii) reduce machine breakdowns (iv) ensure that no defective products are produced.
Reduction of inventory is a natural consequence of such improvements in the production – why would
a company keep inventory if it never produced a defective product, its machines never broke down,
it did not incur any setup time, and its suppliers always supplied the right amount of materials at the
right time. Such a comprehensive view of just-in-production system is called Big JIT. In sum, a just-in-
time production is a robust production system.
It is based on the logic that nothing is produced until it is needed. Need is created by actual
demand for the product.
Waste is to be avoided at all cost. Waste can be of many kinds – time wasted in maintenance,
investment locked in inventories, and idle workers.
It considers organization in terms of value chain that extends from suppliers through
transformation to the final customer.
It organizes employees in teams and makes everyone in the organization conscious of their
work.
The endeavor is to produce goods of high quality and to have continuous improvement as its
goal.
It organizes operations by product or cellular manufacturing – a work center will have all
machines that are required to make a product. For example, a work center may have machines
for boring, drilling, and shaving.
It produces the quantity of units that is needed – no more and no less.
It produces the quantity of units that is needed on date and time that they are required – not
before and not after.
Supplier delivers the exact quantity demanded, at the scheduled time and date.
Stable production schedule is required. Same products are produced in the same sequence,
in the same quantities , and in similar time periods.
It presents a system of enforced problem solving. There are no safety factors as every item is
expected to meet quality standards, every item must arrive at the right place, and at right
time, and every machine must function without breakdown.
Adhering to just-in-time principles leads to reduced lot size, reduction in inventory levels, low
production cost, and superior quality product.
Most companies cannot follow the above principles, and keep inventories at various stages of the
value chain due to following reasons:
A company keeps inventory of raw material because its suppliers are unreliable, and hence it
fears that it may run out of stock – production is stalled due to unavailability of material. It
buys in bulk to take advantage of economies of scale in purchasing, and hence has to keep
inventory.
It expects its production stages to produce some defective components, and hence keeps in-
process inventory between production stages – if it does not keep in-process inventory, it will
lose production of final products equal to the number of defective components.
It expects some of its final products to be defective, and hence keeps inventory of final
products – if it does not keep inventory of final products, it will lose sales equal to the number
of defective final products.
It schedules large production runs to spread the fixed cost of machine set up over a larger
number of units, and hence has to keep large cycling inventories.
It keeps in-process inventory between production stages to protect against delays resulting
from machine breakdowns and employee absenteeism, and also to keep production stages
running during product changeovers at other production stages.
Demand of final product is uncertain, and it keeps safety inventory to take care of variability
in customer demand.
The right way to implement just-in-time principles is to reduce inventory gradually – say 10 per cent
every month. As inventory is reduced, there would be work stoppage at some production stages due
one or few of the following problems – defective products, long setup time, machine breakdown, and
delayed supply from a supplier. The problem which causes work stoppages at maximum number of
production stages is attacked first – for example, if defective products cause work stoppages at
maximum number of production stages, the company improves its quality system. Once it has
improved its quality system, it reduces its inventory by another 10 per cent, and then attacks the next
problem that causes work stoppages at the maximum number of production stages.
Products are manufactured and pushed through the supply chain where it is then up to the
salespeople to find customers. The company makes a forecast, and develops a Master Production
Schedule (MPS), which specifies the quantity of products to be produced in various time periods – for
example, 100 units of product X and 150 units of product Y are to be produced on 26.09.2012.
Material is put on the first machine of the line flow, and each machine passes on the material
(semi-finished product) to the next machine after it has processed it. For example, if Master
production schedule specifies that 100 units of product A has to be produced on a particular
day, 100 units of raw material of product A are put on the first machine in the morning of the
day, and all the 100 units move one by one through all the machines till all of them are
processed at all the machines by the end of the day. If more than one product is to be
produced on the same line flow, their production is organized in batches. For example, after
100 units of product A has been produced, 150 units of product B are scheduled for
production, and 150 units of raw material of product B arrive on the first machine after the
line flow is set up for product B. Production is organized in batches because set up time is
incurred each time a line flow is required to produce a different product.
Rate of production of a product is unrelated to rate of demand, especially if a number of
products are produced in batches on the same line flow. For example, daily demand of product
A and Product B may be 100 and 150 units respectively, but when product A is scheduled for
production, its rate of production is much higher than 100 units per day (its daily demand),
and rate of production of product B is zero (while its rate of demand is 150 units per day.
Therefore, when multiple products are produced in batches one after another on a line flow,
rate of production of a product is either higher than its rate of demand or is zero (when other
products are being produced).
Each operator works at his own rhythm disregarding what is going upstream or downstream.
Workstations produce at different rates – workstation 1 may produce one piece in 30 seconds
while upstream workstation 2 may produce one piece in 40 seconds. Workstations are
unbalanced.
Inventory build-up creates panic – operators are stressed.
Machines run idle – operators are not fully occupied.
It is the customer’s order that triggers a demand. This demand pulls the required product through the
supply chain. When a product is sold or its inventory falls below a predetermined level, a replacement
product is pulled from the last stage of the production line. The last stage pulls a semi-finished product
from the next upstream stage, which pulls a semi-finished product from the next upstream stage –
each stage pulls semi-finished product from its upstream stage till the material is pulled from first
stage, which pulls raw materials from suppliers. Just-in-time production system is a pull production
system.
Just-in-time is a unit by unit operation where the operator upstream waits until the immediate
downstream workstation needs a unit before producing it.
Cycle time is the rate at which units are being produced on a production line. To be efficient and to
avoid unnecessary inventory, the cycle time should be equivalent to customers’ demand rate.
Push production system and pull production system differ in following ways:
In a pull system, products are produced to replenish stocks that are depleted by actual sales
– when the inventory of a product falls below a certain level, the final stage of the production
process carries out its designated process on the semi-finished product (in-process inventory)
lying with it, and produces a product. And it signals to its previous production stage that it
should send another semi-finished product – the second last production carries out its
operations, and similarly sends a signal to its previous production stage that it needs another
semi-finished product. This process continues all the way to the first production stage and the
material procurement process – the supplier receives signal that it has to send material to the
first production stage. Few points are noteworthy (i) each stage carries a semi-finished
product or in-process inventory (ii) production of a product is initiated by the final stage of
production in response to actual demand – the depletion of stock is proxy for actual demand
(iii) the final production stage pulls the semi-finished product from its preceding stage, which
pulls from its preceding stage – each production stage pulls semi-finished product from its
preceding stage and finally the first production stage sends a signal to the supplier of raw
material that it needs another raw material.
It is important to keep in mind that rate of production is strongly linked to rate of demand –
the rate of depletion of inventory is related to rate of demand, and since rate of production
matches the rate of depletion of inventory, the rate of production is related to rate of demand.
Lot size: Most production systems produce more than one product – for example, an
automobile company may manufacture three models of cars: X, Y, and Z. And since most
production systems have significant set up times, products cannot be produced in single units
– the automobile company cannot produce one unit of X, and then one unit of Y followed by
one unit of Z, because set up time would take up most of the production time. For example,
if it takes 10 minutes to process one unit of X, Y, and Z, and it takes one hour to set up the
machine, it will take 10(X) + 60(setup time) + 10(Y) + 60 (set up) + 10(Z) = 150 minutes to
produce one unit each of X, Y, and Z, of which 120 minutes is spend in setup – on an average,
it takes 50 minutes to produce one unit. Therefore, most production systems produce in
batches – for example, the automobile company produces X, Y, and Z in batch of 100. And
now it will take 10 x 100(X) + 60(setup time) + 10 x 100(Y) + 60(setup time + 10 x 100(Z) = 3120
minutes to produce 100 units each of X, Y, and Z – on an average, it take 3120 / 300 = 10.4
minutes to produce one unit. The average production time will decrease as the batch size
increases, but since the average inventory being held is half of batch size, holding cost will go
up – therefore, the tradeoff is between the average production time and average inventory
being held. But, a production system can work with smaller batch size if it has been able to
reduce its set up time, and since JIT production systems have lower set up times, it can have
lower batch sizes – production at the last stage Is scheduled in batches when variety of
products are being produced.
It is important to remember that pull or JIT production system cannot produce variety of
products if it has any significant amount of set up time, because by principle JIT production
system should have small batch size – in fact, it should produce in batch size of one. Since
most production systems have not been able to reduce set up times to close to zero, JIT
production systems work best when just one product is manufactured.
Responsiveness: Push production systems maintain inventory to meet unexpected surges in
demand, but pull production systems keep extra capacity to meet unexpected surges in
demand. Though inventory and capacity are substitutes of each other, it is important to
remember that holding inventories of products that are perishable or have high levels
technology or fashion imbued in them can be dangerous – the products deteriorate or become
obsolete, and have to sold at discounted prices or scrapped.
Highly trained employees are committed to removing waste and doing only those activities where
value is added. Empowered employees analyze every aspect of their work and squeeze out waste.
Elimination of waste
1. Defective products: Do not make defective products. Improve production process to make
perfect products rather than rely on inspection to sort out good products from bad ones.
2. Overproduction: Produce only what is needed.
3. Motion and effort: Eliminate unnecessary human motions. Reengineer the necessary human
motions so that less effort has to be expended in performing them. Try to automate as many
processes as possible – use more machines.
4. Waiting: Machines and workers should not be idle. Employ more multi-skilled people and
install more general purpose machines.
5. Transportation: Facility layout should be such that movement of materials, employees, and
customers is minimum.
6. Unneeded production processes: Production processes that do not add value must be
eliminated.
7. Work-in-progress inventory: Reduce set up time, and eliminate work-in-progress inventory.
1. Flexible resources: Multifunctional workers can operate more than one machine at a time,
and they can easily be shifted from one machine to another – this reduces idle time of workers
and machines. General purpose machines can carry out more than one operation – for
example, a general purpose machine can bore holes in an engine block, and then do drilling,
milling, and threading operations at the same station. Wastes of moving the engine to other
stations and setting up other machines at those stations are eliminated when multipurpose
machines are used.
2. Cellular layouts: A cell groups dissimilar machines to process a family of components with
similar processing requirements. Machines of a cell are arranged in a line flow, and a
component moves from one machine to the next as in a line flow. Since a cell processes similar
components, set up requirements are low, and components can be produced in small batch
sizes.
3. Pull system: In a pull system, a worker goes back to his preceding (upstream) workstation and
takes the materials that he needs to produce his output, and does so immediately. The worker
at the preceding workstation produces the materials (his output) that the worker of the
succeeding (downstream) workstation took away – the worker at this workstation will not
produce his output till the one that he has already produced is taken away by the worker of
succeeding stage to produce his own output. Therefore, the rate of production of a stage is
determined by the rate of production of its succeeding stage – there is neither overproduction
nor underproduction.
In a push system, a schedule is prepared for all the workstations, and each workstation pushes
his output to the next (downstream) workstation.
4. Small production batch size: When production is done in small batches, defective products
are quickly identified – since work-in process inventory is small, a workstation processes
output from his preceding workstation as soon as it receives it, and if the output is defective
it is immediately known. Less work-in-progress inventory makes workstations dependent on
each other, and bottleneck workstations are easily identified.
5. Quick setups: Production cannot be done in small batches if set ups are elaborate and take
up a long time. All the set up activities which can be done while the machine is running are
completed before the machine is stopped for setting it up. Further, it is explored if the set up
activities that are being carried out when the machine is stopped can actually be carried out
while the machine is running. The precise procedure in which the setup activities will be
carried out is established. Before the machine is stopped for setting it up, it is ensured that all
the required personnel and materials are available.
6. Constant production rate: Finished product is produced at a rate near its design capacity. If
production rate of finished product is high, machines that produce components would have
to operate at speeds higher than its design capacity which would lead to higher breakdown
rates. If production rate of finished product is low, machines that produce components would
run idle. Forecasts have to be accurate if production rate is to be constant, which essentially
means that forecasting has to be done closer to the actual selling period. And forecasting can
be done close to the selling period only if lead time is small.
Another way to achieve constant production rate is to arrange daily production in the same
ratio as monthly demand – at least some quantity of each product is produced every day, and
the company always has some quantity of a product available to respond to variations in
demand. For example, if demand for a company’s three products (X, Y, Z) are in the ratio 1:2:3,
production is carried out as follows: 1 X is produced, then 2 Y are produced, and then 3 Z are
produced, and then 1 X is produced, and so on. But, a company can follow such a production
plan only if there is almost no set up time, which essentially means that large number of
common components are used, and that assembly processes are very similar – product design
plays a very significant role if set time has to be reduced close to zero. A more viable approach
is this: If capacity of the production line is 120 units, 120 x 1 / (1 + 2 + 3) = 20 units of X is
produced first, then 120 x 2 / (1 + 2 + 3) = 40 units of Y is produced next, and then 120 x 3 / (1
+ 2 + 3) = 60 units of Z is produced next. Such a schedule can be managed even if there is some
amount of set up time. Such a schedule also ensures that components are produced in smaller
batches, which means that their inventory levels are low – here too, the essential requirement
is that machines that produce the components have small set up times.
7. Zero defects: It is a huge waste to produce defective products and then rework or reject them.
Quality problems are identified and solved at their source, and a defective product is never
allowed to move to the next stage. Workers, and not inspectors, are responsible for quality,
and they have the authority to stop the production line if they detect a quality problem.
8. Total productive maintenance: Lost production and poor quality products from broken down
machines can be very significant wastes. It is extremely important that machines never
breakdown as all downstream machines will run idle as soon as a machine breaks down –
because there are no work-in-progress inventories. Total productive maintenance involves
operators maintaining their own machines with daily care, periodic inspections, and
preventive repair activities.
9. Supplier network: A lean production system cannot tolerate late deliveries or supplies of
defective products as it has no inventory to take care of production till the right quality
products arrive. Therefore, suppliers are expected to deliver the ordered quantities of right
quality at just the right time. Buyers develop long-term relationships with suppliers, award
them long-term contracts, and they collaborate to improve quality and reduce cost of
components being supplied by suppliers. In return of assured business with their buyers,
suppliers locate near their buyers if possible, use small, side-loaded trucks and ship mix loads,
establish warehouses near their buyers, and use standardized containers and make deliveries
according to delivery schedule prescribed by their buyers. A certified supplier gets payment
at regular intervals rather than on delivery, and its supplies are subjected to limited amount
of quality and quantity checks.
Production Kanban
In a line flow, the rate of production is determined by the production stage that takes the
longest time – the production stage that takes the longest time is called the bottleneck stage.
For example, if five stages of a line flow have the processing times of 40, 45, 35, 50, and 40
seconds, the fourth stage is the bottleneck stage since it has the longest processing time of 50
seconds, and a product is produced every 50 seconds – it does not matter that other stages
have processing times of less than 50 seconds. It also follows that if processing time of the
bottleneck stage can be reduced, the rate of output can be increased. For example, if the
processing time of fourth machine is reduced to 40 seconds by employing a faster machine,
the bottleneck stage shifts to the second stage, and now a product can be produced every 45
seconds. It also follows that any reduction in processing times of non-bottleneck stages does
not have any influence in increasing rate of output. For example, if processing time of the first
stage is reduced to 35 seconds by employing a faster machine, it does not result in increase
of rate of output since the rate of output is still governed by the fourth stage – any reduction
in processing times of non-bottleneck stages simply result in increased waiting for products
or increased idleness for machines.
10. The bottleneck stage is the most critical stage of a line flow, and should deserve
management’s maximum attention. If management has money to invest in improving the line
flow, it should first focus on the bottleneck stage. The best operator should run the bottleneck
stage, and if any inventory has to be kept, it should be kept at the bottleneck stage. If the
bottleneck machine breaks down, it should be attended to first, since any time lost on the
bottleneck stage is time lost for the whole line flow, and hence it directly affects volume of
output.
In JIT production system, the rate of output is related to rate of demand, but if rate of
demand exceeds the rate of production as determined by the bottleneck stage, the JIT
production stage cannot feed the demand rate. For example, if the rate of demand is one
unit every 30 seconds, the JIT system cannot serve this demand since the bottleneck stage
has processing time of 50 seconds, and hence it can produce an output only every 50
seconds. Similarly, in push production system, material can be fed on the first stage at any
speed, but the rate of output can only be one unit every 50 seconds
A company has to find ways to continually improve operations – it is not restricted to improving
quality, but applies to process improvement as well. Focus is on business processes rather than
functions – a process like new product development cuts across different functions, and sometimes
even different companies. Therefore, process improvement teams are cross-functional, and they may
also include suppliers and other partners. It tries to understand as to how all the functions, suppliers,
and other partners work together to carry out the process. It involves benchmarking one’s processes
with the best ones of its industry or even outside of its industry, and trying to match their level of
excellence. It involves eliminating activities that do not add value and thus are wasteful. It also involves
instilling a sense of employee ownership in the process – it is based on the belief that people most
closely associated with a process are in the best position to identify as to how it can be improved.
Workers are trained to use methods of statistical process control and other tools for improving quality
and processes. They are also trained to use plan-do-check-act cycle for problem solving.
OPERATIONS STRATEGY
Strategy provides direction, and strategy decisions are widespread in their effect. They define the
position of the company relative to environment, and they move the company closer to its long term
goals. A company’s strategy provides answers to following two questions: (i) where is the company
heading and (ii) how will it get there. A company’s strategy should clearly delineate as to what it
intends to do for its target customers, and how will it be different from or better than its competitors
in doing whatever it is trying to do for its customers. Hence, effective strategy can be achieved in two
ways – by performing different activities than those of its competitors or by performing the same
activities better.
Competitive advantage is a company’s ability to be better than its competitors in providing value to
customers. A company can achieve competitive advantage by being better than its competitors on
quality, speed of delivery, innovativeness, and so on. A company uses its competencies like its superior
technology or its creative employees to create competitive advantage – such competencies are called
core competencies.
The company produces a few mature products (or only one) in high volumes using line flow process,
and hence its cost of production is low. Since the products are mature, there are hardly any changes
in product design or process design. It focuses on achieving high productivity and capacity utilization.
The company provides high level of performance in a few areas of concern to customers – for example,
a company may deliver its products faster than all its competitors, while product of another company
may be more innovative than the products of all its competitors. A company’s ability to differentiate
its products depends on it possessing the corresponding core competencies – the company that
delivers its products faster than its competitors has elaborate distribution structure in place, and the
company that launches innovative products has large pool of creative people among its ranks.
The company sells large variety of products. It is able to do so because it has installed large number of
different types of equipments, and it has hired large number of employees possessing different types
of skills.
The company focuses on reducing the time gap between a customer ordering a product and him
receiving it. Such a company uses information technologies to get a customer’s order into its
production system fast, keeps inventory of components, and keeps extra capacity to ensure that the
customer’s product does not have to wait to be produced, and uses fast means of transport to deliver
the product to the customer.
The company provides the most prompt and / or comprehensive after-sale services. The company sets
up its service centers near where its customers are, and uses information technologies to let
customers’ request flow into its service delivery system.
OPERATIONS SUB-STRATEGIES
A company’s operations strategy is driven by the kind of products that it offers. Product can be
classified as make-to-order, make-to-stock, or assemble-to-order. Make-to-order products are
designed, produced, and delivered to customer specifications in response to customer orders. They
are unique, and cost is generally not an issue. Important operational issues concern understanding
and meeting customers’ needs as each customer wants something unique, and delivering the product
to the customer in the shortest possible time – on-time delivery, quality, and the capability to design
and manufacture different products will determine success. Make-to-stock products are designed and
produced for average customers in anticipation of demand. Shelves are prestocked with products, and
customers select from the products available for purchase. Important operational issues concern
forecasting demand, reducing costs, and maintaining inventory to serve customers – high-volume
production, standardization in design, low manufacturing cost, and high availability of product will
determine success. Assemble-to-order products are assembled from standard components / modules
according to customer specifications. Components are made-to-stock and then assembled to order
after the customer has placed his order. Important operational issues concern minimizing the
inventory level of standard components, and delivering the product to the customer in the shortest
possible time.
Technology strategy
Technology comprises the equipment, processes, and people that are used to produce products. A
company’s technological capabilities and the way it uses technology to serve customers will
increasingly decide whether a company will succeed or not. Choice of technology affects (i) scale of
operation (ii) location (iii) level of personnel skills and training (iv) scheduling (v) tooling (vi)
maintenance (vii) and safety.
It is important that managers are always keen to explore the possibility of using new technologies to
serve customers – technology averse managers are impediments in companies’ adopting new
technologies in the ways they work and serve customers. All managers need not be engineers, but
they must be keenly aware as to how technology is weaved into their company’s strategy. Some
companies outsource their production processes, and hence lose their expertise of process
technology. It is important to keep in mind that a company’s innovation process is most effective when
product innovation and process innovation are done together – knowledge of process technology
enables designers to know if the design specifications can be produced by the existing process
technology, and they also understand how the existing or emerging process technologies can be used
to make new products.
Large number of technologies are embedded in products, and it is important that companies keep a
tab on technological developments that are happening outside its organizational borders – a company
is likely to be stumped by a new technology than anything in its competitive environment. Therefore,
it is important that a company mandates its employees to visit companies outside its industry, and
attend seminars on science and technology.
It is also important that a company employs latest relevant technologies in not only producing their
products, but also in managing other aspects of their company.
Facility strategy
It takes money and time to establish, expand, or close production and distribution facilities, and
therefore, a company must have a clearly defined capacity strategy. It is important to remember that
size, location, and production processes of a facility are related to each other – for example, if a
company’s decides to build a huge production facility instead of large number of small facilities, it will
be located centrally with respect to its markets, and it will employ a production process that will allow
it to reap benefits of economies of scale. Facility strategy is linked to marketing strategy in services –
a service provider may decide to open large number of small outlets so that its customers do not have
to travel long distances to avail its services, or it may open small number of large outlets so that it can
provide large number of services in each outlet.
Some production processes exhibit significant economies of scale – the per unit cost decreases as
production volume increases. Therefore, it is important to have a clear capacity strategy:
A company can have one large production facility, and hence it will have low per unit cost, but
its distribution cost will increase. Therefore, if a company’s production process does not
exhibit significant economies of scale, it may decide to open large number of small production
facilities, and since now each production facility will be nearer to customers, its distribution
cost would be lower.
A company’s production process may be such that a small facility is simply unviable – for
example, an automobile assembly line must be designed to produce thousands of units every
day.
In some services, customers will not travel large distances to avail the company’s services, and
hence the company must have large number of outlets close to its customers – for example,
people will not travel long distances to have meals. In some other services, customers need
large number of services at one place, and hence a company has small number of large outlets
– for example, hospitals need different types of medical and diagnostic services to attract
patients.
It takes money and time to expand capacity, and therefore a company should have a clear capacity
expansion strategy. A company has four options:
It adds capacity before it is needed – it has extra capacity to meet surges in demand. It is a
good strategy when profit per customer is high.
It lets capacity lag behind demand – it maximizes capacity utilization, and hence it operates at
low costs. It is a good strategy when production cost structure has high fixed cost and low
variable cost.
It tries to match capacity to demand – is possible when its production process is labor
intensive, and hence its capacity is proportional to the number of employees. It is a good
strategy when hiring / firing costs are low, and production processes are simple – and, hence
newly hired workers are able to reach peak productivity quickly.
It adds capacity on a regular basis, say every year – it has excess capacity when it has just
added capacity, and demand may slowly outstrip its capacity till it adds capacity in the next
round. It is a good strategy when embedded technologies of equipments that constitute the
production system are updated frequently.
Location strategy
Once a company has located a facility at a particular place, it is not easy to relocate. Therefore, a
company should have a clear location strategy.
The cost of producing a product may be low at particular location for following reasons: (ii) Good
quality raw materials may be available at low prices (ii) There may be competent suppliers of
components (iii) There may be abundant supply of cheap but competent labor (iv) Taxes may be low
(v) Government may be providing incentives (vi) Cost of infrastructural facilities like land, power and
water may be lower.
When a company has to decide the location of its facilities, it should clearly know if it is more important
to be close to its customers or it is important that its cost of production is low. If it decides that it is
important that its cost of production is low, it must locate its facilities at locations where cost of
production is low. It has been seen that most customers do not care much about where the products
that they buy are being produced, as long as their quality is acceptable. It has also been found that if
a location has significant advantages in terms of enabling low cost production, lower production cost
easily dwarfs higher distribution cost. Therefore, most manufacturers of products are locating their
manufacturing units at locations where cost of production is low.
Being spatially close to customers is important for some services, and hence service providers must
locate their outlets wherever their customers are, even when cost of production is high. Therefore, a
gourmet coffee shop must be located in the most posh locality of the city, because that is where its
customers are, even though its cost of production may be prohibitively high. In deciding where to
locate its outlets, a service provider must consider (i) Current and predicted spatial distribution of
customers (ii) Interaction among its own outlets (iii) Locations of its competitors’ outlets. Some service
providers are location leaders – they are first to locate in areas with a growing customer population.
Location leaders pays less for land and they get the best site in terms of centrality and highway access.
And since they are the first service provider in the region, they are able to establish a base of loyal
customers. Some service providers are location followers – they wait till a competitor or a
complementary service provider has opened an outlet, and then they open one of their own. Location
followers incur less search and evaluation cost, and face less risk of opening an outlet in an undesirable
area, but they have to struggle to get customers. It is also important that a service provider decides as
to how many outlets it will open in an area, and what is the time frame in which it will open these
outlets. For example, if a service provider plans to open only one outlet in an area, it will open the
outlet at a central place, but if it plans to open 5 outlets over a period of 5 years, it may not be a good
idea to open the first outlet at the central place, because when it would have opened all its outlets it
might find that a few of its outlets are too close or too far away from each other. It is the best that a
service provider decides the locations of all its 5 outlets before it opens its first outlet, and opens its
outlets according to that plan – if a service provider plans to open one outlet a year for the next 5
years, the best location for the first outlet may not be the same as it would be if the service provider
had to open just one outlet.
Companies are increasingly becoming global – they are selling their products to country markets other
than the ones they belong to, and they also have manufacturing facilities in more than one country.
Therefore, companies must have a international location strategy in place.
The costs of land, equipment, materials, and labor are different in different countries and so
are government regulations and tax rates. Therefore, a company needs to know cost of
production in different countries, and locate its production facility in the country in which its
cost of production will be low. Companies whose production processes constitute of low-skill,
labor-intensive operations should have production facility in a country where labor rate is low.
Another international location strategy that companies adopt is to develop different expertise
at different locations throughout the world. For example, an automobile company may make
engines for all its cars in one country, and sheet metals for all its cars in another country, while
the cars themselves may be assembled at multiple locations throughout the world.
Changes in exchange rates, inflation rates, wage rates, and labor productivity change the cost
of production, and it may happen that a country that has low cost of production sees
unfavorable changes in exchange rates, inflation rates, wages rates, and labor productivity,
and hence finds its cost of production going up. A company that has production facility in only
one country may be at serious disadvantage if the that country’s exchange rates, inflation
rates, and labor rates move in ways that raises the cost of production in that country. A
company that has production facilities in many countries can shift most of its production to
countries where their exchange rates, inflation rates, wage rates, and labor productivity are
moving in ways that will lower the cost of production in those countries.
Though most customers do not overtly care where the products that they buy are produced,
they do believe that that some countries have systems, practices, and values that enable
companies located in those countries to produce products of high quality. Similarly, they
believe that companies located in some countries are simply incapable of producing products
of high quality. Countries have been known to specialize in production of certain products,
and customers are keen to buy products of those companies which are located in those
countries. Some customers also like to buy products of companies which are located in the
countries where the live – swadeshi syndrome. Companies that make products for other
companies find out that the buyer companies would want the seller companies to locate
where they are – deliveries in small batches can be made quickly and they can cooperate in
multiple ways if buyers and sellers are located close to each other. Some countries have home
content laws that impose import restrictions – a country may have a law that mandates
companies located in its country to buy all its components and supplies from local companies
only.
Process strategy
A company’s production system has two dimensions: (i) Flexibility in terms of the variety and volume
of products it can produce (ii) Efficiency in terms of its cost. At one end of the flexibility-efficiency
continuum is project process. Project process has maximum flexibility, and is used to produce unique
products such as computer programs, buildings, and space shuttles. Project process cannot exploit the
production efficiencies of specialization and repetitiveness.
At the other end of the flexibility-efficiency continuum is line flow process. Line flow process can
produce a narrow range of products in large volumes at low cost. All products move through all
equipments in the same sequence. A line flow process uses specialized equipments, and each worker
carries out only a narrow set of tasks. Most line flow processes are automated. Line flow process has
limited flexibility as it can produce only a limited range of products – it is most efficient when only one
product is produced.
In the middle of the flexibility-efficiency continuum are job shop process, batch flow process, and
cellular process. These processes can produce moderate to wide range of products in small to
moderate volumes. In job shop process, similar equipments and workers with similar skills are grouped
in separate departments – all equipments and workers required to carry out a task are grouped
together. For example, in a hospital, all X-ray machines are placed together in a separate department.
A product moves from one department to another depending on what tasks needs to be carried out
on it – all products need not go through all departments, and even when they go through the same
departments, they may not do so in the same sequence. Therefore, a job shop process is very flexible
– large variety of products can be produced. For example, a patient with a broken limb will go to X-ray
department whereas a patient with stomach infection will go the ultrasound department. A job shop
process is inefficient because products wait on equipments to be processed, and equipments remain
idle – some equipments may have long waiting lines while some equipments may be idle, and even
the same equipment may sometimes have long waiting lines and sometimes may be idle. Job shop
process also has high set up times because an equipment has to be reconfigured each time a different
product has to be processed on it.
When identical products are moved from one department to another in batches, and the products of
the batch are processed one after another on an equipment, the job shop process is converted to
batch flow process. Batch flow process is as flexible as job shop process because similar equipments
are placed in separate departments as they are in job shop process, but is more efficient than job shop
process because no set up time is incurred when products of one batch are being processed on an
equipment.
When large number of products require to be processed on two or more equipments in the same
sequence, these equipments are taken out from the job shop process, and put together to form a
small line flow process. For example, if 20 per cent of products are required to be processed on
equipments C, E, and F in the sequence E – F – C, equipments C, E, and F are taken out from the job
process, and are put together in the sequence E – F – C to form a small line flow. A cellular process is
as flexible as a job shop process because similar equipments are placed in separate departments as
they are in job shop process, but is more efficient than job shop process because no waiting time is
incurred when products are passing through the small line flow process.
Therefore, the process that a company installs depends on the variety and volume of products that it
has to produce, and the order winning dimensions that it is using to compete.
Choice of process also depends on the stage of product life cycle. Products go through four stages in
their life cycle: (i) Introduction (ii) Growth (iii) Maturity (iv) Decline. Product, and the process used to
produce it change as the product moves through the stages in its life cycle. In the introduction stage,
it is not yet clearly known as to what features and functions customers want in the product, and
volume is low as large number of customers are not yet buying the product. Design changes are
frequent in the introduction stage, and hence the production system must be flexible enough to
produce the new designs. Also, alternate production methods must be tested to arrive at the
production method which would be suitable for producing the product in high volume. Therefore, a
job shop process is most suitable for a product in the introduction stage of its life cycle.
In the growth stage, product design is stable, and large numbers of customers buy the product, and
hence it has to be produced in high volume. Hence, a line flow is most suitable for a product in the
growth stage of its life cycle.
In the maturity phase, competition becomes intense, and high quality product has to be produced in
large volume at low cost. Line flow process continues to be the most suitable process, but focus in on
improving quality and reducing cost through making improvements in the way work is done at each
workstation.
In the decline phase, demand for the product goes down, and it is produced in smaller volume. Line
flow process continues to be used, but the company starts preparing for discontinuance of the
product. It considers if the line flow process can be modified to produce another product.
Quality strategy
Quality is an important order winning factor, and hence It is imperative that the production system
produces products of high quality. Product quality is strongly related to the production process that a
company employs to produce its products, and to skill and motivation of its employees. Though
standard measures like defect rates are routinely use to measure quality levels, it is important that a
company allows its customers to evaluate its product. Successful companies implement principles of
Total Quality Management, some of which are:
What constitutes quality is decided by the customer – quality is customer focused. Markers
unearth the benefits that a product should provide and the features that it should possess,
and designers incorporate these benefits and features in the design of the product.
The company knows how quality affects its long term as well as routine decisions – there is
companywide philosophy with respect to quality, and each decision that the company makes
is evaluated for its affect on quality.
Every employee feels responsible for quality – quality product is made first time and every
time.
The company believes that high product quality does not necessarily translate into high cost
– high quality and low cost can go together as defect rates are lower when quality is improved.
Employees are empowered to make changes and suggestions that improve product quality –
employees are treated with respect, and are trained and motivated to make changes that
improve quality.
Personnel strategy
A company’s personnel policies should be consistent with its choice of technology, process, and quality
strategies. For example, if a company has decided that its products will be embedded with latest
technologies, it must have a policy of hiring and nurturing the best scientists and technologists. It must
allow them to have greater control over their jobs and allow them to pursue their own interests.
Similarly, if a company expects its employees on the shop floor to make suggestions on how quality
can be improved and cost can be reduced, it must train them and empower them.
Information strategy
Acquiring and processing information quickly and accurately can be a competitive advantage. Retailers
who can know quickly what their customers are buying can reorder products that customers are
buying and discontinue buying products that customers are not buying. Companies that can know
what features and functions their customers want in their products can quickly introduce products
bearing those functions and features.
Flow of information should be clearly delineated – who should send what information to whom, and
what actions are expected of the party which receives the information. For example, when a
storekeeper learns that the quantity of a component has fallen below its reorder level, it places an
order with the supplier, who dispatches the ordered quantity in two days. It is important that a
company employs the latest information technology equipments to acquire, process, and transmit
information. It is important to remember that while it is expensive to install information technology
equipments, their operating cost is minimal, and the operating cost does not increase significantly
with the increase in data that it handles. For example, when a retailer installs a point-of-sales system
which is linked to its suppliers’ ordering systems, the system tracks sales of products, checks inventory
status of products, and places orders with products’ suppliers whenever their inventory levels drop
below reorder levels – the cost of carrying out these operations are minimal, and it does not matter
much whether the system handles few hundred products or a few thousand products.
Sourcing strategy
A vertically integrated company sells the product, makes all the components, and extracts all the raw
materials that are needed to make the components. But, most companies do not produce all the
components that go into a product. Therefore, a company has to decide as to which components it
should outsource, and which components should it produce in-house – this decision involves
competency-building, dependence, proprietary knowledge, and cost.