Annals of the CIRP Vol. 56/1/2007 -451- doi:10.1016/j.cirp.2007.05.

108
Techniques for Planning and Control Dependent on Different Types of Flexibility
P. Schnsleben
1
1
Centre for Enterprise Sciences BWI, Department of Management, Technology, and Economics D-MTEC,
ETH Zurich, Switzerland;
Submitted by J. Milberg (1), Munich, Germany
Abstract
The implementation of a particular technique for planning and control of capacities, materials, and other
resources for production is strongly dependent on what type of flexibility, in a strategic, tactical, or operations
view, is the center of attention. For decision making, two-dimensional visualizations, with the horizontal axis
and the vertical axis each representing a certain type of flexibility, have proved practical. This decision aid
proves advantageous especially when the decision is made jointly and the reasons for the decision need to be
communicated to other persons in the job shop, for example. The paper presents examples of concepts and
techniques for capacity and materials planning, and further areas where the decision aids can be used.
Keywords:
Flexibility, Decision Making, Operations Management
1 INTRODUCTION
Flexibility is the capability to adapt to new, different, or
changing requirements [1].
In the literature, the above very general definition is
specialized in various directions:
Various subtypes of flexibility. [2] distinguishes mix
flexibility, design changeover flexibility, modification
flexibility, volume flexibility, rerouting flexibility, and
material flexibility. [3] provides a comprehensive list of
types of flexibility; [4] provides a literature review.
Aspects of flexibility that mostly are meant to describe
the company. It is common to use adjectives like lean,
agile, adaptable, transformable, reconfigurable,
changeable, resilient, and the corresponding nouns.
The examples that are given for these terms are all
quite similar, and they can be understood as flexibility
potentials. See here [5], [6], [7], [8], and [9].
The implementation of a particular technique for planning
and control of resources for production is strongly
dependent on what type of flexibility, in a strategic,
tactical, or operations view, is the center of attention. For
decision making, two-dimensional visualizations, with the
horizontal axis and the vertical axis each representing a
certain type of flexibility, have proved practical. Most
people can (intuitively) understand logical connections
better if they can be represented in two dimensions, on a
plane, for example on a piece of paper. Using the
visualization, a team in a job shop, for instance, can reach
a common understanding, which is a requirement for
successful collaboration.
The paper presents examples of concepts and techniques
for capacity and materials planning.
2 TECHNIQUES FOR CAPACITY PLANNING IN
DEPENDENCY UPON FLEXIBILITY OF CAPACITY
AND FLEXIBILITY OF ORDER DUE DATE
2.1 Characteristic Features that Indicate Flexibility,
and Classes for Capacity Planning and Control
Depending on the main objectives of the enterprise, the
values for some of the characteristic features of planning
and control will differ.
The quantitative flexibility of capacity describes its
temporal flexibility. If maximum capacity utilization is
required, there will be no quantitative flexibility of
capacity.
The flexibility of the order due date indicates whether
customers (internal or external) are flexible when
stipulating the delivery due date. If maximum fill rate
and delivery reliability rate are required, there will be
no flexibility of the order due date of the production or
procurement order.
There are various techniques for capacity planning.
Figure 1 groups them into two classes, in dependency
upon the two types of flexibility just mentioned before.
Infinite loading techniques
above the line
rather simple techniques
Finite loading
techniques
below the line
rather
complicated
techniques
Quantitative
flexibility of
capacity along
the time axis
In-
flexible
Not
very
flexible
Flexible
Not very
flexible
Inflexible
(set due date)
Flexible
Flexibility of
order due
date
Order-oriented
(infinite), CRP
Kanban
CPFP
Order-oriented
(infinite),
CRP
Operations-
oriented,
Order-oriented
(finite)
Order-wise
finite
Order-wise
infinite
Corma
Order-oriented
(finite)
Constraint-
oriented
Loor
Figure 1: Classes of techniques for capacity planning in
dependency upon quantitative flexibility of capacity and
flexibility of order due date.
Infinite loading means calculating the work center
loads by time period, at first without regard to
-452-
capacity. The primary objective of infinite loading is to
meet dates as scheduled, with greatest possible
control of fluctuations in capacity requirements.
Therefore, infinite loading is important when meeting
due dates must take priority over high capacity
utilization, such as is the case in customer order
production in a job shop production environment. The
planning techniques are rather simple.
Finite loading considers capacity from the start and
does not permit overloads. To prevent overloads, the
planner changes start dates or completion dates. The
primary objective of finite loading is good use of the
capacity available through the course of time, with
greatest possible avoidance of delays in order
processing. Therefore, finite loading is most useful if
limited capacity is the major planning problem, such
as in the process industry in a continuous production
environment. Often, this condition is given in very
short-term planning, in execution and control. The
planning techniques are rather complicated.
In addition to quantitative flexibility of capacity, there is
also qualitative flexibility of capacity:
The qualitative flexibility of capacity determines
whether capacity can be implemented for various or
for particular processes only. If a company puts the
focus on flexibility in the utilization of resources, then
qualitative flexibility of capacity (employees and the
production infrastructure) is absolutely necessary.
If there is qualitative flexibility in capacity, this can
increase its quantitative flexibility. For example, if
employees can be moved from one work center to
another, this is the same as if each work center showed
quantitative flexibility in assigning employees.
2.2 Techniques for Capacity Planning
In addition to these two classes of techniques, Figure 1
groups techniques for scheduling and capacity
management in nine sectors in dependency upon
quantitative flexibility of capacity and flexibility of the order
due date. The techniques shown were chosen because
they are readily available today, in many cases also in
software packages such as SAP. The task is to identify
the correct technique in accord with the desired type of
flexibility.
For details of the different techniques, see [10] or [11].
CRP stands for capacity requirements planning,
particularly in connection with software (there exist also
some variations of CRP). CPFP is the abbreviation of
cumulative production figures principle, Loor for load-
oriented order release (see also [12]), Corma for capacity-
oriented materials management.
The techniques can be compared with respect to their
overall capacity planning flexibility.
Overall capacity planning flexibility is defined as the sum
of the quantitative flexibility of capacity along the time axis
and the flexibility of the order due date.
Note that there is no technique in the three sectors at
top right of the figure: Here, the overall capacity
planning flexibility is high enough to accept and
execute any order at any time. This case is very
advantageous with regard to capacity planning, but it
is usually too expensive due to overcapacity.
Note the numerous techniques in the three sectors
from top left to bottom right. Here, there is sufficient
overall capacity planning flexibility in order to allow a
computer algorithm to plan all the orders without
intervention by the planner. After completion, the
computer program presents unusual situations to the
planner. The planner will intervene in order to execute
appropriate planning measures perhaps daily or
weekly.
Note that there are few techniques in the two sectors
above and to the right of the bottom left sector. Here,
there is no flexibility on one axis and only low flexibility
on the other. Thus, there is little overall capacity
planning flexibility. Planning takes place order for
order (order-wise). Each new order must be
integrated individually into the already planned orders.
In extreme cases, the planner may have to intervene
following each operation and change set values for
planning (completion date or capacity). Already
planned orders may have to be re-planned. This
procedure is usually very time consuming and is
therefore efficient only for orders with considerable
added value.
Finally, note that there is no technique in the sector at
bottom left. Here, there is no flexibility of capacity or
due date. As a consequence, there can be none of the
required balancing, and the planning problem cannot
be resolved.
3 TECHNIQUES FOR MATERIALS PLANNING IN
DEPENDENCY UPON THE FREQUENCY OF
CUSTOMER DEMAND, THE ITEMS UNIT COST,
AND THE ORDER PENETRATION POINT (OPP)
3.1 Characteristic Features that Indicate Flexibility,
and Classes for Materials Planning and Control
Depending on the main objectives of the enterprise, the
values for some of the characteristic features of planning
and control will differ.
Frequency of customer demand means the number of
times within defined observation time periods that the
entirety of the (internal or external) customers demand
a product or product family. If the company puts the
focus on adaptation of the product to the customers
specific requirements, then the demand pattern of a
product (family) will be discontinuous.
An items unit cost is defined as the total cost for
producing or purchasing one unit of measure of the
item, e.g., one part, one gallon, one pound. It includes
labor, material, and overhead cost. For low-cost items,
the quality of materials planning is less important than
for high-cost items. Thus, there is more flexibility in
the choice of a planning technique.
The (customer) order penetration point (OPP) is a key
variable in a logistics configuration. It is the point in
time at which a product becomes earmarked for a
particular customer. Downstream from this point, the
system is driven by customer orders; upstream
processes are driven by forecasts and plans (see [2]).
Thus, the OPP indicates the flexibility in order
fulfillment. If the customer tolerance time is at least as
long as the cumulative lead time, the product can be
engineered, procured, produced, or delivered when
actual demand in the form of a customer order is
placed. Otherwise, all goods (such as semifinished
goods, single parts, raw materials, and information)
from which the end product cannot be manufactured
and delivered within the customer tolerance time must
be ordered before there is known demand. The goods
must all be procured and stocked on the basis of
demand forecast. If the customer tolerance time is
zero, the end product must be procured before
demand is known. This is the same as stocking the
end product in a warehouse.
In materials planning, there are two different
classifications to consider. Classification of demand
according to accuracy is defined as follows:
-453-
Deterministic demand is demand downstream from
the (customer) order penetration point (OPP).
Stochastic demand is demand upstream from the
(customer) order penetration point (OPP).
Classification of demand according to accuracy is thus
dependent upon the relationship between the customer
tolerance time and (cumulative) lead time. Classification
of demand according to its relationship with other demand
is defined as follows (according to [2]):
Independent demand is the demand for an item that is
unrelated to the demand for other items.
Dependent demand is demand that is directly related
to or derived from the demand for other items.
Company-external demand, or (customer) demand for
end products or service parts, is independent demand, as
is also a companys own internal demand for office
supplies or partly indirect materials. The demand for
assemblies, semifinished goods, components, raw
materials, and in part auxiliary materials are examples
of dependent demand.

3.2 Techniques for Materials Planning
Figures 2 and 3 distinguish among techniques of detailed
planning in materials management that are readily
available today, in many cases also in software packages
such as SAP. First of all, Figure 2 classifies materials
planning techniques according to the characteristic
features frequency of customer demand and unit cost of
items.
Frequency of
customer demand
low high
unique
disconti-
nuous /
lumpy
conti-
nuous /
regular
rather
complicated
techniques
rather simple
techniques
(Item)
unit cost
CPFP (Cumulative Production Figures
Principle)
Kanban
Order point
additional classification necessary
see Figure 3.
Figure 2: Classification of detailed planning techniques in
materials planning.
According to Figure 2, demand for low-cost items (with
the exception of unique demand) or demand for high-cost
items with a continuous or regular demand pattern is
determined using stochastic techniques.
In general, forecasting techniques determine future
demand analytically or intuitively. Once demand has
been forecasted, different stochastic planning
techniques are available, and all of them are relatively
simple. Again, CPFP is the abbreviation of cumulative
production figures principle.
Dependent demand is calculated as if it were
independent demand that is, ignoring its possible
derivation from independent demand.
For low-cost items, a very high service level has
priority. This holds especially in the event where the
item appears on the bill of material with many
components. Due to the low carrying cost involved,
low stock inventory is of secondary importance.
For high-cost items, short lead times in the flow of
goods, meaning rapid processes, take priority,
requiring simple data and control flow. Inventory is
possible: the continuous or regular demand pattern
guarantees a future demand (for end products: a
customer order) within a short time. However,
because of the high unit cost, the inventories should
be low, which generally requires small batch sizes.
Downstream from the order penetration point (OPP),
the customer order can also show specific features
that lead to a custom-made product (keywords mass
customization, generic Kanban, and
postponement). The demand for this custom-made
product is then actually sporadic, while the demand for
the underlying product family is continuous.
Overall, simple planning and control techniques require,
as shown in Figure 2, low cost items or at least
continuous frequency of customer demand. In the case of
dependent demand for expensive components, more
continuous demand can be achieved through reducing lot
sizes, for example, but also through a product concept
with fewer variants or even standard components.
Simpler techniques can then be implemented in the place
of more complicated techniques of materials
management. To do this, some important methods were
developed within the lean / JIT concept. They lead to
production or procurement with frequent order repetition.
Continuous frequency of customer demand allows
production or procurement with order release according to
consumption, or a simple (stock) replenishment.
For all other kinds of demand, that is, for unique demand
or demand for high-cost items with a lumpy demand
pattern, rather complicated planning techniques are
necessary. For these, Figure 3 shows an additional
classification of planning techniques, this time according
to the accuracy of the demand and its relationship with
other demand (see the definitions above).
Accuracy of the demand
Relationship
with other
demand
dependent
demand
(e.g. for
components)
stochastic
demand
(for items
upstream
from or at the
order pene-
tration point)
deterministic
demand
(for items
downstream
from the
order pene-
tration point)
customer
order
situation to
avoid when-
ever possible
(unsatisfactory
results)
good situation
intuitive or
heuristic
technique
MRP
(used as a
stochastic
technique)
MRP
independent
demand
(e.g. for
end products)
2 1
3 4
Figure 3: Additional classification of detailed planning
techniques in materials planning for unique demand or
demand for high-cost items with a lumpy demand pattern.
The combination of the particular two values on the two
axes of Figure 3 results in the following situations:
1. Deterministic independent demand can be met
according to the actual demand that is, the customer
order. From this perspective, customer order
processing and customer blanket order processing are
techniques for determining deterministic independent
-454-
demand and thus in a sense also belong to
deterministic materials management.
2. Deterministic dependent demand can be calculated
from higher level independent demand. The algorithm
called MRP (material requirements planning) uses
explodes of the bill of material that is, the product
structure into its components. This type of demand
calculation is relatively complicated. However, due to
the priority of both high delivery reliability rate and low
or even no inventory, it is appropriate.
Thus, although they are sometimes rather complicated,
planning techniques for items downstream from the order
penetration point with unique demand or for high-cost
items with a lumpy demand pattern present no great
difficulty. However, planning of such items upstream from
the OPP generally leads to unsatisfactory results:
3. Stochastic independent demand is determined more
or less intuitively. As demand is lumpy, forecasting
techniques tend to be inaccurate and therefore ask for
a lot of additional intuition. The materials management
technique is often a manual procedure performed by
the scheduler using a very personal heuristic. It is
often a risky technique that should be avoided
whenever possible.
4. Stochastic dependent demand is derived by quasi-
deterministic techniques. Here, independent demand
is determined using demand forecast techniques.
Calculation of dependent demand is then based on
independent demand by means of explosion of the bill
of material. This is called quasi-deterministic explosion
of the bill of material. As this demand pattern requires
forecasting, there is a risk of a low service level or
high carrying cost due to capital costs or depreciation
as a consequence of technical obsolescence or
expiration due to perishability. As a consequence, any
materials management technique handling this case
will generally yield unsatisfactory results. Therefore, it
should be avoided whenever possible. However, for
many businesses, being in that situation is a fact of
life.
Overall, Figure 3 points out the reasons for a good
situation or a situation that should be avoided whenever
possible in materials management. Following Figure 3,
the situation becomes better the further upwards that
the order penetration point (OPP) can be set. As the
assumption must be that the customer tolerance time
does not lengthen, the cumulative lead time must be
reduced.
It is interesting to consider that at least in case 4, due to
the dependent nature of the demand many or all of the
value-adding processes are under the control of the
company. A thorough analysis of these processes can
lead to appropriate modifications that entail more items
downstream from the OPP, or a more continuous demand
pattern both situations being desirable. Some of the
best-known lean / JIT methods and techniques are
precisely those that increase the potential for short lead
times as well as a more continuous demand pattern.
4 SUMMARY AND FURTHER AREAS FOR USE
This paper showed how two-dimensional visualizations
can aid decision making on the use of techniques in
capacity and materials planning in dependency on
different types of flexibility. These decision aids can also
be used successfully in some additional areas in
strategic, tactical, and operations planning & control of
value adding, such as, for example:
Design of the production infrastructure, for selecting
different production types, such as batch production,
mass production, mass customization, or one-of-a-
kind production.
Supplier portfolios in strategic procurement, for
distinguishing among market-oriented relationships,
buyer-dominated relationships, supplier-dominated
relationships, and partnership relationships of a
company with its suppliers.
Measures for implementing Supply Chain
Management, in dependency on project phase and
organizational level in the company.
Facility location planning in production, distribution,
and service networks, for distinguishing among
centralized production for the global market,
decentralized production for the local market, in part
centralized production for the local market, and in part
decentralized production for the global market.
For further details on the use of the techniques in these
areas, see [10] or [11].
5 REFERENCES
[1] Merriam-Webster, 2003, Merriam-Websters Colle-
giate Dictionary, Eleventh Edition, Springfield, MA
[2] APICSThe Educational Society for Resource
Management, 2004, APICS Dictionary, Eleventh
Edition, Alexandria, VA
[3] Gottschalk, L., 2006, Flexibilittsprofile Analyse
und Konfiguration von Strategien zur Kapazitts-
anpassung in der industriellen Produktion,
Dissertation ETH Zurich, Nr.16333, Zurich.
[4] Toni A., de Tonchia S., 1998, Manufacturing
Flexibility: a literature review, International Journal of
Production Research, 36/6:1587-1617.
[5] Koste L. L. and Malhotra M. K., 1999, A Theoretical
Framework for Analyzing the Dimensions of
Manufacturing Flexibility, Journal of Operations
Management 18/1:7593
[6] Freiheit, T., Shpitaini, M., Hu, S., Koren, Y., 2003,
Designing Productive Manufacturing Systems
Without Buffers, Annals of the CIRP, 52/1:105-108.
[7] Nyhuis, P., Kolakowski, M., Heger, C., 2005,
Evaluation of Factory Transformability. In:
Proceedings of the 3rd International CIRP-
Sponsored Conference on Reconfigurable
Manufacturing , May 10-12, 2005, University of
Michigan Ann Arbor (only electronic version)
[8] Wiendahl, H.-P.; Hernndez, R., 2001, The
Transformable Factory - Strategies, Methods and
Examples. In: Proceedings of the 1st International
CIRP-Sponsored Conference on Agile, Reconfigu-
rable Manufacturing, May 20-21, 2001, University of
Michigan Ann Arbor (only electronic version)
[9] Koren, Y., Heisel, U., Jovane, F., Moriwaki, T.,
Pritschow, G., Ulsoy, G. and Van Brussel, H., 1999,
Reconfigurable Manufacturing Systems, Annals of
the CIRP, 48/2:527-540.
[10] Schnsleben, P., 2007, Integral Logistics Manage-
ment Operations and Supply Chain Management
in Comprehensive Value-added Networks, 3rd
edition, Auerbach Publications, New York
[11] Schnsleben, P., 2007, Integrales Logistik-
management Operations und Supply Chain
Management in umfassenden Wertschpfungs-
netzwerken, 5th edition, Springer, Berlin,
Heidelberg, New York
[12] Wiendahl, H. P., 1995, Load-Oriented Manufacturing
Control, Springer, Berlin, New York