FLEXIBLE

MANUFACTURING
Click to add text

SYSTEM

1
What is Flexible Manufacturing System?
A flexible manufacturing system (FMS) is a form of flexible automation in which several machine tools are linked
together by a material-handling system, and all aspects of the system are controlled by a central computer.

To qualify as being flexible, an automated

manufacturing system should satisfy the
following four tests of flexibility:
1. Part-variety test. Can the system process
different part or product styles in a mixed model
(non-batch) mode?
2. Schedule-change test. Can the system readily
accept changes in production schedule,
that is, changes in part mix and/or production
quantities?
3. Error-recovery test. Can the system recover
gracefully from equipment malfunctions and
breakdowns, so that production is not completely
disrupted?
4. New-part test. Can new part designs be
introduced into the existing part mix with
relative ease if their features qualify them as
being members of the part family for
which the system was designed? Also, can
design changes be made in existing parts
without undue challenge to the system?
2
3
 What are The Features of FMS?
An FMS is distinguished from an automated production line by its
ability to process more than one product style simultaneously.
At any moment, each machine in the system may be processing a
different part type.
FMS can let us make changes in production schedule in order to
meet the demands on different products.

4
New product styles can be introduced into production with an FMS, so long
as they are to be used on the products that the system can process.
This kind of system is, therefore, ideal when there are likely to be changes in
demands.

5
Distinguishing Characteristics:
An automatic materials handling subsystem links machines in the
system and provides for automatic interchange of work pieces in
each machine
Automatic continuous cycling of individual machines
Complete control of the manufacturing system by the host
computer
Lightly manned, or possibly unmanned

6
There are three levels of manufacturing
flexibility.
1-Basic Flexibilities
Machine flexibility - the ease with which a machine can process various operations
Material handling flexibility -a measure of the ease with which different part types
can be transported and properly positioned at the various machine tools in a system
Operation flexibility - a measure of the ease with which alternative operation
sequences can be used for processing a part type

7
2-System flexibilities:

Volume flexibility
Expansion flexibility
Routing flexibility
Process flexibility
Product flexibility

8
3-Aggregate flexibilities

Program Flexibility

Production-Flexible options are a type of flexibility option. These options are the options that provide room for
Production Flexibility change in input or output in production. These options are of high importance in industries that allow an extension
in the product’s lead time, commonly over many years.  (time between an order and its delivery).

Market Flexibility

9
Major historical developments
 Weaving Looms with paper tapes,
 NC machines with paper tapes
 Hard wired NC machines
 Computer controlled NC machines (CNC)
 Direct Numerical Control (DNC)

10
 Components of FMS Systems
 Robotics
 Material Handling /
Transport
 Machines
 Manual / Automated
Assembly Cells
 Computers
 Controllers
 Software
 Networks

11
12
Benefits of FMS
FMS systems are intended to solve the following problems:

Reduced work in process

Increased machine utilization
Better management control
Reduced direct and indirect labor
Reduced manufacturing lead-time
Consistent and better quality
Reduced inventory

13
The Disadvantages of FMS
Expensive, costing millions of dollars
Substantial pre-planning activity
Sophisticated manufacturing systems
Limited ability to adapt to changes in product
Technological problems of exact component positioning and precise
timing necessary to process a component

14
Future Benefits of FMS
Technology will make 100% inspection feasible
Computer diagnosis will improve estimation of machine failure, and guide work
crews repairing failures
The use of robots that have vision, and tactile sensing
Minimum human labor in manufacturing systems
More sophisticated tools with increased computing power
Better management software, hardware, and fixturing techniques
Developed standards that will let us install new machines easily

15
Future Benefits of FMS
Reduced marketing of products
Custom orders for customers will be made immediately with exact specifications
Improved network systems between manufacturers and suppliers

16
Differences Between FMS and FMC
FMS
Has four or more machines
Larger and more sophisticated
computer control system
Minimized effect of machine
breakdowns
FMC
Has two or three machines
Simpler computer control system
Limited error recovery by fewer
machines
17
In studying FMS, we need to keep in mind what Peter Drucker said: "We
must become managers of technology not merely users of technology".

18
FMS analysis techniques can be classified into:

(1) deterministic models, (2) queueing models, (3) discrete event simulation, and (4) other approaches,
including heuristics
Deterministic models are useful in obtaining starting estimates of system performance (estimates of system
parameters such as production rate, capacity, and utilization )

Queueing models can be used to describe some of the dynamics not accounted for in deterministic
approaches. These models are based on the mathematical theory of queues.

Discrete event simulation probably offers the most accurate method for modeling the specific aspects of
a given flexible manufacturing system. The computer model can be constructed to closely resemble the
details of a complex FMS operation. Characteristics such as layout configuration, number of pallets
in the system, and production scheduling rules can be incorporated into the simulation model.

Other techniques that have been applied to analyze FMS design and operational problems include
mathematical programming and various heuristic approaches.

19
Bottleneck model
Important aspects of FMS performance can be mathematically described by a deterministic model called the
bottleneck model, developed by Solberg

The term bottleneck refers to the

fact that the output of the
production system has an upper
limit, given that the product mix
flowing through the system is
fixed.

The model can be applied to any

production system that possesses
this bottleneck feature, for
example, a manually operated
group technology cell or a
production job shop.

It is not limited to flexible

manufacturing systems

20
Quantitative Analysis of
Bottleneck model
 Bottleneck Model
• In this model, output of production system has
an upper limit, given that the product mix
flowing through the system is FIXED.

1. Product mix: P

p 
j 1
1.0
pj = fraction of total
j system output of
style j.
P = total number of different parts styles made in
the FMS during period of interest.
2. Workstation and Servers: it is possible
to have two or more machines capable
of performing the same operations.
si  number of servers at workstation i,
i= where
1,2……n.
3. Process routing: the process routing defines
sequence of operations, the w/stn at which
they are performed and associated
processing times. tijk  Processing time
which is total time that a production unit
occupies a given w/stn server. i= station; j=
part or product; k= sequence of operation
4.Work handling system: work handling system is designated as n+1 and
sn1  number of carriers in the FMS handling system.

5.Transport time: t n1  the mean transport time required to move a part
from one w/stn to next stn in the process routing.
6.Operation frequency: it is defined as the expected number of times a
given operation in the process routing is performed for each work unit.

eg. An inspection is performed once every four units, the freq = 0.25.
f ijk  operation freq for operation k in the process plan j at stn i.
1. FMS Operational Parameters
The average work load for the given station
is defined as the mean total time spent at
the station per part. It is calculated as:
WL   t ijk f ijk p j
i
j k

WLi = average work load for station i

(min),
tijk
 processing time for operation k in
process
the j at station i
f ijk(min),
 operation freq for oper k in part j at stn
pj =i.part mix fraction for part
j.
The average number of transports is equal to
the mean number of operations in the
process routing minus one.
n   f pj 1
t i j ijk
k

nt  mean number of transports

Computing the work load of the
handling WL n  1  n t t n 
system:
1
WLn1  work load of handling system
(min),
nt  mean number of transports
Determining nt

Consider a manufacturing system with two stations: (1) a load/unload station and (2) a
machining station. The system processes just one part, part A,
so the part-mix fraction pA = 1.0.
The frequency of all operations is fiAk = 1.0.
The parts are loaded at station 1, routed to station 2 for machining, and then sent back to station
1 for unloading (three operations in the routing). Using Equation (19.3),

Looking at it another way, the process routing is (station 1) S (station 2) S (station 1). Counting
the number of arrows gives the number of transports: nt = 2

27
 2. System Performance Measures
Assumptions: 1. FMS producing at max possible rate; 2. Rate is constrained by
bottleneck station in the system (highest workload per server).
Work load per server is

WL i / si
The bottleneck is identified by finding max value of the ratio among all stations.

Let WL * , s * , t * equal to WL, No. of servers, processing

time for the bottleneck station resp
FMS max production rate of all parts
is s
*

*p

R WL *

The above equation is valid if product mix

is constant.
Individual part production rates can be
*
obtained
by multiplying R by the respective part
p

mix ratios.
R
*
*pj
 p
s*
j R  
p p j
WL
*

*pj

max prod rate of part style j
R
and pj =(pc/min)
part mix fraction for style
j
The mean utilization of each workstation is
the proportion of time that the servers at
the station are working and not idle.
*
WL i s
U i  WL i Rp
*
.
si s i WL
 *

Ui  Utilization of station i, WL  workload

station
i
of servers at
i (min/pc), s i * number of
workstations i, and R p  overall
rate (pc/min). The utilization of the
production *
station is 100% R p
bottleneck
at
The average station utilization
including transport system as
n1

U i 1
i
U 
n 1 an unweighted U  is
of all workstations utilization.
average
Useful measure is overall FMS utilization
which is based
n
on number of servers at each
station  s iU i
i 1
U s  n
U  overall FMS

s
s
i
1
utilization
i
Number of busy servers at each station is
*
s
 *
BS i  WLi Rp  i
WL*
BS WL
i
number of busy servers on average
at i, and WL  workload at station
station i

i.
 Problem
A flexible machining system consists of a load/unload station and two machining workstations. Station 1 is the load/unload
1 milling and consists of three identical CNC milling machines.
station with one server (human worker). Station 2 performs
Station 3 performs drilling and consists of two identical CNC drill presses. The stations are connected by a part-handling
system that has two carriers. The mean transport time is 2.5 min. The FMS produces three parts, A, B, and C. The part-mix
fractions and process routings for the three parts are presented in the table below. The operation frequency fijk = 1.0 for all i, j,
and k. Determine (a) maximum production rate of the FMS, (b) corresponding production rates of each product, (c)
utilization of each station, (d) average utilization of the processing stations, and (e) number of busy servers at each station.
 Solution
a) WL   t ijk f ijk p j
i
j k

n   f pj 1
t i j ijk
k
WL n1  n tt n
1
 *p
 s**
R WL

Individual part production rates can be obtained by

multiplying
s*
R*
pj  p j R  
*
p  p j
WL
*

b)
 WL i WL i
Ui  
si R *
p . s
s i WL*
 *
 *
s
BS i  WLi R
*
p  i
WL*
 WL
Extended Bottleneck Model
The extended bottleneck model replaces the assumption made in
the original bottleneck theory of a system with 100% utilization, with
a model that assumes a closed queuing network in which there are
always a certain number of workparts in the FMS.
This number of parts in the system is depicted as N, and the
assumption is made that when one part is exiting the system, a raw
part is entering; thus N remains constant.
If N is small, then there will be some idle time for some stations
owing to starving. If this is the case, then the production rate of the
FMS will be less than R p*.
If N is large, then the system will be fully loaded with parts, and
there will be queues of parts waiting in front of workstations to be
processed.
In this case, Rp* does provide a good estimate of the production
capacity of the system, although with a large work-in-process and
high manufacturing lead time.
work-in-process corresponds to N, and manufacturing lead
time (MLT) becomes the sum of processing times at the
workstation, plus waiting time.

Tw is the mean waiting time experienced by a part due to

queues at the stations

The two cases are:

Case 1 When N is small, production rate is less than the bottleneck case
because the bottleneck station is not fully utilized. The waiting time (Tw) is
theoretically zero, which gives the equation:
MLT1 denotes manufacturing lead time for case 1. Production rate can be
estimated, using Little’s formula:

Production rates of the individual parts are given by:

Case 2
When N is large, the estimate of maximum production rate made previously hold
true:

where * denotes that production rate is constrained by the bottleneck station.

Production rates of the individual products are given by:

Using Little’s formula, manufacturing lead time for case 2 is:

 The mean time a part spends in the system is estimated as:

The decision to use either case 1 or case 2 depends on the value of N. The dividing line
is determined by recourse to the following formula:

where N* is the critical value of N. If N* is greater than N then we apply case 1; if

N* is less than or equal to N then we apply case 2.
An FMS consists of four stations. Station 1 is a load/unload station with one server. Station 2
performs milling operations with three servers (three identical CNC milling machines).
Station 3 performs drilling operations with two servers (two identical CNC drill presses).
Station 4 is an inspection station with one server that performs inspections on a sampling of
the parts. The stations are connected by a part handling system that has two work carriers and
whose mean transport time = 35 min. The FMS produces four parts. A, B, C, and D. The part
mix fractions and process routings for the four parts are presented in the table below. Note that
the operation frequency at the inspection station (fijk) is less than 1.0 to account for the fact
that only a fraction of the parts are inspected. Determine: (a) maximum production rate of the
FMS, (b) corresponding production rate of each part, (e) utilization of each station in the
system, and (d) the overall FMS utilization.
Use the extended bottleneck model on the data given in previous Example to
compute production rate, manufacturing lead time, and waiting time for three,
values of N:(a)N = 2, (b) N = 3,and (c)N = 4.
 Sizing the Flexible Manufacturing System

The bottleneck model can be used to calculate the number of servers required at each workstation to achieve a
specified production rate.
To make the computation, we need to know the part mix, process routings, and processing times so that
workloads may be established. Given the workloads, the number of servers required at station ‘I’ is:

Because the number of servers at each workstation must be an integer, station utilization
may be less than 100% for most if not all of the stations.
Suppose the part mix, process routings, and processing times for the family of parts to be machined on a
proposed FMS are those given in previous example. Determine how many servers at each station i will be
required to achieve an annual production rate of 60,000 parts/yr. The FMS will operate 24 hr/day, 5 day/wk.
50 wk/yr. Anticipated availability of the system is 95%.