Designing prototypes of JIT systems using
system dynamics principles
Julio Macedo
University of Montreal, Business School,
and Institut de Stratégies Industrielles,
229 Forest, Pincourt, PQ., J7V8E8, Canada
Rafael Ruiz Usano
University of Sevilla, School of Industrial
Engineering, R. Mercedes, 41012 Sevilla,Spain
Abstract. American and European firms use just in time
prototypes which are subsets of the japanese just in time
system. Hence, these firms need tools to select those
components of ._ the japanese JIT system best suited to their
particular characteristics and environments. This paper
proposes one of such tools and illustrates its application on
a case study. The proposed tool is a computer aided procedure
that implements system dynamics principles using a continuous-
discrete simulation language.
1. Introduction.
The japanese just in time (JIT) manufacturing system
consists in most of the changes shown in table 1. A just in
time prototype (JITP) is a subset of the japanese JIT system
(Shingo, 1981). The JITP includes those JIT components best
suited to the particular characteristics of the firm and its
environment. Empirical surveys (Gilbert, 1990; Voss, 1987) and
detailed case studies (Voss, 1990; Piper, 1990; Celley, 1986)
show that American and European firms implement JITPs rather
than JITs.
Current tools to design JITPs include “detailed checklists
(Bartezzaghi, 1989) and cause-effect diagrams (Mizuno, 1988;
Fukuda, 1989). However, these tools do not capture. the
dynamics due -to the interactions of the system components.
These interactions can be generated by feedbacks that are
difficult to detect and can be missed by the managers
(Sterman, 1989).
In section two of: this paper, a computer aided procedure
that overcomes these shortcomings is presented. In section
three; this procedure is applied to design a JIT prototype on
a case study.
2. A computer aided procedure to design JIT prototypes
Figure 1 shows the proposed procedure to design a JITP. This
procedure is a computer aided learning process that focuses
the current behavior of the manufacturing system on the
Page-319
Page 320
System Dynamics '91
Table 1. Current successful changes In the organization of a manufacturing
system. Most of these changes are elements of the J.1.T. system
Nature of the change
Reference for the detalled
design of the change
1) Product and manufacturing process analysis
1.1 Product value analysis and simplification
1.2 Simultaneous engeneering design of the product
2) Flow manufacturing
2.1 Group technology and parts commonality
ze Manufacturing cells
2/6 Automation with a human touch
2.7 Small in line movable machines
2.8 In house equipement and site technology
3) Operations analysis
3.1 Set-up time reduction
3.2 Standard operations and value added motions
4) Human factors
4.1 Principles of orderliness and cleanliness
4.2 Multiprocess handling and muliskiled operators
4.3 Mutual line assitance
4.4 Efficiency control using standard times
4.5 Zero defect quality control system
4.6 Employee involvement in continuous problem solving
4.7 Total productive maintenance
48 Visual control and signal boards
4.9 Focused plant human organisation
5) Production planning and Information flows
5.1 Fractionned lots production planning
5.2 Mixed lots production planning
5.3 Synchronised scheduling and cycle control
5.4 Undercapacity scheduling
5.5 Pull-Kanban production planning
5.6 Push-M.R.P. production planning
5.7 Shop floor planning control
5.8 Developping the suppliers
5.9 Permanent materials kit and standardised containers
6) Automation
6.1 Computer aided design
6.3 Robotics techno
6.4 Automated material handling
6.5 Computer process control and automated testing
6.6 Automated assembly systems
6.2 Numerical control machines and computer aided manufacturing
‘SUZUE T, (1990)
‘SUZUE T. (1990)
BURBIDGE JL. (1975)
MONDEN Y. (1983)
HARMON R. (1
)
HIRANO R. (1988)
HIRANO R. (1988)
KOBAYACHI I. (1990)
SHINGO S. (1985)
JM. A. (1990)
SUGIYAMA Y. (1989)
. (1988)
KOBAYACHI |. (1990)
SHINGO S. (1986)
JLHLR.A (1988)
NAKAJIMA S. (1989)
HIRANO H. (1990)
HARMON R. (1990)
WANTUCK K, (1989)
‘SUZAKIK. (1987)
WANTUCK K. (1989)
SHINGO S. (1981)
MONDEN Y. (1983)
FOGARTY D. (1983)
GROOVER M. (1984)
SUZAKIK. (198;
HARMON R. (1980)
GROOVER M. (1984)
GROOVER M. (1984)
GROOVER M. (1984)
GROOVER M. (1984)
GROOVER M. (1984)
GROOVER M. (1984)
System Dynamics '91 Page 321
desired one. The convergence towards the desired behavior is
obtained by the progressive agreement of the suggestion team
and the approval committee. The suggestion team is formed by
the individuals related to the sources of the current behavior
of the system. The approval committee includes all the
managers that possess the necessary background for accepting
or rejecting the proposals of the suggestion team. These two
groups work independently but they have at least one common
member in order to ensure that the suggestions are in
accordance with the objectives of the approval committee.
The procedure of figure 1 respects the total quality control
philosophy (Mizuno, 1989). This philosophy states that any
characteristic of the product that does not satisfy the
customer triggers the redesign of the current manufacturing
system. The new. system must incorporate the necessary .changes
to ensure that the product fully satisfies the customer. Table
3 shows the product characteristics currently desired by the
customers. This list must be used only to initially define the
desired characteristics (block. -A in figure 1). These
characteristics will then evolve during the application of the
procedure (as the effects of the control factors on the
current characteristics of the product are learned by the
participating groups) .
The proposed procedure generates the profiles of. the changes
(block J in figure 1) but does not .specify. their nature.
Hence, the obtained profiles must be converted to
implementable terms using table 1 as a reference list. This
table does not include all kinds of changes and is rather a
collection of the japanese JIT components. These components
are currently the most succesful changes. around the world
(Schonberger, 1987; Harmon, 1990; Hayes, 1984).
The procedure of figure 1 is inspired on the reference
approach (Macedo, 1990), a strategy design method based on
system dynamics methodology (Richardson, 1981). However, the
system dynamics principles (table 2) are implemented here
using the continuous-discrete manufacturing simulation
language SIMAN (Pegden, 1990). This is different from the
traditional use of the discrete simulation languages
(Krajewski, 1987).
The use of a continuous-discrete simulation language is
necessary for the following reasons. The redesign of a
problematic system consists of appropiately modifying the
factors that control its current behavior (block G in figure
1). When these control factors can be captured by modelling a
few of the most significant patterns of the problematic
system, the differential equations are useful (these equations
are then solved using any continuous language, for example
DYNAMO or the continuous features of SIMAN). However, when
these control factors belong to the internal organization
(arrangement) of the system components they are not easily
modelled with the differential equations (Simon, 1969). In
this case, any discrete language (for example, the discrete
features of SIMAN) must be used. Notice, however, that most of
Page 322
System Dynamics '91
Figure 1. Procedure to redesign a manufacturing system so that it generates
a product that satisfies some desired characteristics. The decisions of the
Individuals In the participating groups are aggregated using the nominal
group technique.
A
‘Specify the product desired characteristics by assigning values to the appropriate
coefficients of table 3. These values can be found using benchmarking techniques
(Camp, 1968)
Y
‘Observe and quantify the current product characteristics using the most appropriate
perception tools (Ozeki, 1990)
1
Is there a gap (s) between the observed and the desired product characteristics 2) STOP
yes
The observed product characteristics become the problematic characteristics
1
Using system dynamics principles (table 2) and SIMAN language (Pegden, 1990),
construct a base model whose simulation reproduces the problematic characteristics
[ Participating group: suggestion team |
1
By intensive simulation of the model, identify the factors that control the patterns of the
problematic characteristics (Richardson, 1981)
[ Participating group: suggestion team ]
¥
By intensive simulation of the model, redesign the control factors so that the
pattems of the problematic characteristics are focused on the desired patterns
(Richardson, 1981) [ Participating group: suggestion team ]
t Y
Specify the new desired H | Do the simulated patterns satisfy the
patterns desired ones and are these patterns
[ Participating group: robust to the variations in the
approval committee } NO uncontrollable factors (Forrester and
Senge, 1980) ? [ Participating group:
approval committee ]
y Yes
The new control factors are the profiles of the desired structural changes. Translate
these profiles into implementable terms using table 1.
[ Participating group: suggestion team }
System Dynamics '91 Page 323
the control factors related to the design of a JIT prototype
belong to the organization of the manufacturing system
components (table 1).
3. A case study
In this section the proposed procedure is used to design a
just in time protoype ona case study. The relevant results
are the following ones.
When the time between the arrival orders to an assembly line
is 5 days, the product characteristics desired by the customer
are respected: high quality (all the assembled products show
zero defects), relatively low cost (because the stock of parts
waiting to. be assembled oscillates between zero and one
products) and low delivery delay (between 2 and 4 days). As a
consequence, the customer order size is very good, it
oscillates between 2 units/order and 4 units/order.
However, when the time between the arrival orders reduces to
3 days, most of the characteristics desired by the customer
are not met and the sales rate falls. Following the step E of
figure 1, the base model of figure 2 is constructed. Its
simulation for 10 orders with a time between orders of 3 days
generates the behaviors of figures 3 to 7 and table 4. These
dynamics are far from those desired by the customers
generating the decline of the customer order size (figure 6).
Following the steps indicated in blocks F to I of figure 1,
two profile changes that generate the desired characteristics
of the product were obtained (figures 3 to 7 and table 4). The
first profile change (table 3 in figure 8) consists of
reducing the assembly delay to 1.25. days/unit. In addition, ;
this delay becomes indeperident of the number of parts waiting
to be assembled. The second profile change (table -.4 in figure
8) consists of always producing good assembled products.
The obtained profile changes can be implemented in many ways
(table 1) and a detailed study of each possible implementation
must precede the final choice. For example, the first profile
can be implemented uSing assembly line organization with
multiskilled operators and undercapacity scheduling (in order
to reduce the assembly delay) and a total productive
maintenance program (in order to make the assembly delay
independent of the number of parts waiting to be assembled).
The second profile change can be implemented using a zero
defect quality control system for inspecting all the parts to
be assembled. . i
4. Conclusions.
Empirical studies show that American and European firms
implement prototypes of the japanese JIT system. However, the
currently available tools to design these JIT prototypes are
not very powerful. In this paper anew tool that applies
system dynamics principles, using a continuous-discrete
simulation language, was presented. This tool guarantees the
System Dynamics '91
Page 324
WNgss Bebo
HAg)=3 Peozos
I
nooeeeeesansetecetnee i
Table?
ta A)
Figure 2. Cause-effect diagram of the system that ganerates the current characteristics of the product (Base mode).
‘The customer order size depends of two factors, the sales effectiveness and the quality of the product. In addition, these factors
‘depend on the product delivery delay (tables 1 and 2), On the other hand, more long is the, queue of components waiting for
‘assembly, loss frequently are the machines maintained breaking more frequertly and producing a more long assembly delay (table 3).
Finally, the quaity of the components is verified by sampling procedures bofore they entor into the assembly line, More long is.
the quove of waiting components, less officio is this vertication producing more defective assembled products (table 4).
System Dynamics '91 Page 325
100
0 so 100 pays 8 50.
50 100 pars.
Figo. Figwed. Fou:
Individual product delivery delay. Obtained by Number of parts end detactve prosucts waitng for Number of products ordered. Obtained by
2 Thebase moda ¢———-)andb)Thw__aasonty btalned by alnulating 8) The base _smdtg ae Dae eel (=) are) fw ba
Rene meal ih tn echt ngs sded mt”) nd tn toe mc dh e- - nddh sicard cargs eed 8)
toit( — }. Number
sear
Sie Srey andi nC met! nda ote ae (rn
and) te nes model with the stuctural changes added to It
changes added (mm), (-—) Number of ofecive products reassembled
cblaed by simulating fhe base model (—-—).
100 pao 50 100 pas.
Figue7. Pogue 8,
Obtained by simuating Ses afectrrns fc, Cae by sndang Teo) be base mal (—-)1) Med 1
1). be model (——-) web) be bse model 0) Bu bao meal (—-) ab] basa model fu he obi etm gab fw prot
shut changes added tit —). ‘hucutal chonges eed 19 k(——). _ dedited characte
‘ual foc Outuned by ung | Fo bao “Tele 4 inte ase model (me )b) oct
structural changes added to t( —). ote rosa charsowisba (=)
1) The goal- 2) The systemic 3) The 4) The adaptive — 5) The structural
oriented principle: _ principle: Begin endogeneous aggregation principle: principle: Include
Before building any the construction of principle: Include —_Use in the model a in the model the
model, identify a the model at the in the model level of detail enough closed and the
problematic (s) _problematic(s) _—the true sources of to reproduce the ‘open cause-effect
behavior (s). A behavior (s) and _the problematic (s) protlematc 6) relationships that
(5). Itis not
behavior is include in the behavior (s). In link the factors
problematic when model fact, the desired necessary to model’ which cause the ~
it differs ail the factors strategy consists each partof the problematic (s)
from a desired whose interactions of properly problematic system —_behavior (s)..
one, generates the modifying these _in detail, adequate
problematic (s) ‘sources. aggregation must be
behavior (s). used when necessary.
Page 326 System Dynamics '91
Table 3. Current product characteristics desired by the customers
Product desired Coefficients that partially capture the product characteristics
characteristic in the manufacturing system
41) Maximum Defects rate; Reworking rate; Number of post-guarantee interventions;
quality ‘Quantities of corrective work; Rate of customers complaints; Avetage time
to repair retuned products; Intervention timé.
2) Minimum Average stocks of materials, work in process and finished pro
Price/cost Materials, work in process and finished products tumovers; Pcrcaniaiisé of
shortages of scheduled material for production; Machines downtime;
Machines utilization ratio; Space utilization; Employee hours worked
exceeding target levels; Employee absenteis:n rate; Percentage of goods
shipped on time.
3) Minimum Mean delivery time; Mean delivery delay; Percentage of products shipped
delivery delay | ontime.
4) Meximum Number of different finished product codes offered; Number of product
variety codes in a given time period; Time to convert to new production levels; Time
to introduce new products; Number of products realisable in the minimum
planning horizon; Maximum number of changes in the product design.
Table 4. Values of the coefficients that capture the product characteristics In the
base model and when the structural changes are Introduced in thls model
(uzunits ; d=days)
st desbed Coefficient that partially Value of the coefficient in:
Prod capture the product ‘The base model with
caracterictle cheracterictic The base model | she structural changes
Included
4) Maximum ‘Number of products 6u 20u
quality mace good at the first (SIMAN output) (SIMAN output)
wi
Number of defective Ou
__ |. products reassembled (fig 5)
2) Minimum Number of parts and min: Ou
price/cost defective products max: 1u
waiting for assembly average: 0.42 u
(stocks) (fig 4)
3) Minimum Individual product min:1.25d
delivery delay | delivery delay max: 2.50 d
average: 56.11 d average: 1.88 d
(fig 3) (fig 3)
Makespan 124d 295d
(fig 5) (fig 5)
4) Maximum ‘Total number of 14u 20u
market share | _ products ordered (fig 5) (fig 5)
‘Order size average: 1.14 ~ average: 2.5 u
(fig 6) (fig 6)
System Dynamics '91 Page 327
coherence of the JIT prototype components before their
detailed designs. This tool must be used on a regular basis to
keep improving the manufacturing system as suggests Mazaaki
(1986).
References
Bartezzaghi E., F. Turco (1989), The impact of just in time on
production system performance: An analytical framework,
International journal of operations and production
management, 9(8), 40-62.
Burbidge J. L. (1975), The introduction of group technology,
Heineman, London.
Camp R. C. (1989), Benchmarking: The search for industry best
practices that lead to superior performance, Quality Press,
Milwaukee, WI.
Celley A., W. Clegg, A. Smith, M. Vonderembse. (1986),
Implementation of JIT. in the United States, Journal of
purchasing and materials management, Winter, 9-15.
Fogarty D. W., Th. R. Hoffman (1983), Production and inventory
Management, South Western Publishing Co., Cincinnati, OH.
Forrester J., P. Senge (1980), Tests for building confidence
in system dynamics models, TIMS studies in the management
‘sciences, 14, 209-228.
Fukuda R. (1989), CEDAC: A tool for continuous systematic
improvement, Productivity Press, Cambridge, MA.
Gilbert J. P. (1990), The state of JIT implementation and
development in the USA, International journal of production
research, 28(6), 1099-1109.
Groover M., E. Zimmers (1984), CAD/CAM: Computer aided design
and manufacturing, Prentice Hall, Englewood Cliffs, N.J.
Harmon R. L., L. D. Peterson (1990), Reinventing the factory:
Productivity breakthroughs in manufacturing today, The Free
Press, New York. .
Hayes R., S. Wheelwright (1984), Restoring our competitive
edge: Competing through manufacturing, J. Wiley, New York.
Hirano H. (1988), JIT factory revolution: A pictorial guide to
factory design of the future, Productivity Press, Cambridge,
MA.
J.H.R.A.: Japan Human Relations Association (1988), The idea
book: Improvement through total employee involvement,
Productivity Press, Cambridge, MA.
J.M.A.: Japan Management Association (1990), Kanban and just
in time in Toyota: Management begins at the work place,
Productivity Press, Cambridge, MA.
Kobayashi I. (1990), Twenty keys to workplace improvement,
Productivity Press, Cambridge, MA.
Krajewski L. J., B. King, L. Ritzman, D. Wong (1987), Kanban,
MRP and shaping the manufacturing environment, Management
science, 33, 39-57.
Macedo J. (1990), Designing a manufacturing function as a
competitive weapon using the reference approach, in D. F.
Page 328 System Dynamics '91
Andersen, G. Richardson, J. Sterman, System dynamics 90,
686-700, System Dynamics Society, Cambridge, MA.
Mazaaki I. (1986), Kaizen: The key to Japan's competitive
success, Random House Business Division, New York, NY.
Monden Y. (1983), Toyota production system, Industrial
Engineering and Management Press, Atlanta, GA.
Mizuno Sh. (1988), Management for quality improvement: The
seven new QC tools, Productivity Pres, Cambridge, MA.
Mizuno Sh. (1989), Company wide total quality control, Asian
Productivity Organization, Tokyo.
Nakajima S. (1989), TPM development program: Implementing
total productive maintenance, Productivity Press, Cambridge,
MA.
Ozeki K., T. Asaka (1990), Handbook of quality tools: The
japanese approach, Productivity Press, Cambridge, MA.
Pegden C. D., R. Shannon, R. Sadowski (1990), Introduction to
simulation using SIMAN, Mc Graw Hill, New York, NY.
Piper Ch., R. Mc Lachlin (1990), Just in time production:
Eleven achievable dimensions, Operations management review,
TS) 71-82
Richardson G., A. Pugh III (1981), Introduction to system
dynamics modeling with DYNAMO, MIT Press, Cambridge, MA.
Schonberger R. J. (1987), World class manufacturing casebook:
Implementing JIT and TQC, The Free Press, New York, NY.
Shingo S. (1981), Study of the Toyota production system, Japan
Management Association, Tokyo. 7
Shingo S. (1985), A revolution in manufacturing: The SMED
system, Productivity Press, Cambridge, MA.
Shingo Ss. (1986), Zero quality control: Source inspection and
the poka yoke system, Productivity Press, Cambridge, MA.
Simon H. A. (1969), The sources of the artificial, MIT Press,
Cambridge, MA.
Sink D. Ss. (1983), Using the nominal group technique
effectively, National productivity review, Spring, 173-184.
Sterman J. D. (1989), Modelling behavior: missperceptions of
feedback in a dynamic decision making experiment, Management
science, 35, 321-339.
Sugiyama T. (1989), The improvement book: Creating the problem
free-work place, Productivity Press, Cambridge, MA.
Suzaki K. (1987), The new manufacturing challenge: Techniques
for continuous improvement, The Free Press, New York, NY.
Suzue T., A. Kohdate (1990), Variety reduction program: A
production strategy for product diversification,
Productivity Press, Cambridge, MA.
voss Ch. A., S. J. Robinson (1987), Application of just in
time manufacturing techniques in the United Kingdom,
International journal of operations and production
Management, 7(4), 46-52.
Voss Ch. A., L. Okazaki (1990), The transfer and adaptation of
JIT manufacturing practices by japanese companies in the UK,
Operations management review, 7(3), 24-29.
Wantuck K. A. (1989), Just in time for America: A common sense
production strategy, The Forum Ltd., Milwaukee, WI.
CC BY-NC-SA 4.0