IFAC MCPL 2007

The 4th International Federation of Automatic Control Conference on Management and Control of Production and Logistics
September 27-30, Sibiu - Romania

OPERATIONS PLANNING BASED ON THEORY OF CONSTRAINTS DECISION SYSTEM

Gabriela Proştean

University Politehnica of Timisoara,

Faculty of Management in Production and Transportation
Str. Remus nr. 14, 300191, Timisoara, Romania,
Tel ++40256404039, Fax ++40256403214, e-mail gprostean@eng.upt.ro

Abstract: The raw material, the machines, and the execution operations required by
production are not the only necessary resources. Information is already more important
than the physical resource. The designing and the maintaining of an updated information
system represent a more difficult and expensive approach. The Theory of Constraints
Decision System (TOCDS) is an experimental system and deals with constraint handling
decision in manufacturing operations. It is purposed to define the optimal sequence of the
various operations in a flow shop, considering machines and tools available. The
decisions are based upon comparative reports between the marginal analyses and a report
called the “product contribution”. Copyright © 2007 IFAC
Keywords: optimisation, decision, methods, control, operations, scheduling

1. INTRODUCTION feasible a process of ongoing improvement: system’s

A flow shop type process includes machines and
people that are sequenced along the factory in a
succession required by the established target. The
work may also be unique, or it may be a repetition of
the standard design (e.g. furniture factories,
electronical products factories, automotive factories).
Flow shops are not actually continuous (for instance,
as chemical processes are), because the items within
the flow are discreet, obtained from independent
parts. The manufacturing operations scheduling are
generally based on a Manufacturing Resources
Planning II (MRP-II) system that uses a push system
to load the resources based on anticipated customer
demand, resulting in frequent situations in
rescheduling and expediting of critical items.
In the 80s, Goldratt and Cox developed the Theory of
Constraints for managing production operations
planning and scheduling. The Theory of Constraints
is focused on the principle that the goal of any
industrial organization is to make money, now and in
the future, and that a system’s constraints determine
its capacity and rhythm to make money. Goldratt
specified a five step oriented process to make
constraint identification, decision for constraint
exploitation, subordination of all the other capacities
at the level of the exploitation decision from previous
step, constraints elevation, and new constraint
identification. (Goldratt and Cox, 1984), (Goldratt,
1997).
The efficient and effective operations planning
require the assistance of methodologies supported by
computer to become more practicable.
2. THE MODULES OF THE TOCDS
TOCDS includes a few modules. The first module is
called Operations planning and examines the
operations requirements in concordances with the
market demands. This module examines the planning
feasibility for the operations system. The user has to
communicate with this module interactively for the
number of the products, the weekly demands for each
product, the selling price for each product, the row
material allocated for each product, the material
costs, the operations flow. (Hull and Wu, 1994)

591
The Constraints Identification is also very important There are taken into account 3 raw materials and the
in the system structure. The user has to communicate final package, out of which, through the
the time availability for each machine and tools, combinations required by the technological processes
respectively the technological flow for each product, there result products P1 & P2. (Fig.2)
including the processing time for each operation. The Product P1 is obtained from the raw material 1-1 unit
Constraints Identification module specifies which and the raw material 2-1 unit, combined with the
machine encounters constraint capacity in terms of additional purchased part. Product P2 is obtained
overallocated time. from raw material 2 and raw material 3 (with 1 unit
Constraint Exploitation has the goal to identify the each).
best exploitation procedure for the constraint
machine. It develops the decision process by
establishing a comparative examination between the
marginal analyses and a report called “the product
contribution/constraints”.
Constraint (s) Elevation module suggests changes,
firstly by redistributing the processing time for each
operation and secondly, within the firm available
possibilities by making investment for purchasing
more equipments or by redistributing the available
personnel or equipment.
Reports and External Constraint Identification Fig.1. The operations flow - logical diagram
module provide useful reports for entire decision
process, respectively restart the whole decision
process in case of identification of a new constraint,
as far as the present situation is concerned, will be
followed up (Prostean, 2002), (Prostean, 2005).

3. TOCDS – PROGRAM SPECIFICATIONS AND

IMPLEMENTATION

3.1 Program specifications

Taking into account that the system integrates many Fig.2 Process initialization
algorithms, it was implemented in C++ using O.O.P
Raw material 1 is processed by means of
(Object Oriented Programming)
machines 1, 3 and 4, but not by means of machine 2.
- Each product is an object, storing the information
Raw material 2 is processed by means of machines 2,
that is specifically (weekly demand, selling price,
3 and 4, but not by means of machine 1. Raw
row material price, units supplied, weekly
materials 1 and 2 are taken in quantities of one unit
throughput, throughput, time required for each
each, being processed up to the final stage by means
operation), and can be used later in constraint
of machine 4, and then packed within the same
handling decision in manufacturing operations,
compartment, so that at the output of the process
respectively in operations planning.
there results product P1.
- Each machine is an object, storing the information
that is specifically (machine number, time available)
- There are defined 2 arrays that store the objects, the
first array is used for storing the object having the
product class type and the second array is used for
storing the object having the machine class type.
These classes provide functions for data initialization
and processing during the progress of the program
(time, reports, verifying, a.s.o.).
- The algorithms follow the Goldratt’s Theory of
Constraints five step focusing process (constraint
identification, constraint exploitation, reports for
subordination, constraints elevation, new constraint
identification); (Prostean, 2004), (Prostean, 2006)

3.2 Program implementation

P1 & P2 represent a simple process which produces

only two products, “P1” and “P2”, (Youngman,
2007),(Fig. 1). The objective of the application is
that of establishing a decision through which the
process profit can be maximized.
Fig. 3 Material specifications

592

Raw material 2 is processed by means of machines 2,
3 and 4, but not by means of machine 1. Raw
material 3 is processed by means of machines 1, 2,
but not by means of machines 3 and 4. One unit from
raw material 2 and one unit from raw material 3 are
combined on machine 4 in order to obtain product
P2. (Fig.3, Fig.4)
Fig.4. The technological flow initialization
There are known the following data: -the selling
price for product P1 is 105 MU/piece (Monetary
Units/piece), - the market demand is of 70 pcs./week,
- the selling price for product P2 is 85 MU /piece, -
the market demand is of 110 pcs./week. The
operation costs for the whole process are of 5800
MU /week. The availability of the resources is of: 60
minutes/hour, 8 hours/day, 5 days/week respective in
2400 minutes/week. The user has to provide the
primary data to the system to can identify the
production bottleneck, respectively the assignment of
materials and operations for each machine and the
operations time.
There is no waiting time. At the end of one step, the
following step begins at once.
Fig. 5.a Report – units supplied in conformity with
the weekly demand for Product 1
The processing machines are not multifunctional.
Equipments (machines) 1, 2, 3, 4 can work according
593
to formulas for which they are dedicated. (There are
situations when the machines or the personnel are
“multitasking”).
There are defined the following relations:
- The profit = “throughput” – operation costs
- The throughput= the selling price – the cost of the
materials.
Fig. 5.b Report – units supplied in conformity with
the weekly demand for Product 2
Fig. 5.c Report – weekly net profit for units supplied
in conformity with the weekly demand
- There can be asked the following question:
What is the profit that can be generated weekly by
this process?
Many managers decide to do their best to satisfy
market demand. In such a case, the results will be as
presented in the following Fig. 5 a, b, c.
In the first step of the analysis carried out according
to this approach, the marginal cost of the throughputs
obtained through the difference between the selling
price and the cost of the materials:
- for product P1, the marginal cost is:
Cmp= 105-35=70 MU
- for product P2, the marginal cost is:
Cmq= 85-40=45 MU
In the second step, there is calculated the value of the
throughput, from which there are subtracted the
weekly operation costs, and thus there is obtained the
net profit which is equal to 4050 MU. (Fig. 5.c)
In the third step, there is checked whether the
capacity of the processing machines is enough to
cover the demand.
Thus, there is checked whether the time required at
each processing machine is not longer than the
available time. In this stage the system identifies the
constraint machine, resulting in overallocated time
capacity. (Fig. 6.)
Fig.6. Constraint(s) identification
- Machine 1 – in order to obtain 70 units of P1,
machine 1 requires 70 times an available time
interval of 10 minutes (for raw material 1) and to
obtain 110 units of product B, machine 1 requires
110 times an available time interval of 10 minutes
(for raw material 3). There results a total of 1800
minutes for machine (resource) 1, with an available
time interval of 2400 minutes.
- Machine 2 – for 70 units of product P1, machine 2
requires 70x5 minutes (raw material 2), and to obtain
110 units of product P2, machine 2 requires 110x5
minutes (raw material 2), to which there are added
110x10 minutes (raw material 3). There results that Fig. 8.a Report – units supplied in conformity with
2000 minutes are needed under the conditions of an marginal analysis for Product 1
available time interval of 2400 minutes.
- Machine 3 requires 3400 minutes (70x15 + 70x10 +
110x15).
Machine 3 has a shortage as far as its capacity is
concerned.
- Machine 4 also requires 1530 minutes (70x14 +
110x5).
Following the calculation of the capacity required for
each machine, there has been identified a constraint
concerning machine 3. (Fig. 6.)
By applying the steps of the constraint theory, there
follows the establishing of the way of this constraint
exploiting, and planning, respectively. Fig. 8.b. Report – units supplied in conformity with
The target of the exploiting is to maximize the marginal analysis for Product 2
capacity of machine 3 and to go on having a good
profit by the end of the week. The demand for product P1 is of 70 pieces. For
If this problem is solved as far as the costs analysis is machine 3 there are required 70x10 minutes (raw
concerned, the maximization of the process is material 2) and 70x15 minutes (raw material 3).
obtained by identifying the product with the lowest There results a total of 1750 minutes, required for
cost. product P1. If from the available time interval of
If the identification is made as function of the raw 2400 minutes there is subtracted the requisite for
material cost, then the answer is that product P1 at product P1, i.e. 1750 minutes (2400-1750=650),
35MU is to be preferred to product P2 at 40 MU. there will remain 650 minutes for product P2.
If the identification is made as function of the Product P2 is processed for 15 minutes on machine 3,
execution operations cost, product P2 which is and there result 650/15 = 43 units of product P2.
executed in 45 minutes (10+15+15+5) is to be The result of exploiting machine 3 decided through
preferred to product P1, executed in 54 minutes costs analysis is presented in (Fig. 8. a, b), and there
(10+5+25+14). can be observed a weekly net profit of 1035 UM.
As far as the maximization of the selling price is (Fig. 8.c)
concerned, namely of the final sold product, product
P1, which is sold with 105 MU, is to be preferred to
product P2, which is sold with 85 MU. As concerns
the marginal price, the price of product P1 is 70 MU,
and it is to be preferred to product P2 whose price is
45 MU.
By analyzing the costs involved in the execution of
the two products, the marginal incomes, respectively,
it is obvious that product P1 has prior claim to
consideration, as far as its execution is concerned, in
comparison with product P2. (Fig. 7)
Fig.8.c Report – weekly net profit for units supplied
in conformity with marginal analysis

Conclusion: The weekly net profit in conformity with

Fig. 7. Report – Marginal Cost Analysis
marginal analysis is less then the weekly net profit
for units supplied in conformity with the weekly
demand.
In the first part – the marginal analysis there has been
approached the exploitation of the constraint by
maximizing the outputs of the constraint, where the
following have been taken into account:

594
 - the cost of the execution operations
- the cost of the raw material
- the highest selling price
- the highest marginal price.
However, the constraint exploitation means to
wholly exploit the constraint potential.
- The processing time for product P1 by means of
machine 3 is 25 minutes;
- The processing time for product P2 by means of
machine 3 is 15 minutes;
How much throughput is obtained during each
processing time?
- Product P1 gains 70 MU
- Product P2 gains 45 MU
There follows the calculation of the ratio throughput Fig. 10 c Report – weekly net profit for units
/number of minutes within the constraint (machine 3) supplied according with the “Product Contribution”
- For P1 – 70MU/25 minutes = 2.8MU/minute
- For P2 – 45MU/15 minutes = 3MU/minute In Fig. 10.c, there can be observed a weekly net
profit higher than that of Fig 8.c. Consequently, by
Conclusion: going on in applying the steps of the theory of
- Product P2 generates money for the system with constraints, all the other equipments will be
7.14% quicklier than product P1. subordinated to this exploiting decision (according to
- Consequently, product P2 is to be preferred for Fig. 10.a.,b), although their capacity has a greater
execution as compared to product P1. (Fig. 9) volume.
Out of the weekly demand, there have been left 40
units of product P1 not integrated in the process.
TOCDS generates the constraint (s) elevation
suggesting changes, firstly by redistributing the
processing time for each operation and secondly,
within the firm available possibilities by making
investment for purchasing more equipments or by
Fig. 9. Report – Product Contribution redistributing the available personnel or equipment
(Fig. 11)
The report called “Product contribution” is the key
for reaching the goal to make money, respectively to
make feasible a process of ongoing improvement

Fig.10 a. Report – units supplied in conformity with

the “Product Contribution” for Product 1
Fig. 11 Constraint elevation procedure

In the procedure of constraint elevation the user

Fig.10 b. Report – units supplied in conformity with
the “Product Contribution” for Product 2
initiated an investment of 2950 MU, through which
equipment 3 has been released with a time margin of
one minute, while equipment 4 has been loaded up
with 2 minute.
The final result indicates an increase from 15 minutes
to 16 minutes for the processing of raw material 2.
There will be saved 25 minutes for equipment 3 (for
the processing of raw material 2). The saved minutes
will be used to produce additional units of product
P1, with a rate of 2.8 MU/minute. (Fig. 12). The
weekly net profit has been increased. (Fig.13)

595

Fig. 12 Report – units supplied for Product 1 after
constraint elevation
Fig.12 Report – weekly net profit after constraint
elevation
TOCDS provide useful reports for entire decision
process, respectively restart the whole decision
process in case of identification of a new constraint.
(Fig. 13.)
Fig. 13 Report for entire decision process
4. CONCLUSIONS
TOCDS integrates the Goldratt’s prescribed
methodology in the proposed framework of a flow
shop scheduling, being an experimental system.
The module Constraint Exploitation establishes a
complex framework for deciding the best way of
constraint exploitation, meaning to wholly exploit
the constraint potential. This module integrates the
Goldratt’s theory that the rhythm of efficiency is
generated by the constraint (s) and to go on having
the maximized possible profit by the end of the week
means to identify the product that generates money
on the constraint (s) quicklier than other products.
596
TOCDS is an object oriented scheduling system,
having the following advantages:
• Information is distributed – each product and
machine stores the information that is specifically
(weekly demand, price, time required, time available,
a.s.o), related to it and can be used later in constraint
handling decision in manufacturing operations,
respectively in operations planning.
• Computation is distributed in time – the system
continuously adjust the planner as the environment
changes
• Flexibility – once the flow shop behaviour is
defined at the operation level, the decisions of
adjusting and rescheduling are more flexible.
REFERENCES
Goldratt, E. M., Cox, J. (1984), The Goal, a process
of ongoing improvement, The North River Press.
Goldratt, E. M. (1997), The critical chain, The North
River Press
Hull, R. and Wu, B. (1994), Design process planning
within a computer aided manufacturing systems
design framework, International Journal of
Manufacturing System Design, Vol.1, No. 4,
367-387
Prostean, G. Prostean, O. (2002), Inteligent Project
Manager using a Plannifying, Supervising and
Diagnosing Integrated Process, International
Conference on Artificial Intelligence, IC-AI, Las
Vegas, USA,Vol 3, pp 1404 - 1409, CSREA
Press
Prostean, G. and Other (2004), Startegy of control
developed for project management intelligent
system, “Manufacturing and Management in 21st
Century”, Ohrid, Republic of Macedonia, CD-
ROM
Prostean, G., Izvercianu, M. Giuca, M. (2005),
Knowledge based expert system for production
and operation management (KBES - POM), The
4th International Conference on Management of
Technological Change, Chania, Creta, pag.159-
164, Published by Technical University of Crete
Prostean, G. and Other (2006), Scheduling intelligent
system for time shortening,Advancing Business
and Management in Knowledge-based Society
MIC'06 – Management International
Conference, pp 1873 – 1882, Published Faculty
of Management Koper,University of Primorska,
Portorož, Slovenia.
Youngman K. J., A Guide to Implementing the
Theory of Constraints (TOC), 2007,
http://www.dbrmfg.co.nz.