30220140501001 by iaemedu


									 International Journal of Production Technology and Management (IJPTM), TECHNOLOGY
 ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME
                            AND MANAGEMENT (IJPTM)

ISSN 0976- 6383 (Print)
ISSN 0976 - 6391 (Online)
Volume 5, Issue 1, January - February (2014), pp. 01-09
© IAEME: www.iaeme.com/ijptm.asp                                      ©IAEME
Journal Impact Factor (2014): 1.8513 (Calculated by GISI)

                   HEURISTIC METHOD

                                         Ch. Srinivas
            Professor and Principal, Vaageswari Engineering College, Karimnagar,
                                   Andhra Pradesh, INDIA


         This paper describes a methodology for a pull production inventory control strategy
 which is based on optimization using heuristic algorithm for a mathematical model and then
 it is evaluated using simulation. The approach is described through the examples of
 production lines that process a single part type and is planned according to demand and lead
 time. The pull system has drawn the attention of researchers due to substantial advantages of
 being able to directly control WIP using the CONWIP cards, and can be applied to a wider
 variety of manufacturing environments. The information sharing will help with the
 integrating the echelons but with some complexity. To analyze the CONWIP controlled
 production line mathematical modeling is often applied and also simulation study is another
 performance evaluation tools so that it gives a valuable aid for gaining insights into and
 making decisions about the manufacturing systems. Each node is considered as a machine in
 a CONWIP SC. Our objective in this paper is to extend CONWIP control to a production
 inventory control system setting with an emphasis on customer satisfaction. Given this goal,
 inventory levels must be set for the whole system and the product to satisfy demands fairly.
 We also address a secondary objective of minimizing inventory costs by designing our
 procedure to find the smallest effective inventory level. Formulating a solvable problem
 meant shifting the focus to throughput, but the allocation found by the throughput driven
 heuristic can be utilized to provide good customer service. CONWIP cards can be
 implemented with a simple visual control at almost any level, at a machine, a work center,
 plant, or even an entire supply chain treating each echelon as a work center.

 Key words: Pull and Push Production Control Systems, CONWIP, Supply Chain,
 Genetic Algorithm, Simulation.

International Journal of Production Technology and Management (IJPTM), ISSN 0976 – 6383 (Print),
ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME


         It has been noticed for the last several years that due to change in Globalized business
and communication system a tremendous intensive competition being focussed in business
environment let it be in manufacturing, healthcare, banking etc., in this situation enterprises
must be able to quickly respond to the diverse needs of customers. The Production control
systems can be generally be subdivided into two ways ie., push and pull systems (Spearman
et al. 1990). The principle of pull system is implemented in several production control
strategies such as KANBAN, CONWIP.
         In push system the production is initiated by a central planning system which makes
use of forecasts for future demands. Production is initiated before the occurrence of demand,
otherwise the goods cannot be delivered in time. Therefore the production lead times have to
be known or approximated. Where as in Pull production system, production starts when
demand actually occurs. The production is initiated by a decentralized control system. To
avoid long waiting times for customers, parts and finished products must be stored in buffers.
Therefore, the pull system is called minimum inventory level system and the push system is
zero inventory system.
         Since it is very fact that production without some inventory can only be realized when
the system works without any sort of failures or break downs. But this is very fictitious in
industry scenario, always there will be certain amount of WIP. In pull system a certain
inventory level of parts and finished products is planned to fulfill the customer demand.
         CONstant Work In Process (CONWIP) control system first proposed by Spearman et
al. (1990) uses a single card type to control the total amount of WIP permitted in the entire
line. It is a generalized of the Kanban system and exceptionally it can be viewed as a single
stage Kanban system as a whole. A CONWIP system behaves as follow: when ever a job
order arrives to a CONWIP production line at the beginning of line, a card is attached to the
job, provided cards are available, otherwise, the job must have to wait in queue for allotment
of card meanwhile the job order can be treated as backlog order. When a job is processed at
the final station, the card is removed and sent back to the beginning of the line, where it
might be attached to the next job waiting in the backlog, no work order can enter the line
without its corresponding work permitted card ie., CONWIP card. Along the production line
the total WIP is constant when the system is sufficiently loaded to work non stop (thus the
name CONWIP). The primary difference between CONWIP and Kanban systems is that
CONWIP pulls a job into the beginning of the line and the job goes with a card between the
workstations, while Kanban pulls jobs between all stations (Hopp and Speraman 2001). The
basic rule assigned to each station in a Kanban model applies to the whole line in the
CONWIP model. Under the CONWIP system, the materials are pulled into the production
system by the completion of products in order to restrict the level of inventory, then the
pulled materials are pushed from one station to another through the whole production system.
         CONWIP does not send signal from the bottleneck, but sends only form the final step
in the line, however in CONWIP with fully integrated supply chain it is quite possible in
sharing information between the echelons.
         Inventory is one of the most widely discussed areas for improving supply chain
echelon efficiency. Since the holding of inventories can cost anywhere between 20% to 40%
of product value, hence an effective management of inventory is critical and most essential
(Ballou, 1992). Supply chain integration has become the focus and goal of many firms and it
is used as strategy through which such integration can be achieved. In this environment,

International Journal of Production Technology and Management (IJPTM), ISSN 0976 – 6383 (Print),
ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME

‘supply chain management’ has become effective business tool to reduce supply chain
echelon inventory. Many researchers have approached the management of inventory in
supply chain from operational perspective. The main issues that have been addressed include
deployment of strategies, and control policies. Information and communication technology
(electronic data interchange, internet), globalization, intensifying worldwide production and
competition, shorter product life cycles, higher innovation rates or higher customer
requirements are all the reasons for evolution of SCM.
         There are two types of inventory control in supply chain management. In a stage
based inventory control system, inventory is managed at each stage only and inventory limits
are determined by individual stage. However, in an echelon based inventory control system,
inventory is counted from the current stage to the last stages inventory and the inventory
limits are determined by considering different stages altogether rather than a single stage.
         CONWIP supply chain (CONWIP SC), is an approach by which we attempt to
improve the supply chain (SC) performance, through an extension of the closed production
control system. CONWIP SC is defined in this paper as a production–distribution system, in
which the production line of each firm has a similarity to a ‘‘work center’’ being a part of a
‘‘global line’’ of supply. The set of cards mentioned in the description of CONWIP system,
extends now to a virtual center of control that governs the SC and manages the parts flow and
the inventories along the chain. When orders arrive at the final node, the production orders
and required materials are released to the first node considering its production capacity
constraints. There is a unique and centralized control of the backorders of the SC. Thus, the
centralized information control through Internet type of tools is critical in this context (Ovalle
and Marquez, 2003).
         Distinctive heuristic procedures such as customer dispatching rules, local search and
Meta heuristics procedures such as Tabu Search, Simulated Annealing, and Genetic
Algorithm (GA) have been applied to solve the production control problems and find the
optimal. Genetic Algorithm (Holland, 1975, Goldberg, 1989 and Michalewicz, 1996, Srinivas
and Rao, 2010) belongs to the class of evolutionary computation that was based on Darwin’s
principle of the survival of the fittest. It is a stochastic global search technique which can
effectively search good feasible solutions by mimicking the natural process of evolution and
by using genetic operators. In order to apply the genetic algorithm to solve the inventory
problem, the first step is to encode the candidate solutions with a system of concatenated,
multi parameter, mapped, fixed point coding (Goldberg, 1989). Individuals which represent
the candidate solutions are then grouped into a set called population, and the number of
individuals contained in the population is called the population size. Hence, individuals form
a population and strive for survival in accordance with their fitness function values. The
fittest individuals are selected to undergo a sequence of perturbations (by using crossover and
mutation operations) to breed a new population of individuals for the next generation. During
the search process, the topological information of the solution space is extracted, and the
most promising regions of the solution space are enumerated to locate the optimum. After a
number of generations, the search converges. The overall best individual is then decoded to
identify the final optimal or sub optimal solution.


       There are many studies on control policies for manufacturing systems, Spearman et
al.(1990) proposed that the CONWIP concept could be applied to an assembly system fed by

International Journal of Production Technology and Management (IJPTM), ISSN 0976 – 6383 (Print),
ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME

two fabrication lines. Hopp and Roof (1998) studied such assembly systems using statistical
throughput control method, for setting WIP levels to meet target production rates in the
CONWIP system. Duri et al (2000) developed an approximation method to obtain some
performance measures in three stage production lines under CONWIP control policy with
random processing time and random inspection. Framinan et al. (2001, 2003, 2006) studied
the input control and dispatching rules that might be used in a flow shop controlled by the
CONWIP system within a make-to-stock environment.
        Christelle et al.(2000) analyzed a CONWIP system which consists of three stations in
series. They proposed an analytical method to evaluate performance of CONWIP systems
with inspection for the two following cases: saturated systems and system with external
demands. Yang (2000) investigated the performance of single kanban, Dual kanban, and
CONWIP for the production of different parts on a single flow line. CONWIP is a well
known production control system, and some papers have shown it has better performance
than the KANBAN system (Yaghoub, 2009) but however Kanban is more flexible.
        Houlihan (1985) is credited for coining the term supply chain (SC) with insight
concepts and a strong case for viewing it as a strategy for global business decisions. Many
definitions of SCM have been mentioned in the literature and in practice, although the
underlying philosophy is the same. The lack of a universal definition for SCM is because of
the multidisciplinary origin and evolution of the concept. Simchi-Levi et al.(2000) defined
SCM as a set of approaches utilized to efficiently integrate suppliers, manufacturers,
warehouse and stores, so that merchandise is produced and distributed at the right quantities,
to the right location and at the right time in order to minimize system wise cost, while
satisfying service level requirements. On the other hand, Christopher (2000) defined SCM as
the management of upstream and downstream relationships with suppliers and customers to
deliver superior customer value at minimal cost in the supply chain as a whole. Each echelon
of SC perform independent business with integrated information sharing among all the
echelons and it holds some inventories which may be unavoidable due to existing uncertainty
in the business (Srinivas and Rao, 2004).


        Customers orders have to be analyzed in detailed for the purpose of segregation of
items that allows work orders are to be placed in to list of orders based on dispatching rules
or any customer production priority rule. These can be placed in visual master plan.
        The development of CONWIP control has highlighted the benefits of control policies
that pull work into the facility in response to demand while limiting inventory. The proposed
solution procedure consists of two stages; the first stage is a heuristic for solving the problem,
and the second stage is using simulation to analyze the characteristic behavior of system. We
assumed that, authorized card will never wait for raw material at the input station. Product
mix is allowed, it is possible because a sole card for each lot, ie., each lot will be having a
independent card.
        Starting from an initial allocation of a small number of cards for each product type,
each successive card is given on a trial basis to each of the product types. The type that
makes the best use of the additional card, by moving the total system throughput closest to
the target is allowed to keep the card. The process continues until each type’s throughput
attains its requirement. To overcome bottleneck situation in CONWIP production system a
simple and logical procedure may be applied. Cards must be available for material to be

International Journal of Production Technology and Management (IJPTM), ISSN 0976 – 6383 (Print),
ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME

pulled from upstream to downstream workstations and must have enough cards to pull
processes material from the bottleneck, so that the bottleneck is not blocked.
       The key to the card reduction heuristic is to reduce system WIP by reducing the
number of cards while still meeting or exceeding a desired throughput goal. Estimate the total
number of cards needed for the CONWIP system using the analytical formula

                 wrbTo            wrb
        Th =              Th =            ,
               w + W0 − 1      w + W0 − 1

Where, rb is the rate of bottleneck workstation in job per minute.
W0 is the WIP level attained for a line with maximum throughput operating at the rate of the
bottleneck = rbTo , where To is the sum of the average processing time availability, blocking,
machine breakdowns, supply chain failures.
        The CONWIP system acts inside as the push system, so the throughput rate may be
taken to be equal to the arrival rate of jobs per minute. Using the estimated global card level,
find the current workstation utilization and system output levels. Whenever the number of
cards reduces, the time in queue increase, and the time in the physical system decreases. As
the system WIP is reduced, the order spend more time to get card in the first stage where as
the raw material spends less time in shop floor. Thus card dealing heuristic is designed to
efficiently search for the smallest WIP with an effective allocation. Since, the fitness function
is “max. throughput”, and hence the maximum balanced throughput represents the target for
the card allocation. The proposed CONWIPSC model is shown in Fig. 1. and the heuristic
model is shown in Fig. 2.


                                       Card Flow

                                                    Material Flow
                                                    Cards flow

                    Figure. 1 Cards flow and material flow in CONWIP SC

      In the present work, initial population is generated having a fixed number of
chromosomes and it is called population size (pop_size). Initial population contains suitable
number of solutions for the problem.

The GA Chromosomes structure is: (n,bj,os,k,bn,i)

       n:       orders
       bj :     backlog which can be joined with regular order
       os :     shipped orders
       k :      available production cards
International Journal of Production Technology and Management (IJPTM), ISSN 0976 – 6383 (Print),
ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME

       bn :          production backlog cannot enter in production
       i :           available total FGI in the node

                                                     Start                     Input parameters:
                                                                                  Job type
                                                                                  Order quantity
                                           Production order details               Previous backorder
                                                                                  Due date
                                                                                  Number of cards
                                                  Encoding                        Customer dispatching priority conditions

                                             Generation of Initial

                                           Calculation of objective and
                                               Fitness evaluation

                                                                      no yes
                                                is termination
                                                   criterion                     Stop
                    Population                                                 Results
                after regeneration
                                                is regeneration
                       scheme                      required?

         Pi+I = PI



                               Fig.2 Flow chart showing the Heuristic method

        The present work considers pop_size equal to 150 and it is generated randomly. If the
optimization criterion is not fulfilled, then creation of new generation begins. Parent strings
are selected according to their fitness for the production of offspring and combined to
produce superior offspring chromosomes using crossover and mutation operations with a
certain probability. This process is performed with number of generations as 300 which is the
termination criteria. The probability of cross over is 0.7 and probability of mutation is 0.05.
        Simulation is used before an existing system is altered or a new system is built, to
reduce the chances of failure to meet specification, to eliminate unforeseen bottlenecks, to
prevent under or over utilization of resources, and to optimize system performance (Geoffrey,
1978). Thus, simulation modeling can be used both as an analysis tool for predicting the
effect of changes to existing systems, and as a design tool to predict the performance of new
systems under varying sets of circumstances. A performance evaluation is carried out based
on the throughput rates and inventory levels. The CONWIP SC simulation model developed
using Planimate™ is described in Fig.3.

International Journal of Production Technology and Management (IJPTM), ISSN 0976 – 6383 (Print),
ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME

                              Figure 3. CONWIP SC simulation model


         The evolutionary algorithms proposed are coded using VC++ on Pentium 4 with 1.60
GHZ, 2 GB RAM and SP3 and for simulation technique a Planimate™ simulation software is
used. The simulation model has been simulated for 52 weeks with a warm up period of 5% of
simulation time ie., during the first three weeks (warm up time) it is observed that a steady
state is reached. Starting from the moment in which WIP and finished goods inventory in the
chain reaches the steady state, as new orders arrive, sufficient number of cards were always
available for releasing the necessary orders, in order to produce and meet customer demand.
The average throughput rate is reached a steady state after warm up time of initial simulation
period (Fig.5). and the mean delivery times are shown in Table 1.

                                  Figure 5. Average throughput

International Journal of Production Technology and Management (IJPTM), ISSN 0976 – 6383 (Print),
ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME

                          Table 1. Mean delivery times (minutes)
                           From          To         Time (minutes)
                         Supplier 1 Manufacturer          48
                         Supplier 2 Manufacturer          52
                        Manufacturer Customer 1           48
                        Manufacturer Customer 1           72


        This research paper intends to develop the ways for applying CONWIP production
mechanism to supply chain system and to establish a suitable inventory management scheme.
The proposed CONWIP supply chain is evaluated by a GA and Simulation. The simulation
results show that CONWIP supply chain reduces the fluctuation and system inventory.
        However, we derived a throughput target that balances the production system. The
card dealing is based on the given conditions which derived from the heuristics, the stopping
criterion is based on the number of generations which is fulfilled each product type
throughput. Our computational analysis suggests that equitable customer service can be
provided by finding an allocation that achieves a throughput reasonably close to the target.
The proposed model is validated through analytical and simulation study. The proposed
system can derive the multi echelon WIP inventory limits effectively compared to the
traditional stage based inventory monitoring scheme and can also obtain higher service levels
with lowest inventory.
        In the future, the studies can be analyzed for multi echelons with a product mix
complex CONWIP SC network with more realistic demand data to get realistic simulated
results using Radio-frequency identification (RFID) as the source of information sharing.


 1.   M.L.Spearman, D.L.Woodruff and W.J.Hopp (1990), “CONWIP: a pull alternative to
      kanban”, International Journal of Production Research, vol.28, pp.879-894.
 2.   Hopp, W.J. and M.L. Spearman (2001), “Factory Physics: Foundations of
      Manufacturing Management”, McGraw-Hill, New York.
 3.   Ballou, R.H. (1992), “Business Logistics Management”, Prentice-Hal, Englewood
      Cliffs, New Jersey, 3rd ed.
 4.   O.R.Ovalle and A.C.Marquez, (2003), “Exploring the utilization of a CONWIP system
      for supply chain management.A comparison with fully integrated supply chains”,
      International Journal of Production Economics, vol.83, pp. 195-215.
 5.   Holland, J.H. (1975), “Adaptation in Nature and Artificial Systems”, University of
      Michigan press, Ann Arbor, MI.
 6.   Goldberg, D. E. (1989), “Genetic Algorithms in Search, Optimization, and Machine
      Learning”, Addison Wesley.
 7.   Michalewicz, Z. (1996), “Genetic Algorithms + Data Structures = Evolution
      Programs”, 3ed, New York, Springer Publishers.

International Journal of Production Technology and Management (IJPTM), ISSN 0976 – 6383 (Print),
ISSN 0976 – 6391 (Online), Volume 5, Issue 1, January - February (2014), pp. 01-09 © IAEME

 8.    Ch.Srinivas and Rao, (2010), “Optimization of supply chains for single-vendor-
       multibuyer consignment stock policy with genetic algorithm”, International Journla of
       Advanced Manufacturing Technol, vol.48, pp.407-420.
 9.    Hopp, W.J and Roof, M.L (1998), “ Setting WIP level with stastitical throughput
       control (STC) in CONWIP production lines”, Internatio nal journal of production
       research vol.36, pp.867-882.
 10.   Duri, C., Frein, Y., & Lee, H.-S. (2000), “Performance evaluation and design of a
       CONWIP system with inspections”, International Journal of Production Economics,
       vol.64, pp.219-229.
 11.   Framinan, J. M., Ruiz-Usano, R., & Leisten, R. (2001). “Sequencing CONWIP flow-
       shops: analysis and heuristics”, International Journal of Production Research, vol.39,
 12.   Framinan, J.M., Gonzàlez, P.L., & Ruiz-Usano, R. (2003), “The CONWIP production
       control system: review and research issues”, Production Planning & Control, vol.14,
 13.   Framinan, J.M., Gonzàlez, P.L., & Ruiz-Usano, R. (2006). “Dynamic card controlling
       in a Conwip system”, International Journal of Production Economics, vol.99, pp.102-
 14.   Christelle D, Yannick F, Lee H-S (2000), “Performance evaluation and design of
       CONWIP system with inspection”, International journal of production economics,
       vol.64, pp.219-229.
 15.   Kum Khiong Yang (2000), “Managing A Flow Line With Single-Kanban, Dual-
       Kanban Or Conwip”, Production and Operations Management, vol.9, pp.349-366.
 16.   Yaghoub Khojasteh-Ghamari (2009), “A performance comparison between Kanban
       and CONWIP controlled assembly systems”, Journal of Intelligent Manufacturing, vol.
       20, pp. 751-760.
 17.   J.B.Houlihan,(1985), “International supply chain”, International Journal of Physcial
       distribution and Materials Management, vol.15, pp.22-38.
 18.   Simchi-Levi, D., Kaminsky, P., and Simchi-Levi, E. (2000), “Designing and Managing
       the Supply Chain: Concepts, Strategies, and Case Studies” McGraw-Hill, New York.
 19.   Christopher, M. (2000), “Logistics and Supply Chain Management, Financial Times,
       Pitman publishing”, Pauls Press, New Delhi.
 20.   Ch.Srinivas and Rao,(2004), “Simulation of Supply Chains under uncertainty
       inventory levels”, Proceedings 33rd International Conference on Computers &
       Industrial Engineering, Jeju, South Korea, 25 – 27th March, pp. 1-4.
 21.   Geoffrey, G. (1978), “Systems simulation”, Prentice Hall, Englewood Cliffs, New
 22.   C. P. Aruna Kumari and Dr. Y. Vijaya Kumar, “An Effective Way to Optimize Key
       Performance Factors of Supply Chain Management (SCM)”, International Journal of
       Management (IJM), Volume 4, Issue 3, 2013, pp. 8 - 13, ISSN Print: 0976-6502,
       ISSN Online: 0976-6510.
 23.   Amit Raj Varshney, Sanjay Paliwal and Yogesh Atray, “A Systematic Review of
       Existing Supply Chain Management: Definition, Framework and Key Factor”
       International Journal of Mechanical Engineering & Technology (IJMET), Volume 4,
       Issue 2, 2013, pp. 298 - 309, ISSN Print: 0976 – 6340, ISSN Online: 0976 – 6359.


To top