You are on page 1of 18

Inventory Optimization in a supply chain

INTRODUCTION
Market Practices to sustain:

1. Reducing lead-times and costs


2. Increasing customer service levels
3. Improving product quality is the need of current.

 Conventionally, sourcing (obtaining), manufacture, distribution and marketing have been


working autonomously.

 Supply chain management is the efficient coordination and incorporation of different


organizations with different objectives towards a common goal -Optimal balance between the
inventory levels and stock out levels for a desired level of

Three kinds of costs associated with inventory systems.


1. Reorder cost or set cost
2. Inventory holding cost
3. Shortage cost is a loss of profit, goodwill, etc., when run out of stock.
Introduction Contd..
Genetic Algorithm (GA) :

 Developed by Holland and his colleagues in the 1960s and 1970s.

 Encouraged by the evolutionist theory explaining the derivation of species.

 A solution vector x€X is called an individual or a chromosome.

 Chromosomes are made of discrete units termed genes (Binary Numbers). Each gene
controls one or more features of the chromosome.

Chromosome corresponds to a unique solution x in the solution space. This requires a


mapping mechanism(encoding) between the solution space and the chromosomes.

 GA use two operators to generate new solutions from existing ones: crossover and mutation.
 In crossover (iteratively), generally two chromosomes, called parents, are united together to
form new chromosomes, called offspring. Parent’s preference towards fitness.

 The mutation operator introduces random changes into characteristics of chromosomes.


Generally applied at the gene level. Probability of Mutation is low
RESEARCH GAP

Firms now operate in global markets, deals with suppliers and customers spread worldwide, are
required to manage inventories in new and innovative ways, and are faced with possible
channel reorganization. It would be almost impossible to deal immediately the unknown and
predict the future course research in Supply chain optimization is likely to take. However, there
are a few issues within inventory optimization which are identified as suitable and appropriate
for further research exploration.

Although many non-traditional methods have been applied for inventory optimization,
one unexplored research area is to implement the Genetic Algorithm to specifically determine
the optimal excess stock level and shortage level required for inventory optimization in the
supply chain which leads to minimum total supply costs.
PROBLEM FORMULATION

 The supply chain modelled for the study is a three level one, consisting of 2 factories, 1
Distribution Centre and 3 Agents. Four products (P1, P2, P3, P4) have been considered in the
supply chain

Agent 1

Factory 1
Agent 2
Distribution Center

Factory 2 Agent 3

Objective :

Find the optimium stock levels at various supply chain levels (Factory, Distribution and
Agent) so as to minimize the total (Overhead + Direct ) cost.
METHODOLOGY
Gen=0

Create Initial Random


Population

Evaluate Fitness of Each


Individual in Population

YES
Designate Results Termination Criteria
Satisfied ?
Gen=Gen+1
NO

END Perform Crossover and


Mutation

Perform Reproduction

Genetic Algorithm Structure


Methodology Contd..

A. Generation of Random Individual’s Population

 Each individual (with 13 genes here)generated with random values.


 Number of occurrences of the individual in the past records is found.
 In MATLAB this is done by the function count () and the total number of occurrences of that
individual for the particular product is determined.

F1 F1 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P3 P4

400 -102 91 26 -85 24 -67 -58 -98 68 91 -84 -98

Chromosome representation
B. Evaluate the Fitness Function

Nrec Number of reappearance of records in the record set


Nt Numbers of the records present in records set.
N Total number of chromosomes for which the fitness
function is calculated.

C. Genetic Operations (Selection, Crossover and mutation)

C.1 Selection –

Responsible for the selection of the fittest chromosome for further


genetic operations. Ranking based on Fitness function values.
C.2 Crossover (Two point)

F1 F1 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P3 P4

Before Crossover
400 -102 91 26 -85 24 -67 -58 -98 68 91 -84 -98

F1 F1 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P3 P4
300 45 87 65 -45 42 25 -65 -36 63 99 -14 75

F1 F1 F2 F2 DC DC DC DC A1 A1 A2 A3 A3

After Crossover
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P3 P4
400 -102 87 26 -85 24 -67 -58 -36 68 91 -84 -98

F1 F1 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P3 P4
300 45 91 65 -45 42 25 -65 98 63 99 -14 75
C.3 Mutation

Before Mutation
F1 F1 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P3 P4

400 -102 87 26 -85 24 -67 -58 -36 68 91 -84 -98

After Mutation
F1 F1 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P3 P4

400 -102 87 24 -85 26 -67 -58 -36 -84 91 68 -98

 After obtaining the new chromosome, again a random chromosome will be generated. The
process will be repeated along with the new chromosome obtained from the previous process.
 Finally, an individual which is the optimal one among all the possible individuals is achieved.

 Best chromosome, has the optimal information about stock levels of a particular product at
each participant of the supply chain.

 From the information it can be understood that the particular product and its corresponding
stock levels play a significant role in the increase of supply chain cost. By controlling the stock
level of that particular product at the respective participant of the supply chain in the future
periods, the supply chain cost can be lessened.
SOLUTION

 For superior accuracy, the number of iterations should be sufficiently augmented and run on
the most commonly updated large database of historical records.

 As long as minimization of the fitness function is possible, then the iteration continues till
time that no improvement in the fitness function value is obvious. After a certain number of
iterations, if the fitness function value is not improving from the preceding iterations, then this is
assigned that the fitness function value is stabilizing and the algorithm has congregated towards
optimal solution

F1 F2 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P2 P4 P1 P2 P3 P3 P4

300 -200 10 66 -82 54 -8 -45 30 -72 -99 40 35

THE FINAL BEST CHROMOSOME OBTAINED AFTER THE ‘200’ ITERATIONS


RESULTS AND DISCUSSIONS

F1 F2 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P2 P4 P1 P2 P3 P3 P4

300 -200 10 66 -82 54 -8 -45 30 -72 -99 40 35

The final best chromosome obtained after the ‘200’ iterations

 MATLAB Used for the GA Procedures.

 MATLAB script used to generate stock levels (individuals), used for evaluating the
performance of the GA.
 Assumed records of past periods -

F1 F1 F2 F2 DC DC DC DC A1 A1 A2 A3 A3
P1 P2 P3 P4 P1 P2 P3 P4 P1 P2 P3 P3 P4

-371 -736 -199 532 445 758 952 -214 125 -124 456 254 -754

-502 379 -981 -361 214 -524 635 254 -458 -254 789 -214 524

-158 254 486 354 756 -216 -102 -98 142 521 -635 -946 -657

200 -370 35 324 -754 698 -475 -241 -854 -951 -74 325 -98

150 -350 -108 -250 -154 125 145 98 287 -268 361 -381 169

122 -66 -62 48 251 36 -98 65 -24 -89 715 -87 -78

-500 651 -851 -658 -987 -996 -214 256 367 452 -582 -98 -102

502 -984 -653 975 -69 512 -225 -265 246 -657 164 145 -98

120 -368 -652 234 264 125 215 238 486 999 722 245 -842

1000 254 -356 998 -889 -554 245 125 124 128 -268 -364 -128

138 163 -65 99 -135 -146 -185 -478 256 321 -245 -68 524
 Controlling the best resultant chromosome is adequate to reduce the loss either due to the
compiling inventory or due to the shortage of stocks.

 By focusing on the excess/shortage inventory levels and introducing suitable steps to


eradicate the same at each member of the chain, it is possible to optimize the inventory levels in
the future period and thus minimize the supply chain cost.
CONCLUSION
 Inventory management is a important constituent of supply chain management. Proposed a method
built on GA to optimize inventory in supply chain management we also emphasis on to precisely determine
the most probable excess stock level and shortage level necessary for inventory optimization in the supply
chain such that the total supply chain cost is minimized .

 We apply our methods on three stage supply chain model for optimization and Estimated using MATLAB.

 The analysis extracts an inventory level that made a notable contribution towards the upsurge of supply
chain cost, and in turn enabled to predict the future optimal inventory levels to be upheld in all the supply
chain members with the assistance of these levels.

 Hence it is possible to minimize the supply chain cost by upholding the optimal stock levels that was
projected from the inventory analysis, and thus making the inventory management more effective and
efficient.

You might also like