| Author | I Nyoman Pujawan |
| Call Number | AIT Thesis no. ISE-97-43 |
| Subject(s) | Just-in-time systems
|
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Advanced Technologies |
| Publisher | Asian Institute of Technology |
| Series Statement | Thesis ; no. ISE-97-43 |
| Abstract | There has been a growing body of literature dealing with kanban system. Various models have
been proposed with a wide range of solution techniques. However, there is no publication of
work in the area of kanban system which considers the withdrawal lot size as one of the
decision variables. If transportation cost of materials between workstations is considered to be
important, determination of withdrawal lot sizes is essential. In this situation, a trade off
between inventory holding cost and transportation cost will yield a more economical inventory
system.
In this study we propose a mathematical model to determine simultaneously the number of
production kanbans, the number of withdrawal kanbans, and withdrawal lot size for each item.
The initial model is a non linear integer program and becomes a linear integer after
transforming some non linear equations and fixing one variable. The model can be applied in
a multistage and multiproduct assembly system. Four hypothetical cases are presented to apply
the model. Each case represents different complexities of the simplified multistage and
multiproduct assembly system.
A number of different experiments are conducted to learn the relationships among variables
and parameters. Different scenarios of demand for each hypothetical case are investigated to
learn the effect of different level and different variance of the demand to the number of
production and withdrawal kanbans as well as the cost of the system. We also study the effect
of cost parameters to the optimal size of withdrawal lot. The robustness of the decision
variable values with respect to the change of cost parameters are also studied. A number of
consequences are derived from the findings of the experiments. These consequences provide a
good insight into the development of heuristic algorithm.
Since the model is hard to solve, the computation time for complex problem is unreasonably
long. To alter the computation time we propose a heuristic algorithm for the uncapacitated
version of the model. The algorithm has been tested with many different cases and scenarios.
The tests show that the algorithms are able to produce solution with discrepancy of the
objective function less than 3.5% from the optimal solution with far less amount of
computation. At the end of the study we also suggest the future research related to this topic. |
| Year | 1997 |
| Corresponding Series Added Entry | Asian Institute of Technology. Thesis ; no. ISE-97-43 |
| Type | Thesis |
| School | School of Advanced Technologies (SAT) |
| Department | Department of Industrial Systems Engineering (DISE) |
| Academic Program/FoS | Industrial Systems Engineering (ISE) |
| Chairperson(s) | Anulark Pinnoi; |
| Examination Committee(s) | Pandey, P.C.;Nagarur, Nagendra N.; |
| Scholarship Donor(s) | Higher Education Project ADB Loan 1253 INO Government of Indonesia; |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1997 |