Buffer allocation strategies in shop floor using simulation

Abstract

Buffer allocation is considered to be an important strategy in real life production system because 75% of working capital is tied up in-process inventory in any industry. Therefore an optimum buffer allocation strategy may help in better inventory management of the industry. Consideration of buffer strategy and work distribution play vital role in design of flow lines. However analysis of flow lines using queuing theory become intractable when number of machines increases. Simulation models were made with discrete event simulation being adopted to gain better insight into the problem. To this end, a flow line model has been developed using EXTEND- v.4 to conduct experiments to meet the objective of designing flow lines for buffer allocation and work allocation. The effect of varying number of inter-stage buffers, varying time of each machine, using different process input to study the corresponding outcome is to be experimented ahead. On the basis of analysis of all three models we came to know about some basic patterns like
1) Maximum queue length increase with buffer capacity.
2) Average queue length increases with buffer capacity.
3) Average wait increases with buffer capacity.
4) Throughput increases with buffer capacity.
5) Machine utilization increases with buffer capacity to some extend
6) Maximum wait increases with buffer capacity.
Now coming to the most important part of the project that is analysis of practical model from toy car industry. This industry has a small floor space between the assembling line
which can accommodate a maximum of ten buffers. So they wanted us to analyze their flow process and tell them what would be the best strategy for smooth flow of assembling
process. Then we made the simulation model of there process and analyzed for three most important inter-stage buffers. By increasing the middle buffer than side buffer gave higher output than by increasing the side buffers. This trend was prominent, but only upto the strategy 4-6-4 then the increase in throughput was very marginal and will not be economical considering increase in required floor area, reprocessing investment( in some cases due to long wait). On the basis of all this study we concluded and suggested them that 4 – 6 – 4 buffer allocation strategies will be the best one for their toy car assembling process.