Design of a two-card dynamic kanban system using a simulated annealing algorithm

The design of a kanban system addresses the selection of two important parameters, i.e. the number of kanbans and lot sizes of part types. Kanban-based operational planning and control issues have been tackled in a number of studies by means of analytical or simulation modelling. However, the estimation of these parameters becomes complicated because of issues such as variation in demand, variation in processing times, different types of products, etc. The combinatorial property of such problems warrants the development of efficient methodology or heuristics to obtain a good solution. In this paper, an attempt has been made to select the number of production and withdrawal kanbans at each workstation and the lot size for each part type required to achieve the best performance using a simulated annealing algorithm technique. An object-oriented simulation model of a two-card dynamic kanban system capable of handling different types of part with different demand requirements has been developed and used for the analysis. Each part type has its own number of production ordering kanbans and withdrawal kanbans at each workstation. The lot size can also be different for different part types. A bicriteria objective function comprising mean throughput rate and aggregate average kanban queue has been used for evaluation. Different types of problem have been tried out and the performance of the algorithm is studied.