This work consists of a model which is based on a computerized algorithm and consists of two main parts namely balancing and simulation routines. This model was used to determine the buffer in-process units required between work stations along manual assembly lines. Since the task durations and deviations vary from assembly of one product to another and frum one task to another for the same product; this model was developed to handle different cases of wide and narrow Work variability according tu the type or assembly work.
The model is capable of banding two different assembly line problem cases. Firstly in case of an assembly line production without balancing, the balancing routine is invited first to produce a balancing solution for the problem and then the simulation routine is Invited to determine the number of in-process units needed for each individual work station. Secondly in case of balanced assembly line production, the simulation routine is invited to simulate the work performance in each work station along the line and to determine the number of buffer in-process units needed for each individual work station.
An assembly line problem consisting of fourteen tasks was used to test the balancing and simulation capability of the model. The problem was prepared to produce a perfect balancing solution of three work stations. The resultant balanced problem data was then used to Investigate the relationship between taşk variance, buffer in-process units between work stations and confidence level.