The truck plant of General Motors of Canada produces 60 trucks/hour, 2 shifts/day, in a highly automated factory employing 3,200 people. Flow in the paint shop is prone to bottlenecks when there is a large ratio of two-to monotone jobs. (Two-tone jobs require a second pass through the paint shop.) The shop has four major control points (two conveyor crossovers and two significant conveyor intersections), where decisions for routing jobs can greatly impact paint-shop performance. A simulation model is developed and refined to determine a set of "best" rules for those control points, dynamic rules reflecting input variables, and current system status. The paint-shop simulation model is employed to study overall throughput and time in the system for various control rules, as functions of product mix and the rates of automotive processing and repair. Conclusions are drawn regarding the influence of those rules on the conveyor crossovers and intersections.
