Flexible hoist scheduling for electronics manufacturing

In the manufacturing of printed circuit boards (PCBs), computer-controlled hoists for material-handling are commonly used in automated PCB production systems. For many such production systems, there are no storage buffers between successive production stages and the processing time for each stage is (upper and lower) bounded. In addition, because of the detrimental effect of oxidization on the quality of the product, the parts must be transported with “no-wait” between successive processing tanks. However, the sequence of moves of the hoist need not be the same as the move sequence of the parts, and by optimizing the schedule of the hoists, production throughput can be maximized.; Scheduling the movement of programmable hoist for production lines is generally known as the Hoist Scheduling Problem (HSP). HSP has been proven to be NP-complete, even for problems with single-hoist and mono-product configuration. There are two main approaches to this problem. The first is a deterministic approach where, given bounds on the processing times for each stage in the production process, a minimum cyclic schedule is sought. The second is a real-time approach where decisions are made “on-line” as to which tasks the hoists should perform next.; After an extensive literature review on HSP, a constraint-satisfaction approach for the cyclic HSP (CHSP) for lines with single-hoist, mono-product configuration is presented. A binary search procedure is proposed and a tighter bound to the cycle length is introduced to reduce the computation effort.; Compared with single-hoist problem, multi-hoist scheduling problem (MHSP) is significantly more difficult due to the additional constraint of hoist collision avoidance. Most multi-hoist schedules in industry are experimented on-line, which are costly and time consuming. The use of simulation will save both cost and time, and improve current manual approaches so that better schedules can be identified. To this end, we first develop simulation models based on data from several actual PCB electroplating lines to assess the performance of various hoist scheduling heuristics. Experimentation was conducted with different combination of hoist speed and inter-hoist distance to examine their effects on productivity. Next, a solution approach is given to solve MHSP in the presence of hoist or tank failures. Finally, we expand the study of MHSP to multi-product case. The objective is to investigate additional productivity gain of processing multi-product on the PCB production line simultaneously and identify good hoist scheduling heuristic that maximizes production throughput for all products.