Intelligent planning and control of multi-assembly systems.

The global trend towards cost minimisation in manufacturing has intensified during the last two decades. Cost reduction can be achieved either, directly, through elimination of waste, or indirectly, through optimisation of production processes and generating more reliable information regarding the costs incurred. The research presented in this thesis considers cost reduction in three aspects: optimisation of production processes, accurate cost estimation and accounting. Due to the increasing number of combinatorial optimisation problems associated with the production of Printed Circuit Boards (PCB), it has attracted the attention of many researchers who tried to solve these problems with the aim of minimising the production cost. Therefore, PCB production is used in this research as a test-bed for the three aspects mentioned above. Regarding cost reduction in PCB manufacturing, three interrelated combinatorial optimisation problems are considered: the component placement sequencing problem, the feeder assignment problem and the board type sequencing problem. Solving these problems ensures cost reduction by reducing the time required for manufacturing PCBs. As for cost reduction in the costing and accounting aspects, the traditional standard costing and standard accounting have some problems that make them unsuitable for today's manufacturing. Standard costing allocates overhead to labour or machine hours, which leads to a distortion of product costs due to the fact that today's manufacturing relies more on technology and less on human power. As for standard accounting, it has some features and characteristics that contradict with the widely spread lean manufacturing. The deficiencies in standard costing and standard accounting may create more waste and lead to the wrong decisions being taken.A framework is developed to provide solution to the above-mentioned problems in an integrated environment. A mathematical formulation for the three PCB manufacturing-related problems is developed and solved using a metaheuristic-based algorithm. In order to deal with the costing and accounting part of the framework developed, Activity Based-Costing (ABC) and Lean Accounting (LA) are implemented on a PCB manufacturing facility using a case study. ABC is used to estimate the costs of manufacturing PCBs and provide detailed information on how the costs are incurred. As for LA, it is used to reduce the costs associated with the accounting system, which is achieved by eliminating and/or replacing accounting transactions and promoting lean measures.Simulation results obtained show an average reduction in total assembly time of 5.96% and 5.43% when Taboo Search (TS) and Genetic Algorithms (GA) metaheuristics are used respectively. The results also show how ABC can be used to identify the activities used in PCB manufacturing and calculate their costs. By targeting the most costly activities identified by ABC, the production costs can be reduced. Regarding LA, the results indicate how the accounting system costs can be reduced by eliminating some accounting transactions and processes or replacing them with less costly alternatives.