Aggregate assembly process planning for concurrent engineering

Abstract

In today's consumer and economic climate, manufacturers are finding it increasingly difficult to produce finished products with increased functionality whilst fulfilling the aesthetic requirements of the consumer. To remain competitive, manufacturers must always look for ways to meet the faster, better, and cheaper mantra of today's economy. The ability for any industry to mirror the ideal world, where the design, manufacturing, and assembly process of a product would be perfected before it is put mto production, will undoubtedly save a great deal of time and money. This thesis introduces the concept of aggregate assembly process planning for the conceptual stages of design, with the aim of providing the methodology behind such an environment. The methodology is based on an aggregate product model and a connectivity model. Together, they encompass all the requirements needed to fully describe a product in terms of its assembly processes, providing a suitable means for generating assembly sequences. Two general-purpose heuristics methods namely, simulated annealing and genetic algorithms are used for the optimisation of assembly sequences generated, and the loading of the optimal assembly sequences on to workstations, generating an optimal assembly process plan for any given product. The main novelty of this work is in the mapping of the optimisation methods to the issue of assembly sequence generation and line balancing. This includes the formulation of the objective functions for optimismg assembly sequences and resource loading. Also novel to this work is the derivation of standard part assembly methodologies, used to establish and estimate functional tunes for standard assembly operations. The method is demonstrated using CAPABLEAssembly; a suite of interlinked modules that generates a pool of optimised assembly process plans using the concepts above. A total of nine industrial products have been modelled, four of which are the conceptual product models. The process plans generated to date have been tested on industrial assembly lines and in some cases yield an increase in the production rate.