Research Issues for Automatic Assembly

Abstract The research issues currently being explored world wide will be discussed. The work falls under two main headings namely, a) parts mating phenomena-quantitative descriptions of what happens when parts interact during the assembly process, the kind of information available, its quality etc. b) programmability-the capability of a machine to be taught a new task rather than being built specifically to do the task. Exploration of the parts mating question requires investigation into geometric, force-friction, and logical test characteristics of the mating process, involving the generation of and the carrying out of very precise experiments to examine hypothesis, Programmability issues include the analysis of manufactured products to determine kinds of tasks, statistics of their occurence and geometric requirements on the placing and alignment of parts; and the general economic modeling and comparison of manual assembly, assembly by special machines and assembly by programmable machines.