Virtual assembly with biologically inspired intelligence

Abstract

This paper investigates the introduction of biologically inspired intelligence into virtual assembly. It develops an approach to assist product engineers making assembly-related manufacturing decisions without actually realizing the physical products. This approach extracts the knowledge of mechanical assembly by allowing human operators to perform assembly operations directly in the virtual environment. The incorporation of a biologically inspired neural network into an interactive assembly planner further leads to the improvement of flexible product manufacturing, i.e., automatically producing alternative assembly sequences with robot-level instructions for evaluation and optimization. Complexity analysis and simulation study demonstrate the effectiveness and efficiency of this approach.