This paper presents a knowledge-intensive concurrent integration framework for designing, planning, analysis and simulation of assemblies, which is built upon P/T nets and multi-agent systems in AI. In the paper, critical issues are addressed on the implementation of intelligent assembly modeling, analysis and design, planning and simulation using knowledge intensive P/T net models, e.g., knowledge modeling and acquisition, function-behavior-structure modeling, fuzzy relational database, case-based reasoning, and assembly design language, etc. The developed multi-agent intelligent environment, RAPID Assembly system, can process various types of knowledge in assembly design, process and task planning, and assembly system design and simulation so that intelligent design, planning and simulation can be implemented in a virtual environment. It assists designers and manufacturing engineers to design and evaluate the product and its assembly processes and assembly systems (e.g., assembly feasibility, sequences, trajectories, and execution strategies) without building physical prototypes. Design cases are provided to illustrate how the system works.
Challenges in exception handling in multi-agent systems
