Automatic Skill Matching for Production Machines

Abstract

We consider the problem of adding a new machine to an existing production line with minimum engineering effort and downtime (e.g. for plug-and-produce). The supervising system and the new machine have interfaces which are similar but not fully compatible with each other, so the interfaces and interaction logic between these systems must be modified and adapted to ensure functionality and exploit the full potential of the new machine. This adaptation is currently done by hand, where a number of parameters and synchronization steps need to be checked by the engineer. We present an approach for detecting incompatibilities between production machine interfaces automatically. To match the parameters requested by the supervising system with the capabilities provided by the machine, we employ an augmented unification method stemming from logic programming. To analyze the synchronization and termination properties of the interaction, we employ graph algorithms on the synchronous cross product of finite state automata. We developed a software prototype and used it to evaluate the approach with example models. Our approach has the potential to optimize the process of machine addition and replacement and minimize potential failure by applying a zero-engineering approach.