A Cognitive Human Model for Virtual Commissioning of Dynamic Human-Robot Teams

Abstract

Virtual commissioning can be used for planning human-robot workflows. To this end, digital human models enable consideration of human ergonomics and motion times as optimization criteria. This leads to optimal but fixed human-robot schedules with clearly quantified benefits over the manual process. By contrast, recent research on robot co-workers has been putting emphasis on dynamic task sharing. Respective methods enable robots to make decisions recurrently and adapt to unforeseen situations. Measuring the gain in productivity when deploying such systems to the shop floor while leveraging the advantages of virtual commissioning requires replicating the indeterminism of human behaviour in simulation. We, hence, contribute a cognitive human model that captures aspects of dynamic human decision making relevant to everyday work. Our model covers phases of working, pausing, fatigue, and error using Markov Decision Processes with dynamic transition probabilities. Experiments demonstrate the scalability of generated behaviour with regard to different human characteristics.