SAFETY ASSESSMENT OF HUMAN-ROBOT SYSTEMS (2ND REPORT, LOGIC MODELS FOR THE ANALYSIS OF THE ACCIDENT-CAUSING MECHANISMS - PART 1).

Abstract

This paper first develops Boolean logic models to represent accident-causing mechanisms where a robot body or arm strikes a human. The models can be used as a checklist to uncover various oversights which could pose hazards in human-robot systems. The necessity of dividing the system into phases to assure statistical independence among the basic events is emphasized, and examples of the system phase division are demonstrated. Minimal cut sets of the logic models in the typical system phases are then extracted to obtain the main system failure modes.