Environment perception for mobile robots – using artificial intelligence for irregularity detection

Abstract:
Automated guided vehicles (AGVs) in industrial environments usually are safeguarded by sensitive mechanical bumpers or laser scanners to prevent collisions with obstacles. But they can only detect obstacles in one particular plane. Within the scope of the research project S³ (safety sensor technology for service robots in production logistics and other applications), sensor technology was developed which, in addition to safe 3D environmental monitoring, enables further functions such as the differentiation between objects and persons and recognition of irregularities. The sensor technology was evaluated within two use cases: AGVs in the area of production logistics and on a service robot in stationary care. This paper presents the results of the project with a focus on the implementation of irregularity detection with the use of artificial intelligence.