Research Project beyondSPAI - The Safe and Reliable Monitoring of Adaptive Safety Zones in the Proximity of Collaborating Industrial Robots Using an Intelligent InGaAs Camera System

Abstract

Collaborative industrial robots are becoming increasingly costefficent for manufacturing companies. While these systems can be of great help for the human coworker, they simultaneously pose a serious health hazard, if safety measures are implemented inadequately. Conventional safety equipment like fences or light curtains offer very good protection, but such static safeguards are problematic in these new highly dynamic work scenarios. In this work, we present a prototypical interlocked, multi-sensor system for safeguarding dynamic zones around and close to industrial robots. The core of the system is a robust optical material classification method that can distinguish between skin and other materials (e.g. wood) using an intelligent InGaAs camera system. This feature is used to detect reliably human coworkers, consequently enabling a conventional robot to work as a person-aware cobot. The system is modular and can be easily extended with more sensors and different sensor types. It can be adapted to multiple brands of industrial robots and is quickly integrated into existing setups. The four safety outputs can - depending on the penetrated zone - either issue a warning, slow down the movement of the robot to a safe speed or safely stop the robot. Once all zones are identified as clear, the robot can speed up again to its original speed.