Indoor Positioning, Artificial Intelligence and Digital Twins for Enhanced Robotics Safety

Flexible robotics safety solutions allowing the implementation of fenceless robot cells are becoming a reality nowadays. Safety approved sensors such as light curtains, safety scanners, and safety cameras have been deployed already successfully in various industrial robotic solutions. Still, as these safety systems are installed in fixed locations, monitoring predefined regions, the systems can be rigid and inflexible. This paper introduces a novel hybrid safety solution. The solution comprises safety-approved sensors, additional sensors, and artificial intelligence analysis. The system increases flexibility, especially in cases where collaborating humans and robots need monitoring in larger areas. Typically, in such environments, work objects are large and heavy, introducing additional challenges. In addition, the proposed system includes a digital twin implementation that allows a connection between the real and virtual worlds. Already virtual models and robot simulation have been used for designing safe robot applications. However, the efficient use of digital twins in safety planning and safety monitoring is still uncommon