Transforming Collaboration: A Vision for Human-Aware Robots to Enhance Worker Safety and Boost Production Efficiency

Human-robot interaction is becoming a necessary part of the industry to achieve higher efficiency, productivity, and comfort. However, to ensure human safety, the awareness of human workers has to be under supervision. In this research, we defined a new function to estimate awareness through eye contact between the worker and the robot. This mechanism exploits a proposed robust and real-time eye contact classifier based on SeNet-50, that outperforms previous proposals based on ResNet-18.