A Holistic Concept on AI Assistance for Robot-Supported Reconnaissance and Mitigation of Acute Radiation Hazard Situations

Abstract

Hazardous situations involving radioactive, chemical, and other dangerous materials pose significant risks to emergency responders. In such situations, robots can support civil forces in reconnaissance and mitigation. However, this requires the development of robotic systems with novel AI-based assistance functions, operation procedures, and raises new ethical and legal questions. To address these topics, we propose a holistic concept covering the aspects of domain-specific requirements, AI-based assistance functions for mapping, navigation, and manipulation, legal and ethical aspects, end-user training, and evaluation. The proposed concept is implemented in two real-world robotics demonstrators and evaluated in multiple complex simulated and real-world environments.