SafeMove: multi-model navigation for fail-resistant autonomous nuclear material transport

Conference item

The nuclear industry requires safe and efficient transportation of radioactive materials, especially in decommissioning and waste management. As waste package movements increase from monthly to daily frequencies, traditional manual methods have become unsustainable, exposing operators to hazardous environments and creating potential operational bottlenecks. This paper introduces SafeMove, an advanced autonomous transportation system developed by AtkinsRéalis and the Oxford Robotics Institute, designed to address these critical issues.
SafeMove integrates a vehicle-agnostic hardware and software autonomy stack with a sophisticated multisensor fusion system that combines LiDAR, cameras, and inertial sensors. This fusion enables robust localization, detailed 3D mapping, and real-time mission planning, allowing the system to navigate without reliance on GPS or extensive infrastructure modifications. Key functionalities include dynamic obstacle detection, point-to-point mission execution, and risk-aware replanning, ensuring adaptability to complex and evolving conditions within nuclear facilities. Tested in environments analogous to nuclear sites, the system demonstrated exceptional reliability, safety, and operational efficiency. To ensure readiness for deployment in active nuclear facilities, a comprehensive risk assessment based on ISO 26262 standards was conducted.
This paper details the development process, testing outcomes, and deployment considerations of SafeMove, highlighting its potential to revolutionize nuclear waste management. Its implementation offers transformative benefits for the nuclear sector, including significant reductions in human exposure to hazardous conditions, improved operational throughput, and enhanced cost efficiency. Additionally, the system’s adaptability makes it suitable for a wide range of applications, from waste transportation to autonomous inspection and monitoring tasks. By addressing the industry's need for safe, sustainable, and technologically advanced solutions, SafeMove exemplifies the role of robotics and AI in shaping the future of nuclear waste management and decommissioning activities worldwide.

Authors: Marquez-Gamez, G; Harrison, M; Pilon, C; Marwood, R; Stephens, S

• Publication status: Published
• Peer review status: Reviewed (other)
• Publisher: WM Symposia
• Host title: WM2025 Conference Proceedings
• Pages: 4540-4554
• Article number: 25556
• Publication date: 2025-06-01
• Acceptance date: 2025-01-13
• Event title: 51st Annual Waste Management Symposia (WM2025)
• Event series: Waste Management Symposia
• Event location: Phoenix, Arizona, USA
• Event website: https://www.wmsym.org/
• Event start date: 2025-03-09
• Event end date: 2025-03-13
• ISBN: 9798331322441
• Language: English