Towards Autonomous Robotic Organizations
The field of marine robotics is advancing quickly as ocean industries, science, and defence require better spatial and temporal coverage for mapping and monitoring. Autonomous underwater vehicles are moving beyond traditional exploration and inspection toward more complex missions that involve coordinated operations among diverse robotic platforms—including surface and aerial vehicles and satellites—forming a large cyber-physical system-of-systems.
Key challenges, such as limited energy, computing, communication, and underwater navigation, demand more autonomous and computationally efficient GNC solutions. Embodied AI improves situational awareness, reasoning, and decision-making, enabling real-time analysis of multimodal payload data and supporting adaptive mission planning with online risk control.
As operations scale to multi-agent cyber-physical systems, new organizational frameworks are needed. The concept of Autonomous Robotic Organizations (AROs) describes collaborative robotic teams with dynamic coordination and interoperable networks that draw inspiration from human organizations but exceed human operational capabilities.
Scaling autonomy introduces major concerns, including infrastructure security, reliability, and risks stemming from accessible low-cost robotics. Ensuring system integrity throughout mission planning, execution, and demobilization requires advanced situational awareness and rigorous software verification. AROs’ collaborative behaviours further necessitate formal and simulation-based verification, supported by digital twins for scalable testing, training, and real-time decision support.
Finally, online decision-support tools—integrating simulation, data, weather, and mission history—are essential for continuous planning and scenario ranking in dynamic conditions.
The keynote outlines these technical, organizational, and research challenges and presents a roadmap for achieving AROs.