National Nuclear Laboratory

Robotic Manipulation for Nuclear Sort and Segregation (RoMaNS)

Project Description

The EU Horizon 2020 RoMaNS project aimed make significant steps in autonomous, tele-operative and shared control for remote manipulation. Inspiration was drawn from the Box Encapsulation Plant (BEP) project at Sellafield.

RoMaNS specifically focused on developing autonomous technology.

Technical Solution

The RoMaNS consortium comprised the University of Birmingham, NNL, the University of Darmstadt, the Commissariat à l’Energie Atomique (CEA) and French National Centre for Scientific Research (CNRS).

Each partner had a specific focus area in developing a semi-autonomous sort and segregation capability. The University of Birmingham focused on vision-based manipulation algorithms and collision-free trajectory planning. The University of Darmstadt focused on learning algorithms. CEA and CNRS focused on haptic sensors, assisted telepresence system and shared control algorithms.

Finally, NNL provided the industrial and nuclear background and the industrial testing facility.

Benefits – What difference it made

The planning of robotic trajectory is resolved in path planning software, where robotic kinematics are calculated against resolved object geometry (generated from sensor data) and a target end-point. These calculations allow the system to generate the fastest, safest path in order to avoid collisions, from the robot’s current location and orientation to the target point on the sort table.

The RoMaNS project has far reaching cross-sector applications in nuclear, aerospace, oil and gas, space, food and agriculture – and within the nuclear industry itself the research aims to advance waste processing, decommissioning, asset care, maintenance, repair, characterisation and sampling techniques.