Project BI-SENSS

Blended Intelligence for Safe and Efficient Nuclear Sort & Segmentation.

Bi-SENSS consortium is led by Veolia Nuclear Solutioni (VNS). Fabulous teams from VNS, University of Lincoln (UoL) (Lincoln Centre for Autonomous System (L-CAS)), UK, and CREATEC build an exciting technology for Autonomous + HRI technologies in response to the need for nuclear waste cleaning in the world. The consortium brings together proven technologies and deep domain expertise from leading nuclear, agritech and artificial intelligence (AI) practitioners to build a solution around automating DEXTERTM - a commercial force guided and radiation-hardened teleoperated robotic platform and integrating it with advanced characterisation and tracking technologies to identify, classify, transfer, trace and package waste. UoL team consisted of Prof. Rai, Hanhedid, Dr Argin and Dr Poozhiyil. UoL team led by Dr Ghalamzan build/developed all the components necessary for autonomous movements of the Dexter arm (read more about Dexter here). This is the first time in the history of Dexter tat it moves autonomously. This project outputs (1) Autonomous capability of Dexter (step-change in the commercial opputunity for VNS), (2) Synthetic and Real dataset for Nuclear decomissioning, (3) AI models for autonomous grasping and packaging optimization of items.

Completion notes

What a great way to end 2022! Thanks to the funding from Innovate UK (SBRI Phase 2 Competition) that brought together a fabulous team from Veolia Nuclear Solutions, University of Lincoln, Faculty and Createc to develop robotic sorting and segregation of nuclear waste. It was impressive to see the Dexter robot operating in semi-autonomous and teleoperated modes. The consortium successfully demonstrated the Dexter robot`s capability to identify a suite of unknown materials found in nuclear waste and segregate them based on their category. This project is another success story of an industry-industry and industry-academia partnership that helped boost the Dexter robot's commercial potential for undertaking challenging operations in extreme environments.

SBRI, Innovate UK, 2020-2022.

Contact PI: Dr Amir Ghalamzan

Contact Co-I: Prof. Mini Rai

Contact Co-I: Prof. Marc Hanheide

RSS 2022, "Action Conditioned Tactile Prediction: case study on slip prediction" By: Willow Mandil, Kiyanoush Nazari and Amir Ghalamzan
External references:
All projects
Oxford University
University of Bristol
University of Birmingham
London Imperial College
Queen Mary University
Polytechnic University of Milan