The UK legacy nuclear waste clean-up problem is now the largest environmental remediation project in the whole of Europe, with estimates for the clean-up cost ranging as high as £100 billion over the next few decades. For academics, this is a chance to develop truly cutting edge robotics technologies, for an application that has tremendous societal impact and urgency – cleaning up the environment and saving human operators from exposure to hazardous radioactive waste. For the nuclear industry, this is also a significant step forward.
A collaborative EU Horizon 2020 project including the University of Birmingham, National Nuclear Laboratories (NNL), French Alternative Energies and Atomic Energy Commission (CEA), National Centre for Scientific Research (CNRS) and the Autonomous Systems Lab of Technical University of Darmstadt will commence in May 2015 to develop cutting-edge new robotics techniques to assist the future clean-up of legacy nuclear waste.
The project titled ‘Robotic Manipulation for Nuclear Sort and Segregation (RoMaNs) will run for three years and has been awarded 6.4 million euros by the European commission. 1.6million will come directly to the University of Birmingham, who will be leading the project. RoMaNs aims to develop robotic manipulation systems which will be capable of handling the millions of cubic metres of unsorted radioactive waste, which has accumulated in the UK over more than half a century.
The research will focus on robotic arms, hands and sensing systems including: developing new kinds of robotic hardware and mechanisms, new kinds of interfaces for human-controlled (tele-operated) robots, and new AI algorithms for vision, sensing, planning and control to enable the robots to execute grasps and manipulative actions autonomously
Robots previously deployed in the nuclear industry have had each individual joint or motor controlled remotely by a human via a joystick or push buttons. Increasing the amounts of AI and autonomy capabilities in the robots will be needed to enable timely clean-up of the vast quantities and complexities of the UK legacy waste inventory.
Legacy waste containers must be cut open and their contents sorted to extract and segregate the most highly contaminated objects. Such waste contains a huge variety of objects and materials, ranging from rubber gloves and suits to pieces of fuel rod casing to contaminated tools and rubble. This huge variety of sizes, shapes and material types poses an enormous challenge to the vision and perception systems and control algorithms needed to enable robots to autonomously understand the scene and plan, execute grasps and manipulative actions on arbitrarily shaped objects.
This project will develop a major new autonomous robotics test-bed facility at National Nuclear Lab, help fund the creation of jobs for a new generation of robotics pioneers in the industry, and go a long way towards promoting autonomous robotics methods to industry leaders and managers. We are currently recruiting a number of post-doctoral Research Fellows to work on the project, and we are negotiating with the University for space to set up an exciting new robotics lab in which to carry out the research. This is an excellent combination of academic and industry partners, which aims to create very exciting technologies which can be applied world-wide across an enormous sector with enormous societal impact.