NDA Annual Report and Accounts 2021-2022 | case studies
This work started in 2019 when Dounreay colleagues invited a team of researchers from the RAIN Hub (Robotics and AI in Nuclear) at the University of Manchester to visit the site and see some of the real challenges faced as part of their decommissioning process.
Characterisation of redundant facilities was quickly identified as an area with real potential for innovation and, working together, the researchers and Dounreay team were able to pick a laboratory ventilation duct to use to trial new survey technologies under development at the university. Several phases of site trials were undertaken. The first used an existing system, VEGA, and learning from the initial trials was used to modify and tailor the equipment and ultimately develop a new robotic platform, LYRA.
As a confined space below a floor slab, the duct was exactly the sort of area which is hard to reach. Developing new robotic technologies to do this is not only safer, but also provides higher quality data and more of it.
It’s been a win-win from both sides, as the Dounreay site team were able to really focus on what would be an ideal solution, and the Manchester researchers could better understand the complex constraints on the site to develop technology that could really solve NDA group problems.
The remotely operated robotic system that was developed is a low cost tracked platform, combining a range of sensors and a manipulator arm. It is able to undertake visual and radiological surveys and take physical swab samples for later analysis. Through repeated deployment and learning, a robust robotic platform has now been developed and tested on an active site, which paves the way for future remote robotic deployments.
By using a robotic platform, operators are spared unnecessary radiation exposure, keeping them away from harm. More and better data is also generated, which better informs risk assessments and decision making, leading to greater confidence in developing decommissioning programmes based on real-world data rather than speculation and worst-case scenario estimates.
The project showcases exactly how university research can be transferred to meet NDA group challenges through collaboration. As Jason Simpson, the project lead at Dounreay, says: “We really, should be working with universities more! It’s been amazing working with University of Manchester team, they’ve really brought fresh thinking to our challenges. I know we’ve really helped too, not just defining the technical requirements needed, but also helping them to understanding the operational constraints and complexity of working within integrated project teams.”
An even bigger success for me is the wider story that the lead researcher, Matthew Nancekievell, is a former NDA-funded PhD student, and has now moved on to form and lead a spin out company, ICE9 Robotics, specialising in robotic solutions for infrastructure inspection and maintenance, particularly in the nuclear sector. It’s so great to see our funding really help develop those experts we need and move on to deliver those innovations on the ground.
In terms of learning, the project really shows how collaborating has made sure there is a better result for both sides – better solutions for the NDA group, and better understanding for the researchers based on real deployment opportunities. It also shows how it’s important to have that safe space for new technologies to ‘fail’. Early trials encountered a range of technical difficulties, but it was the learning from those that ensured the second-generation system, LYRA, could meet Dounreay’s needs.
It has also been great to have physical deployment trials used as demonstrators. Both teams have been fantastic in sharing access to the trials and the learning, which helps the wider group, and sector, to move forwards.