The University of Manchester is leading a £4.6 million robotics research project to develop systems capable of working collaboratively and autonomously on hazardous nuclear sites too dangerous for human workers.
Decommissioning dated nuclear plants, disposing of nuclear waste and remediating the surrounding areas are challenging enough. Just as big a threat is the potential exposure to dangerous levels of radiation, which means that human access is restricted and the majority of that work needs to be completed by robots.
A consortium led by the University of Manchester, which includes the University of Birmingham, the University of the West of England (UWE) and industrial partners Sellafield, EDF Energy, UKAEA and NuGen, has been funded with £4.6 million from The Engineering and Physical Sciences Research Council to develop more sophisticated robotic solutions.
The group is aiming to develop autonomous robots that can handle tasks with dexterity and collaborate with each other. The project will build on the University of Manchester’s pioneering work in robotics to date, which includes the MIRRAX 150 (below), an adaptable robot designed for use on nuclear sites.
Safe decommissioning of nuclear plants requires investment
The cost of cleaning the UK’s historic nuclear sites now stands at £117 billion ($154 billion). The University of Manchester’s Professor Barry Lennox, who is leading the nuclear robotics research project, highlighted the importance of investment and pointed out that his team’s research could prove useful beyond nuclear projects.
“If we are to be realistic about clearing up contaminated sites, then we have to invest in this type of technology,” he said. “These environments are some of the most extreme that exist, so the benefits of developing this technology can also apply to a wide range of other scenarios.”
“This programme of work will enable us to fundamentally improve [robots and autonomous systems] RAS capabilities, allowing technologies to be reliably deployed into harsh environments, keeping humans away from the dangers of radiation.”
The project will aim to develop robots that solve the challenges faced by previous models, which include an inability to grasp and manipulate objects effectively, as well as difficulties with computer vision and perception. Central to the research is the introduction of autonomy; these robots need to be able to operate without direct supervision.
After announcing the funding to the consortium’s nuclear robotics project, Professor Philip Nelson, chief executive of the Engineering and Physical Sciences Research Council (EPSRC), said, “For several decades, EPSRC has been at the forefront of supporting the UK’s research, training and innovation in robotics, automation and artificial intelligence systems.
“Throughout the world, however, from the United States to South Korea, China to Japan, governments are investing billions of dollars into these new technologies. We are faring very well against this global competition, and we should not slow the momentum. These investments are vital for continuing the pipeline that transforms research into products and services.”
The challenges of building a robotics network in a nuclear environment
Speaking exclusively to Internet of Business, Professor Barry Lennox outlined a few of the many challenges the project will face. While nuclear environments can be deadly to humans, working in them isn’t plain sailing for robots either.
“A major problem with using robotic systems in nuclear environments is that the electronics can be damaged by gamma radiation – several robots have failed at Fukushima Daiichi power plant for example,” he explains.
A priority then will be developing techniques to help enable robotic systems to survive for longer periods of time. The MIRRAX 150, in the video above, was designed to fit through a tiny access port before mapping the facility. “We will be investigating techniques that allow it to map its environment whilst avoiding some of the high dose rate areas within the facility and recharge itself by energy harvesting,” said Lennox.
“We will also be developing fault tolerant control systems that will enable the robot to survive in the event that some of its electronics become damaged.”
Beyond simple survival, Lennox and his team, with help from vision system experts at the University of Birmingham, will need to give their robots spatial awareness and the ability to understand their environment, “so that they are able to make decisions autonomously and not have a heavy reliance on human operators.”
And as for the single biggest challenge ahead for the project? It’s impossible to name just one. “This is the hardest question as there are so many engineering challenges,” said Lennox.
“An ideal robot would be a submersible ROV that is untethered and able to transmit HD video underwater – this is extremely challenging and not possible using existing technology. We will be developing manipulators that are able to grasp objects – how will they be able to autonomously determine the most appropriate way to grasp objects of varying dimension? However, possibly the biggest challenge is to ensure that if we develop a robotic tool that is useful, can we actually convince the nuclear industry, which is naturally highly conservative, to adopt it?”