AI-powered robots to work in environments too dangerous for humans

2 min read

A new generation of AI-powered robots is being developed at Manchester University as part of an R&D programme to help the UK maintain its leading position in automation technologies.

Hot robotics: investigating the flashover effects of a shielded drone in Manchester University's high voltage lab
Hot robotics: investigating the flashover effects of a shielded drone in Manchester University's high voltage lab - Manchester University

These new machines – set to carry out work too dangerous for humans - will be designed to think and act for themselves in hazardous places on Earth and beyond.  

‘Hot robotic’ systems were originally designed to work in radioactive environments in decommissioned nuclear reactors, but future assignments will include deployment in nuclear fusion plants, the offshore energy sector, agriculture and potentially outer space.

As part of an R&D programme to maintain UK leadership in robotic technologies, Manchester experts are applying AI technologies to ‘hot robotics’ as they will increasingly need to act independently of human operators as they enter a range of danger zones to carry out highly complex tasks.    

According to Manchester University, an important challenge in the nuclear industry is to improve robot autonomy so that the technology can be used to deliver safer, faster and cheaper decommissioning of legacy power stations and other radioactive facilities.

To support this task, the Robotics and AI Collaboration (RAICo) has been established in Cumbria as a joint research programme between Manchester University, the UK Atomic Energy Agency (UKAEA), Sellafield Ltd, the Nuclear Decommissioning Authority and the National Nuclear Laboratory. The aim is to develop advanced robotic and AI solutions and transferring these to sites across the Nuclear Decommissioning Authority’s UK estate. RAICo will also provide a pilot for the development and application of robotic systems in other sectors.

Academic engagement for RAICo is being led by Professor Barry Lennox and his team at Manchester University. This group leads the RAIN (Robotics and Artificial Intelligence for Nuclear) hub and are also part of the Manchester Robotics and AI Centre.

“The inclusion of AI is because the goal is to develop automated systems that can operate much more efficiently than if they were operated by people,” Lennox said in a statement. “Within RAICo we are looking at how to improve the operation of remote manipulation and inspection systems. We’re helping Sellafield and other nuclear end-users to develop the next generation of remote surveying and handling equipment so they can improve their operations.”

He continued: “The prefix ‘hot’ was introduced because we were interested in deploying the robots into active environments - but we’re now looking to expand the hot so it can refer to more general applications, including the space, agriculture and offshore sectors. Many of the challenges are similar, although the robots may end up looking a bit different.”

Enhancing the AI capability of these machines is the next big challenge for his team, added Professor Lennox. “AI introduces lots of additional problems related to ensuring that the AI will do what we expect it to do and not cause damage or risk the safety of humans.”

Expanding beyond nuclear decommissioning, the Manchester-led RAIN team are also establishing joint programmes of work with the UK Atomic Energy Authority to support them in the development of robotic systems for nuclear fusion reactors.

Manchester’s expertise in AI and robotic technologies was showcased on June 14 as part of a symposium putting a spotlight on the National Nuclear User Facility Hot Robotics programme.