Nuclear Decommissioning

The UK and the world has an extensive ageing nuclear power fleet.

“Nuclear decommissioning is the most significant environmental challenge facing the world today, second only to climate change.” ~ Dr Michael Foster

Decommissioning nuclear sites is a complex, hazardous and expensive process.

Decommissioning is the process by which old nuclear sites are made safe. This includes the clean-up of all residual radioactivity, dismantling the structures, removing any hazardous materials or containing them for storage, and making the site safe for alternative use. It is a complex and lengthy process. The legacy nuclear fleet needs innovative, cost-effective solutions to make the decommissioning process cheaper, faster, and safer for operators.

In the UK alone, the Nuclear Decommissioning Authority’s estimated cost of decommissioning and clean-up is around £131 billion and they predict the process could take up to 120 years. Of the UK’s 19 sites, Sellafield is the largest and has the most diverse range of nuclear-based activities in Europe. It, therefore, presents the most significant and complex challenge of all the nuclear sites.

0 billion
Expected clean-up cost
0 years
To complete process
0 sites
Potential UK sites

I3DR is designing
bespoke vision systems

i3D robotics is working with partners to develop systems to aid nuclear decommissioning, and is part of the Nuclear Decommissioning Authority team led by Jacobs (formerly Wood Nuclear) to develop 3D mapping technology using stereo vision. Further work is being conducted by I3DR with the National Nuclear Laboratory (NNL), and UKAEA, and Game Changers.

Funding for this work has been granted by Innovate UK with the UKRI, and Sellafield, and managed by FIS360 and the NNL.


I3DR is helping to make nuclear decommissioning a faster, less expensive process and making it safer for human operators.

We are designing and producing bespoke stereo vision cameras and housings to meet the challenges specific to nuclear decommissioning.

The cameras are combined machine learning algorithms and artificial intelligence to map the terrain of unknown environments and identify their contents. These integrated systems can be mounted on robots to be deployed remotely allowing unknown environments to be mapped comprehensively, more quickly than by previous methods of observation and investigation, and with minimum risk to human operators.

  • 3D mapping software creates comprehensive replication of an environment in almost real-time
  • Machine learning algorithms enable the identification of potentially hazardous objects
  • Hardware is small and light enough to be mounted on a robot and manoeuvred around cluttered constrained environments
  • Stereo vision cameras are radiation resistant
  • Robotic mounting allows for remote deployment
  • Tailored solutions for unprecedented levels of resolution and accuracy
  • Can be combined with a Raman spectrometer to provide characterisation of unknown substances
  • Modular design allows for incorporation of other technologies such as virtual reality, thermal imaging or radiation sensors