The new UKRI Centre for Doctoral Training in Dependable and Deployable Artificial Intelligence for Robotics, shortened to CDT-D2AIR, will train new PhD students in verification and certification systems for robotics and AI.

The aim is to ensure that robotics applications in the home and in industry — such as AI assistants to help with independent living, or robotic machinery in manufacturing and construction — can interact safely with the environment and users.

The robotics safety centre is being awarded a share of £117 million funding by UK Research and Innovation (UKRI) as one of 12 new Centres for Doctoral Training being announced by the UK government this week, in the run up to the AI Safety Summit 2023.

It will be based at the Edinburgh Centre for Robotics (ECR), a joint initiative between Heriot-Watt University and the University of Edinburgh.

ECR includes the National Robotarium, the centre for robotics and artificial intelligence based at Heriot-Watt’s Edinburgh campus, and the Bayes Centre, the University of Edinburgh’s innovation hub for data science and artificial intelligence.

The state-of-the-art facilities include living labs, a robotic testing and development space, and more than £20m in robotics equipment.

Students working in the D2AIR Centre for Doctoral Training will be able to simulate and test their solutions on this equipment, and will be trained in the latest methods in AI, verification, design, and robotics, the project partners say.

Recommended reading

• UK Unveils New £118m AI Skills Package
• Robotics Programme Inspiring the “Next Generation of Robot Scientists”
• UKRI Plan Sees Significant Investment in the UK Tech Sector

Ron Petrick, professor of computer science at Heriot-Watt University and a researcher in automated planning and robotics, is director of the new D2AIR Centre for Doctoral Training.

He said: “There is a clear need for robotics and AI systems be certifiable, reliable and capable of interacting safely with people and the environment. While AI methods are being increasingly used in robotics, much of this technology was not originally designed with safety and other important human-centred requirements in mind.

“Making AI truly applicable to, and deployable in, robotic solutions will require advanced sets of skills and a new way of thinking.”

Petrick said examples such as generative AI assistants – embodied as robots or speakers – that are deployed in the home to help with independent living, had to converse and collaborate safely.

Robots that can sense and interact in their environment for manufacturing and construction also had to work in challenging and extreme environments.

“Crucially, in all of these cases, the safety of autonomous systems is key to their deployability, especially when they need to operate around people,” he added.

Petrick also said D2AIR would build a new talent pipeline of graduates who can think in new ways around researching, designing, building, and deploying dependable and safe robotics solutions.

“This pipeline will increase productivity, catapulting the UK as a world-leader in trustworthy robotics and AI technologies,” he continued.

UKRI said £117 million funding in total had been awarded to 12 Centres for Doctoral Training in Artificial Intelligence based at 16 universities.

The investment has been made to help ensure that the UK has the skills needed to seize the potential of the AI era and nurture talent.