Dyson invests £5m into research of robots that can see

Dyson has invested £5 million into a joint robotics lab with Imperial College London to develop ‘intelligent domestic robots’ that can see.

James Dyson holding a digital motor which relies on algorithm controls to measure the speed of impeller rotation and fire digital pulses to magnets which make sure the impeller maintains optimum speed
James Dyson holding a digital motor which relies on algorithm controls to measure the speed of impeller rotation and fire digital pulses to magnets which make sure the impeller maintains optimum speed

The five-year investment programme will focus on the development of ‘vision systems’ – comprising sensing, interpretation of environment and  interaction – which effectively amounts to a sense of sight.

Dyson hopes that its experience with electronics and mechanics will complement the vision systems and lead to the creation of ‘a generation of robots that understand the world around them and can intelligently interact as it changes’.

James Dyson says, ‘Robots still lack understanding; seeing and thinking in the way we do. Mastering this will make our lives easier and lead to previously unthinkable technologies.’

Dyson has been working on researching robotics for the last 15 years. Since 2005 it has been working with Imperial College and Professor Andrew Davison to develop machines that use vision to logically navigate their surroundings.

The Dyson Robotics Laboratory will initially use the cash to recruit 15 scientists – including five PhD researchers and six post-doctoral researchers – who will joined by Dyson research, software, and electronics engineers. Imperial is currently recruiting for the positions.

Davidson, who is professor of robot vision at the department of computing, will be the director of the Dyson Robotics Laboratory at Imperial College.

He has already led research in computer vision and robotics, developing Simultaneous Localisation and Mapping (SLAM) systems.

Davidson’s version uses a single camera to create a 3D map of a room without prior knowledge. It could be used by a robot to navigate a space via a wireless connection.

Davidson says, ‘A truly intelligent domestic robot needs to complete complex everyday tasks while adapting to a constantly changing environment.

‘We will research and develop systems that allow machines to both understand and perceive their surroundings – using vision to achieve it.’

Start the discussionStart the discussion
  • Post a comment

Latest articles