A powered wheelchair that uses artificial intelligence (AI) and the user’s eye movements to navigate itself has won £35,000 of funding in a competition designed to offer mobility innovations a shortcut to the market.
The research prize will help a team at Imperial College London to develop the self-driving, self-navigating wheelchair for paralysed people who can lead with their gaze. They also hope to make it available to the public at low cost.
Dr Aldo Faisal from Imperial’s Departments of Computing and Bioengineering, who leads the project, said: “Our wheelchair will help people to navigate their homes and the world outside. If the user can move their eyes to look at the world, they can operate the wheelchair that reads their intentions from their eyes.”
The prize was announced as part of Mobility Unlimited Challenge Discovery Award, which is funded by Toyota Mobility Foundation and run by Nesta’s Challenge Prize Centre.
The self-driving artificially intelligent wheelchair combines existing widely available technologies, such as eye tracking system and laptops, with electric wheelchairs. It uses a Light Detection and Ranging (LiDAR) sensor, which is an infrared-based sensor commonly used on self-driving cars, to build a 360-degree map of the user’s environment.
The eye tracker gathers information on eye movements, and the AI programme defines where the wheelchair should move to, and guides it to avoid obstacles.
The AI system uses the LiDAR to build maps of the environment as the wheelchair moves about, just as in self-driving cars.
Thanks to the AI and LiDAR components, users can now specify that they want to drive out of a cluttered kitchen, and leave it to the wheelchair to navigate the route and avoids obstacles without further human input.
Dr Faisal said: “Our wheelchair is a great example of a frugal innovation. It also demonstrates that the power of AI is becoming accessible and affordable to regular people. We at Imperial are harnessing its power to improve lives.”
Graduate students on Aldo’s team, Noyan Songur and Mahendran Subramanian, said: “Such intelligent AI systems can learn and improve specific to the user over time, so we are able to work with patients from day one to pinpoint their needs and expectations. The technology will help severely disabled people restore their independence at low cost.”
The next stage of the Mobility Unlimited Challenge will be to compete for $1 million (£700,000) prize, for which the winner will be announced at the Tokyo 2020 Summer Paralympics. If won, this will go towards making necessary improvements and putting the wheelchair on the market.