Hoobox Robotics, a robotics company based in São Paulo, Brazil, has developed the “Wheelie 7”, a wheelchair controlled using facial recognition technology. Incorporating AI developed by Intel, the technology allows users to control the movements of a motorized wheelchair using just their faces.
The technology is envisaged as being particularly helpful for users who cannot use their hands to control a motorized device. The tech consists of a 3D camera that records a user’s facial expressions (no body sensors are required) and an on-board computer that interprets the expressions and sends commands to control the movement of the wheelchair.
The company claims that their facial recognition system is so sensitive that it can differentiate ten different levels of pain, detect drowsiness, agitation, and sedation, and can even detect when a person will sneeze before the event occurs. Even raising an eyebrow is enough to function as a specific command for the wheelchair.
As the technology is available as a kit, users can convert a pre-existing motorized wheelchairs into facial recognition-controlled systems. It supposedly takes just seven minutes to convert a wheelchair, hence the name “Wheelie 7”.
See a video of the system in action below:
Medgadget asked Paulo Pinheiro, Co-Founder and CEO of HOOBOX Robotics, some questions about the system.
Medgadget: How did this idea come about? Is there a large unmet need in terms of people who aren’t currently able to control a motorized wheelchair?
Paulo Pinheiro: I have a PhD in robotics and have always been interested in how we can use robotics to help people. I got the idea for The Wheelie a few years ago, when I was at an airport and saw a young woman in a wheelchair who had the most beautiful smile. And I thought wouldn’t it be amazing if she could control her wheelchair with that smile.
According to the National Spinal Cord Injury Statistical Center, there are around 288,000 people in the U.S. living with spinal cord injuries with about 17,700 new cases every year. Studies have shown that physical mobility has the largest impact on the quality of life for people with spinal cord injuries, and at HOOBOX Robotics, we want to help them regain their mobility.
Medgadget: Please give us an overview of the components of the Wheelie 7 system. How difficult is it to install the Wheelie 7 kit on a conventional motorized wheelchair?
Paulo Pinheiro: The Wheelie is a kit that is installed on a user’s motorized wheelchair in just 7 minutes. It uses an Intel Real Sense Camera to capture facial expressions that are the interpreted by AI algorithms that run on a small Intel NUC mini pc. The facial expressions are interpreted in real time thanks to Intel Core processors and the Intel Distribution of OpenVINO Toolkit. This is so important because the wheelchair has to react to the user’s facial expressions immediately so that they can stop before running into objects or turn when necessary.
Medgadget: Please give us an overview of the AI system that powers the facial recognition functionality, and your collaboration with Intel.
Paulo Pinheiro: We worked closely with Intel on the hardware and software for The Wheelie, because their technology is both cutting-edge and provides incredible precision. For example, thanks to the Intel Real Sense 3D Camera, the Wheelie can detect hard facial expressions like “tongue out” or “puffed cheeks”, and 8 more, all of them regardless of the position of the head.
Medgadget: How easy is the system to use? Do users need prior training?
Paulo Pinheiro: The Wheelie 7 is very easy to use, right from the start. The kit currently recognizes 10 pre-trained facial expressions that each user can pick from to move forward, turn right, turn left and stop. The users don’t need prior training.
Medgadget: The system is currently in beta testing. How has it been received by users so far?
Paulo Pinheiro: There are currently over 60 people using the Wheelie and providing feedback to us. We recently surveyed them and found that of the respondents, 100% of them said they would recommend the Wheelie to their friends and family members. 100% of them said the Wheelie was more effective than other solutions they’ve tried. 96% said that the Wheelie was the most effective interface for adverse terrain and lighting conditions.
Link: Hoobox Robotics…