Brazilian researchers have developed a wheelchair that can be controlled through small facial, head or iris movements. The team at Faculdade de Engenharia Elétrica e de Computação da Universidade Estadual de Campinas (FEEC/Unicamp) says the technology could help people with cerebral palsy, those who have suffered a stroke or live with amyotrophic lateral sclerosis and other conditions that prevent precise hand movements.
The team started to look into brain-computer interface (BCI) techniques (the acquisition and processing of signals that allow communication between the brain and an external device) in 2011. The group then decided to test those techniques in a real life situation.
The researchers built their prototype from a standard motored chair, removed the joystick and equipped it with sensors that can gauge the distance between walls and other objects, and pick up variations on floor surface.
A notebook that sends commands directly to the chair using a 3D camera running on Intel’s RealSense technology was installed, which caters for interaction with a computer through facial and body expressions. RealSense is a stand-alone camera that uses depth-sensing technology and can be attached to any computer. The set includes a standard camera, an infrared laser projector, an infrared camera and a microphone array.
“The camera can identify more than 70 facial points around the mouth, nose and eyes. By moving these points, it is possible to get simple commands, such as forward, backward, left or right and, most importantly, stop,” says the researcher Eleri Cardozo. Voice interaction is also possible, but the researchers say it less reliable because of differences in voice timbre and ambient noise.
The chair was also equipped with a Wi-Fi antenna that allows a caregiver to steer the equipment remotely through the internet, which could be handy should a wheelchair user get tired.
For patients with more serious conditions where facial movement is not possible, the team is also looking into a BCI technology that can pick signals directly from the brain through external electrodes and turn them into commands, though this type of equipment has not been added on the robotic chair yet.
The research team recently received extra funding to move forward with the project so it can be adapted and marketed in Brazil within the next two years. “Our objective is that the final product costs, at most, twice as much as joystick-controlled models, which sell for around R$7,000 (US$1,994).” To do that, a start-up called Hoo-Box has been created, which will focus on developing the Wheelie system further.
The researchers presented their project at the Third Brainn Congress in the city of Campinas, Brazil, in April 2016.