Motorized or not, wheelchairs are typically maneuvered by users with their hands. Their movement is controlled either by manually moving the wheels with hands or with a joystick in case of electric wheelchair. However, how would a person, who is either an amputee or doesn't have very good motor skills, control their wheelchair?

Researchers in Brazil are working on a prototype wheelchair that is controlled by facial expressions and voice commands. The robotic wheelchair, named Wheelie, is connected to a computer or tablet with Intel RealSense technology (3D camera, speech recognition) that tracks around 80 points on the person's face. Gestures like smile, half smile, sticking tongue out, etc. can be assigned to specific wheelchair movements like move forward, turn left/right, and stop. The camera can detect those facial expressions accurately and maneuver the wheelchair in the desired direction. 

For those who may have trouble with facial expressions (think stroke victims whose facial muscles get affected), Wheelie can also be controlled by head movement. A "kiss" facial expression can activate Wheelie and moving the head forward or back can make Wheelie move in those directions.

If a person is able to speak, the wheelchair can also recognize voice commands. Simple commands like "Move Forward", "Turn Right" and "Stop". At present, Wheelie understands commands in English and Portuguese.

But wait, there's more! If someone cannot perform any of the actions described above, Wheelie can also track eye movement for anyone who have problems with moving their head, speaking or making facial expressions. Just looking at specific corners of the computer screen can make the wheelchair move in different directions!

HOO.BOX, the company that developed Wheelie, is working on refining its prototype and expects it to hit the markets in 2018.

Watch the videos below to see how Wheelie works in different scenarios.

Facial Expressions: