Printer friendly version Share

News Release

Take control of a wheelchair with a steady jaw and a wink

11 May 2011 Inderscience

For people with severe physical disabilities, such as spinal cord injury, quadriplegia and hemiplegia or amputation, current technology for controlling a wheelchair or mobility scooter is wholly inadequate. A research paper published this month in the International Journal of Mechatronics and Automation now shows how an inexpensive webcam and a bio-signal sensing headband can be used to control the steering and propulsion of an electric wheelchair.

Lai Wei and Huosheng Hu in the School of Computer Science and Electronic Engineering, University of Essex, in Wivenhoe Park near Colchester, have developed a novel hybrid human-machine interface (HMI) for hands-free control of electric powered wheelchairs.

Their prototype system uses pattern recognition and analysis of bio-signals from the user's forehead electromyography and visual signals to identify which of five winking and jaw clenching movements the user is making. The combination of a left and right wink with and without a jaw clench and a jaw clench alone are mapped to six control commands for the wheelchair.

In tests with users on an indoor obstacle circuit the team found that all users were able to control a wheeled-robot simulator initially. They then demonstrated that they could also control a wheelchair sufficiently well to navigate the obstacle course and to follow specific routes. The researchers point out that individual wheelchair users would have different needs but the recognition software could be fine-tuned to suit particular facial characteristics and muscular deficits.

"There are so many different kinds of disability. Therefore, we need to develop different kinds of assistive technology for people to choose," says Hu.

Future research could also examine whether the same technology might encompass additional facial expressions and movements in order to control other devices.

App animated no finger WCSJ ad eNEWS ad2 Facebook-Twitter3 Translation page link