№ 274/2011 from Sep 02, 2011
Computer scientists from Freie Universität Berlin demonstrated a new type of wheelchair at IFA, an international trade fair for home electronics. The wheelchair, on loan from the Otto Bock company, makes it significantly easier to navigate inside buildings. It is equipped with laser and camera sensors and a computer under the seat. Laser sensors detect the position of walls and obstacles and prevent collisions. A so-called Kinect, developed for Microsoft Xbox 360 game consoles, is also mounted on the wheelchair. The sensor detects the three-dimensional structure of the environment and can, for example, prevent the collision of the wheelchair with people. Demonstrations and explanations are available in videos on YouTube.
A camera is installed for steering using eye movements. To cause the wheelchair go to the right or left, the user needs only to glance toward the right or left. Accelerating and braking are triggered by looking upward or downward. For steering by thought, the wheelchair user wears a cap with 16 sensors that continuously measure brain activation. The system is trained to distinguish four brain patterns: drive left, drive right, accelerate, and brake. After a training period, the user should be able to steer the wheelchair just by thinking. A great deal of concentration is required, as ideally the user should think of only the four practiced patterns the entire time. Since obstacles automatically cause the wheelchair to stop, the person remains accident-free in any case. Previously, the group had demonstrated steering a car using only brain power.
The intelligent wheelchair is based on research from the AutoNOMOS Labs at Freie Universität Berlin. Scientists working in the innovation labs funded by the German Federal Ministry of Research are working on the development of autonomous and driver-assisted systems with the aim of preventing traffic accidents in the future and to improve traffic safety through the use of modern sensor and computer technology. In the labs future scenarios are simulated to investigate steering in various types of vehicles, both conventional and electric cars, as well as wheelchairs. Voice recognition for wheelchairs is in the planning stage. The idea is that the user would be able to command the wheelchair to go into a certain room or to follow a person. Robotic gripper arms will be added to the wheelchair, permitting the user to access objects on higher shelves.
David Latotzky, Institute of Computer Science, Freie Universität Berlin, Artificial Intelligence Group, Tel.: +49 (0)30 / 838- 75127, Email: firstname.lastname@example.org
Brain and eye controlled intelligent wheelchair: www.youtube.com/watch?v=yNpOgEbKHbw
AutoNOMOS Innovations Lab: www.autonomos-labs.de