I developed a non-standard control system for an electric wheelchair in my post-graduate studies (described in my Ph.D. thesis, which I handed in in August 2002). The system demands only a minimum of physical effort from its user and particularly does not require using hands or arms. Instead, intentional contractions of a single muscle of choice are used as input signals, which also reduce noise sensitivity in comparison to EEG-based hands-free systems.
To demonstrate the theoretical feasibility of the approach, a prototype consisting of a notebook computer, a mobile EMG device, and a comparatively huge PC-to-wheelchair interface has been developed. The hardware is therefore very voluminous, and this first prototype with its cumbersome operation is not really intended for practical use, although it was able to prove that hands-free control interfaces relying on muscle contractions as input signals are indeed realizable.
At first, my work at the Institute for Mechatronic Systems in Mechanical Engineering concentrated on the extension of this system. The main task was to devise a new prototype which defines its main focus more in the direction of industrial marketing. It also meant miniaturizing the electronics, as well as analyzing and optimizing the system from a 'mechatronic' viewpoint. For the updated prototype, a micro-controller, several LED´s, and various sensors played some role.
In the meantime, I applied the basic control idea to different target systems, and now, I am dealing with alternative input methods in the context of human-computer interaction (HCI) in general.
Further information available: The AID package – An Alternative Input Device based on intentional muscle contractions