A human-computer interface modelled with Petri nets for disabled transportation

Hardware/software codesign is a difficult issue to deal with, all the more so as a unique, structured and automatic methodology, providing the acceleration of design process and the dynamic evaluation of different compromises, is still lacking. Data flow Petri nets that we studied are an efficient solution to make it possible, combined with the use of a hardware description language, so that they can be implemented on programmable chips. The example described in this paper deals with a wheelchair whose hand lever can adapt to the user's handicap. For some disabled people, the use of a wheelchair can be difficult, for physical or cognitive impairments. Moreover, a choice has been made for modular structure and non-specific material for practical reasons. Therefore, the Fauteull Roulant A Commande Adaptee au Handicap (wheelchair with adaptive command for people with disabilities) project has been developed. The lever is handled by an artificial neural network that records functional limitations of the hand, and then compensates them. First experiments and first implementation studies of a neural network give encouraging results.