Thirty years ago, the SVEN project achieved good results in controlling a multi-degree-of-freedom prosthetic arm. The system used a Bayesian EMG classifier, and the prosthesis motors were controlled in a simple on/off fashion. The poor hardware technology of that time effectively prevented the results from ever becoming a practically available system, but the control results are still unsurpassed.
The Institute of Biomedical Engineering of the University of New Brunswick (UNB), Canada, has pioneered rese arch in myoelectric prosthesis control for decades. As a cooperative effort, the Department of Engineering Cybernetics, NTNU, and UNB now propose to use SVEN methods as a starting point for advanced techniques that allow proportional, coordinated joint co ntrol. The project will also focus on control inputs from proximal sites (e.g. EMG from the shoulder region as opposed to the forearm) to accommodate high-level amputations. Research conducted at the two centres will be incorporated in the new methods. Bo th centres are parts of major international cooperations that are largely complementary to N-ARM and with which these activities will be sought coordinatad.
N-ARM implies a qualitative and quantitative intensification of the centres' present minor cooper ation. UNB has stated its intention to dedicate at least a PhD grant to the N-ARM project, which will additionally at least include one PhD student at NTNU and a postdoctoral fellowship plus 1/2 res. engineer. There is a significant degree of overlap but also complementariness in the centres' laboratory facilities and competence profiles, which necessitates mobility and a real cooperative effort. If successful, the N-ARM activities most likely will result in a continued cooperation in the future.
The "Pr e N-ARM" project, which is the focus of the present application, aims at funding the detailed planning and establishment of the cooperation, including the incorportation of industrial partner(s) in the group.