Multimodal System for Training at Distance in a Virtual or Augmented Reality Environment for Users of Electric-Powered Wheelchairs

摘要

Recently, important progresses for controlling electric-powered wheelchair were made for people with disabilities. However, a significant amount of people affected by severe physical disabilities still cannot take advantage of autonomous mobility. For those situations, the use of biological signals to control the assisted environment emerges as a possible solution. In this scenario, the act of driving an electric-powered wheelchair without training can appear as a serious safety risk, which can be solved by using a virtual driving simulator. Nevertheless, when using biomedical signals as commands (eg.: brain-computer interface - BCIs), it is not possible to ensure a continuous and reliable control of the wheelchair due to state of the art of this technology, being necessary to associate it with features such as semi-automatic obstacle detection or contour, which are difficult and costly to reproduce. Thus, it is interesting to offer to new wheelchair's users the possibility of using a simulator that allows them to learn to drive the wheelchair prototype at distance, making use of appropriated telematics techniques, which is the proposal of this work. This research joints complementary skills from the following universities: Federal University of Amazonas (UFAM), Federal University of Espirito Santo (UFES) and Federal University of Uberlandia (UFU), with the collaboration of researches from the University of Lorraine (UL) in Metz-France.