Interactive collision avoidance based on surrounding mobility type for an intelligent powered wheelchair

Abstract

Electric powered wheelchairs have been developed for people with disabilities and the elderly who have difficulties walking and participating in social activities. However, for users with decreased cognitive abilities and judgment due to further aging, mobility aids that are not only powered but also intelligent are required. Intelligent powered wheelchairs with autonomous locomotion functions are one solution to this problem. This research focused on the complexity of coexisting "mobility types" on sidewalks and a prototype intelligent wheelchair equipped with interactive strategies for collision avoidance based on the surrounding mobility type was developed. In order to classify the mobility type of coexisting traffic participants, image-processing components combined with LIDAR were developed. Based on the classification results, the wheelchair selected and executed either a changing-lanes strategy or deceleration strategy autonomously. © 2013 Ito and Kamata; licensee InTech.