Wheelchair Navigation System using Node Combination Technique for People with Motor Disabilities

Abstract

Traveling alone in a wheelchair without prior knowledge of the accessibility of the planned route is one of the most common issues that people with motor disabilities face. Ordinary people usually take the shortest route, but people with mobility issues may prefer a longer route with appropriate PWD ramps. The purpose of this paper is to evaluate the applicability of a node combination technique based on Dijkstra's algorithm to a derived graph model for people with motor disabilities using a noble wheelchair navigation system. Experiments were carried out to determine the shortest path from a starting node to a destination node using the proposed routing algorithm. The Graph Online simulation program validated the shortest path length and the route generated by the application. Furthermore, field tests were conducted to determine the accuracy of the proposed approach in terms of distance and travel time. It was accomplished by comparing the application route search results to actual field experiments performed with the 130 routes generated during the process. The inter-rater reliability of the system's measured road distance and travel time compared to the field experiment was then determined using Cohen's Kappa. The study results demonstrated that the navigation system could generate the shortest path from the derived graph model for people with motor disabilities using the proposed routing algorithm, as validated by the simulation program. The study findings also revealed that the derived road distance and travel time from the route search results of the wheelchair navigation system have a moderate agreement with the actual field experiment results.