Towards Comfortable and Socially Acceptable Navigation in Autonomous Motorized Wheelchairs

• Kristofer Kappel UFPel
• Paulo R. Ferreira UFPel

Resumo

Autonomous Motorized Wheelchairs (AMWs) are an emerging technology that requires effective navigation strategies while respecting social rules, akin to autonomous cars and service robots. Just as service robots, AMWs need to navigate scenarios while considering social rules and acting in a manner that aligns with human behavior. This is further complicated by the fact that AMWs carry humans, necessitating a consideration of both passenger and pedestrian comfort. In this paper, we present a solution that integrates various social rules into a navigation scenario to enhance comfort for both passengers and pedestrians for indoor environments. The experiments incorporate comfort in the form of proxemics, corridor lane detection, and visibility estimation, enabling socially-aware navigation for AMWs. Experiments conducted in a simulated environment using ROS and Navigation Stack demonstrate the effectiveness of the proposed approach. The results highlight the impact of integrating these social rules on navigation behavior and comfort. This research contributes to the advancement of socially-aware navigation systems for AMWs, with potential applications in healthcare and assistive robotics.

Palavras-chave: Social Navigation, Autonomous Wheelchairs, Human-Robot Interaction