Passenger discomfort map for autonomous navigation in a robotic wheelchair

Yoichi Morales, Atsushi Watanabe, Florent Ferreri, Jani Even, Kazuhiro Shinozawa, Norihiro Hagita

Research output: Contribution to journalArticlepeer-review

13 Citations (Scopus)

Abstract

This work presents a navigational approach that takes into consideration the perception of comfort by a human passenger. Comfort is the state of being at ease and free from stress; thus, comfortable navigation is a ride that, in addition to being safe, is perceived by the passenger as being free from anxiety and stress. This study considers how to compute passenger comfortable paths. To compute such paths, passenger discomfort is studied in locations with good visibility and those with no visibility. In locations with good visibility, passenger preference to ride in the road is studied. For locations with non-visible areas, the relationship between passenger visibility and discomfort is studied. Autonomous-navigation experiments are performed to build a map of human discomfort that is used to compute global paths. A path planner is proposed that minimizes a three-variable cost function: location discomfort cost, area visibility cost, and path length cost. Planner parameters are calibrated toward a composite trajectory histogram built with data taken from participant self-driving trajectories. Finally, autonomous navigation experiments with 30 participants show that the proposed approach is rated as more comfortable than the state-of-the-art shortest planner approach.

Original languageEnglish
Pages (from-to)13-26
Number of pages14
JournalRobotics and Autonomous Systems
Volume103
DOIs
Publication statusPublished - 2018 May

Keywords

  • Autonomous navigation
  • HRI
  • Human comfort
  • Human factors

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Software
  • Mathematics(all)
  • Computer Science Applications

Fingerprint

Dive into the research topics of 'Passenger discomfort map for autonomous navigation in a robotic wheelchair'. Together they form a unique fingerprint.

Cite this