Human-Robot Interaction in Urban Areas – Development and Validation of Acceptance Proto-model in the Scope of Proximity Theory

Abstract

It is envisioned that autonomous logistic robots share road system space with humans in the future. For pedestrians to accept and interact with such robots, it is essential to research how pedestrians behave in relation to such logistic robots in real-life environments. This allows the derivation of rules for logistic robots’ spatial behaviour. In this study, a sensor-equipped logistic robot was controlled in a wizard-of-oz design and appeared to move autonomously through a pedestrian-shared space. Through the controlled variation of road infrastructure factors such as path width along a pre-defined route and the structured quasi-experimental analysis of varying environmental factors such as pedestrian encounters and initial distance and trajectories, we derive factors that relate to pedestrians’ proximal behaviour towards logistic robots in real-life environments. In addition to the robot-sensor-based analyses, human-robot encounters are classified by an accompanying human observer, who marks critical interactions such as the robot blocking the way for a pedestrian or robot-related pedestrian trajectory changes. Effects that are known from laboratory studies are found, e.g. smaller proximal margins of lateral encounters with the robot compared to longitudinal approaches. Basic critical scenarios are identified. A correlation between environmental conditions, pedestrian density, and minimal avoiding distance is observed.