Enabling comfortable passing between humans and autonomous mobile robots

< Overview >

A research team from the Visual Perception and Cognition Laboratory and the Cognitive Neurotechnology Unit, Department of Computer Science and Engineering, Toyohashi University of Technology, Japan, investigated whether human pedestrians exhibit a consistent tendency to move either left or right to avoid autonomous mobile robots during head-on encounters. The results showed no consistent directional preference among the participants; instead, individual differences were observed in the chosen direction of avoidance. However, the direction in which the pedestrians moved, either left or right, could be reliably predicted by analyzing the orientation of their waist during walking. To respond accordingly, the robots were equipped with a function that enabled them to detect human waist orientation and adjust their movement in real time. This study also examined whether the timing of the robots' avoidance behavior influenced pedestrians’ comfort during passing. Overall, the findings indicated that early avoidance behavior by the robots, based on the predicted direction of human movement, led to an improvement in the pedestrians’ perceived comfort during passing. This study was published online in the journal PLoS One on May 14, 2025 (https://doi.org/10.1371/journal.pone.0323632).

< Details >

Autonomous mobile robots are increasingly being deployed in diverse settings such as urban logistics, food and beverage delivery, and tourist information services. To become more familiar to humans and smoothly integrated into human-centered environments, these robots must exhibit movement patterns and behaviors that are intuitively acceptable to people. The research team suggests that incorporating human cognitive characteristics and behavioral tendencies into robot motion planning can facilitate more comfortable human–robot coexistence.

Previous studies conducted in countries with right-side traffic regulations have shown that, in virtual reality environments, pedestrians tend to avoid oncoming individuals by moving to the right. In contrast, when the approaching entity is an inanimate object (e.g., a cylinder), no consistent directional bias is observed. These findings suggest that avoidance behavior may depend on whether the oncoming entity is perceived as human or non-human. However, human responses to robots approaching head-on have remained insufficiently explored.

Consequently, this study investigated whether humans tend to avoid a real robot approaching head-on by moving to the left or right. In the experiment, a robot advanced toward participants from a distance of 5 meters at one of five different approach angles. Participants were instructed to walk toward a designated goal while avoiding the robot by moving either left or right. All participants were right-handed and from Japan, where traffic follows left-side driving rules.

The results revealed no consistent directional bias in avoidance behavior—participants did not systematically prefer the left or right side. Instead, considerable individual differences were observed, which did not appear to depend on handedness or cultural traffic norms. Notably, the orientation of participants’ waists immediately before the avoidance maneuver reliably corresponded to the direction in which they intended to move to avoid the robot.

Building on this finding, we developed an algorithm that predicts the direction of human avoidance based on waist orientation. This predictive mechanism was implemented in a robot, and we examined how the timing of its avoidance behavior influenced pedestrians’ sense of comfort during passing. The experiment compared six conditions, including variations in the timing of the robot’s avoidance maneuvers, as well as a control condition in which the robot did not attempt to avoid the pedestrian. The results showed that pedestrians felt more comfortable when the robot began its avoidance maneuver at an earlier stage.

Tatsuto Yamauchi, a first-year doctoral student and co-first author of the paper, commented:“This research has provided valuable insights into designing movement patterns that allow humans and robots to pass each other more naturally. However, several challenges remain for future development. In particular, improving the accuracy and reliability of the technology for predicting avoidance direction based on waist orientation will be crucial.”

< Future Outlook >

In the future, we plan to improve the accuracy of avoidance direction prediction and conduct experiments in diverse environments to more comprehensively examine the factors that influence human comfort and sense of security. Through these efforts, we aim to contribute to the development of urban spaces where humans and robots can coexist more comfortably.

< Publication Information >

Yamauchi, T†., Tamura, H†*., Minami, T., & Nakauchi, S. (2025). Waist rotation angle as indicator of probable human collision-avoidance direction for autonomous mobile robots. PLoS One, 20(5), e0323632. https://doi.org/10.1371/journal.pone.0323632