Interactions of Automated Vehicles with Road Users

Dey, Debargha; Ackermans, Sander; Martens, Marieke; Pfleging, Bastian; Terken, Jacques · 2022 · Crossref

DOI: 10.1007/978-3-030-77726-5_20

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the critical challenge of intent communication between autonomous vehicles (AVs) and pedestrians, a problem arising from the absence of human drivers who typically signal intent through eye contact or gestures. The authors propose an Intent Communication System (ICS) designed to bridge this gap by providing clear, simple signals to prevent deadlock situations and build trust. The research is motivated by the need to understand the psychological aspects of human-machine interaction and to develop a system that is robust, cost-effective, and intuitive for pedestrians with varying levels of familiarity with AV technology. The study employs a mixed-methods approach combining real-world field tests and computational simulations. For the field tests, an autonomous golf cart was outfitted with an ICS consisting of strobe lights, an LED word display, and speakers to convey messages such as "please cross." The system was designed to avoid the "Uncanny Valley" effect by using familiar road signage elements rather than anthropomorphic features. Seventy-six participants were involved, with 26 engaging in real-world interactions divided into four groups based on the presence of the ICS and prior knowledge of the vehicle. The vehicle was manually controlled via transmitter to ensure safety while maintaining the illusion of autonomy for the participants. Additionally, simulations using a Decentralized Markov Decision Process (Dec-MDP) framework were conducted to model pedestrian behavior and quantify trust, incorporating data from pre-surveys that identified potential pedestrian actions. The results demonstrate that the ICS significantly improves pedestrian interaction outcomes. In real-world tests, participants with access to the ICS (Groups 1 and 2) exhibited higher trust levels, felt safer, and behaved more predictably than those without it (Groups 3 and 4). Specifically, Group 1 (with ICS and prior knowledge) showed the highest trust and shortest interaction times, while Group 4 (no ICS, no prior knowledge) displayed the lowest trust and highest hesitation. Video analysis revealed that participants without the ICS were more likely to look around for cues and walk faster, indicating uncertainty. Simulation results further validated these findings, showing a 142 percent difference in pedestrian trust when the ICS was enabled and the pedestrian had prior knowledge compared to when it was disabled and no prior knowledge existed. The mathematical model successfully quantified how explicit communication reduces the stochasticity of pedestrian actions. The significance of this work lies in its contribution to the design of socially acceptable autonomous systems. By integrating psychological insights with hardware and algorithmic design, the authors provide a framework for building trust between humans and machines. The study highlights that simple, explicit communication is more effective than complex or anthropomorphic interfaces in initial interactions. The findings suggest that ICS can mitigate the uncertainty inherent in AV-pedestrian encounters, thereby enhancing safety and facilitating the broader societal adoption of autonomous vehicles. The paper establishes a baseline for evaluating intent communication systems and underscores the importance of considering user psychology in the development of autonomous robotics.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• ehmi external hmi
• automation surprise
• trust calibration
• acceptance adoption
• automation
• trust in automation foundations

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: self report data
• Theoretical Contribution: computational model