Human Merging Behavior in a Coupled Driving Simulator: How Do We Resolve Conflicts?

Siebinga, Olger; Zgonnikov, Arkady; Abbink, David A. · 2024 · Crossref

DOI: 10.1109/ojits.2024.3349635

archive: archived pipeline: cataloged verified

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

Summary

This study addresses the limitations of naturalistic data in understanding human merging behavior, specifically the inability to observe operational control inputs and control initial vehicle kinematics. To investigate how drivers resolve merging conflicts, the authors conducted a controlled experiment using a coupled, top-down driving simulator. The experiment involved 9 pairs of participants interacting in a simplified merging scenario where they controlled acceleration but not steering. The study aimed to determine how initial positions and velocities influence the outcome of the interaction (who merges first) and the time required to resolve the conflict, while also analyzing the specific control strategies drivers employ. The experimental design featured 11 conditions varying in projected headway and relative velocity, ensuring that maintaining initial velocities would result in a collision. Participants were instructed to prevent collisions without communicating, with vibration feedback aiding speed perception. The researchers analyzed the data using Conflict Resolution Time (CRT), defined as the time from the start of the interaction until the vehicles were no longer on a collision course. Statistical models, including mixed-effects logistic and linear regressions, were used to assess the relationship between initial kinematics, conflict outcomes, and resolution times. The results demonstrated that initial kinematics strongly predict the outcome of the merging conflict. Increasing projected headway advantage significantly increased the likelihood of a driver merging first, while higher relative velocity decreased this probability, indicating that faster drivers tended to yield to slower ones with position advantages. Conflicts where one driver had a pure headway advantage were resolved more quickly than those with pure velocity advantages. The neutral condition, where neither driver had an advantage, exhibited the highest median CRT, reflecting the difficulty of negotiation without a clear kinematic cue. Furthermore, the analysis of individual behavior revealed that drivers did not continuously optimize their actions. Instead, they utilized intermittent piecewise-constant control, applying constant acceleration inputs at key decision moments rather than continuous adjustments. These findings challenge the assumption that drivers continuously optimize expected utility during interactions. The study concludes that human merging behavior is better characterized by discrete decision moments and constant acceleration phases. This evidence advocates for the development of interaction models based on intermittent piecewise-constant control rather than continuous optimization frameworks. By providing controlled insights into reciprocal driver interactions, this work contributes to the fundamental understanding of human driving behavior and supports the design of socially acceptable, human-like autonomous vehicle systems.

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 2026-06-09; verification: verified.

Topics

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

• gap acceptance
• driver vru interaction
• traffic density
• following distance
• mental model of traffic

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: behavioral performance data
• Theoretical Contribution: computational model, theory or model