Driving Etiquette

Peng, Huei; Huang, Xianan · 2019 · ROSA P / University of Michigan. Center for Connected and Automated Transportation

archive: archived pipeline: cataloged verified

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

Summary

This study addresses the challenge of designing autonomous vehicles (AVs) that integrate seamlessly into mixed traffic environments by adhering to local "driving etiquette." The authors argue that while AVs must avoid collisions, they must also behave similarly to human drivers to prevent causing accidents or disrupting traffic flow. To achieve this, the research aims to quantify key parameters of naturalistic human driving behavior to serve as design constraints for AV control algorithms. The researchers utilized data from the Safety Pilot Model Deployment (SPMD) project, a large-scale naturalistic driving dataset collected in Ann Arbor, Michigan. The analysis focused on three specific driving scenarios: free-flow driving, car-following, and lane-changing. For car-following and lane-change events, data was extracted from 98 vehicles equipped with Mobileye cameras, yielding 161,009 car-following events and 422,249 cut-in cases. Free-flow behavior was analyzed using a Gaussian Mixture Model applied to 321,945 trips from 2,468 drivers. Statistical models, including Generalized Extreme Value (GEV) and lognormal distributions, were fitted to characterize acceleration limits, time headways, yaw rates, and gap acceptance behaviors. The results reveal distinct behavioral patterns between highway and local road driving. In free-flow conditions, human drivers on highways significantly exceeded posted speed limits, whereas local road speeds closely matched limits. For car-following, the mean time headway was 1.42 seconds on highways and 2.07 seconds on local roads. Extreme deceleration limits were similar across both environments (approx. -2.7 m/s²), but acceleration limits were higher on local roads (1.19 m/s²) than on highways (0.72 m/s²). Regarding lane changes, local maneuvers were more aggressive, with a mean maximum yaw rate of 1.4 deg/s compared to 0.6 deg/s on highways, and shorter durations (2.0 s vs. 4.3 s). The study also identified that highway lane changes often involve mandatory merges with smaller gaps, distinct from discretionary changes. These findings provide specific, data-driven parameters for designing AV controllers that mimic human-like behavior. By adopting these statistical limits for acceleration, headway, and lane-change dynamics, AVs can avoid appearing overly cautious or aggressive, thereby improving safety and traffic compatibility. The authors conclude that these parameters can guide the development of AV algorithms and simulation software, with future work involving validation in testing facilities like Mcity and analysis of data from other geographic regions to assess generalizability.

Key finding

Human drivers maintain a mean car-following time headway of 1.42 seconds on highways and 2.07 seconds on local roads, with extreme deceleration limits averaging -2.81 m/s² on highways and -2.64 m/s² on local roads.

Methodology

naturalistic

Sample size: 2800

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. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

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

• following distance
• traffic density
• speed choice
• lane changing
• lane positioning
• steering pattern

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
• Methodological Resource: dataset resource
• Theoretical Contribution: computational model