Automated Vehicles and Adverse Weather – Phase 3 [flyer]

NHTSA · 2023 · ROSA P / United States. Department of Transportation. Federal Highway Administration

archive: archived pipeline: cataloged verified

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

Summary

The Federal Highway Administration’s “Automated Vehicles and Adverse Weather – Phase 3” project investigated how automated vehicles (AVs) detect and react to adverse weather and road conditions. Motivated by the potential for atmospheric conditions to degrade vehicle performance and driver behavior, the study aimed to identify specific limitations in AV perception systems under controlled stressors. The research utilized two rounds of field tests conducted at the Transportation Research Center, Inc., in Spring/Summer 2020 and Winter 2020/21. These tests challenged AVs with artificial and natural adverse weather, including crosswinds, glare, snow, and ice, across various traffic scenarios such as work zones, lane closures, and signalized intersections. The experimental design involved two production vehicles with differing sensor suites and automation levels in each phase. Summer Field Test #1 compared Vehicle A (camera and radar-based, SAE Level 2) and Vehicle B (lidar and camera-based, SAE Level 2) at 45 mph. Scenarios included lane changes through barrels, navigating brake marks, and handling disappearing lane lines under conditions ranging from baseline dry roads to nighttime wet roads with glare. Winter Field Test #2 compared Vehicle A (camera-based, SAE Level 2) and Vehicle B (lidar and HD map-based, SAE Level 3) at speeds between 15 and 45 mph. Scenarios focused on lane keeping and changing on snow-covered or ice-covered roads, as well as navigating signalized intersections and detecting stopped cars on ice. Key findings revealed significant inconsistencies in AV performance both across different vehicles and between repeated runs of the same vehicle. In both phases, the vehicle equipped with lidar and HD map technology (Vehicle B in both tests) demonstrated superior performance compared to the camera-reliant Vehicle A. Specific failures included localization loss, excessive deviation from programmed paths, rapid acceleration and deceleration at snow-covered intersections, and the inability to drive close to centerlines with varying snow depths. The tests successfully challenged the limitations of the automation systems, highlighting that while vehicles might complete maneuvers in most scenarios, their inconsistent performance in adverse conditions poses a risk of driver over-trust, distracted driving, and inappropriate reliance on automation. The study concludes that substantial differences in automation approaches—specifically regarding sensing, processing, and alert presentation—significantly impact performance. Crucially, the results underscore the necessity for redundant sensing and control systems. The observed losses in localization and steering control during winter conditions indicate that redundancy in perception, braking, actuation, and localization is essential for the safe operation of automated driving systems under all weather and environmental conditions. The findings suggest that current systems are not yet robust enough to handle complex adverse weather without significant risk, necessitating further development in sensor redundancy and algorithmic resilience.

Key finding

Automated vehicles exhibited significant performance inconsistencies in adverse weather, with lidar-based systems demonstrating superior reliability compared to camera-based systems, underscoring the critical need for redundant perception and control systems.

Methodology

field_study

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.

• situational awareness