A Storm Brewing: Severe Weather and Its Impacts on the Commercial AV Industry

Authored by Gregory J. Quinn, Director, Meteorology Practice
Published June 10, 2026

As autonomous vehicles move from novelty to commercial reality, the legal and insurance industries face an emerging and complex question: when a self-driving vehicle is involved in a weather-related accident, who is liable? Gregory J. Quinn, Director of Rimkus’ Meteorology Practice and a seasoned expert witness meteorologist, examines how forensic meteorology is poised to play a central role in AV accident investigations and litigation — and why understanding weather’s role in these cases matters now, before the next generation of claims arrives.

Adverse Weather and Autonomous Vehicles

Severe weather conditions, such as heavy rain, high wind, and snow or ice, reduce visibility and traction on roadways, increasing the risk of vehicle accidents. According to a study performed by the American Automobile Association (AAA) Foundation for Traffic Safety between 2010 and 2014 (Tefft, B.C, 2016), 21% of vehicle accidents in the U.S. were caused by adverse weather conditions. Over the last decade, the Autonomous Vehicle (AV) industry has rapidly expanded due to both the growth in Artificial Intelligence (AI) and increased collaboration between auto manufacturers and software development companies. While AVs promise safer roads by eliminating human error, severe weather conditions will likely remain a critical factor in future liability claims and litigation surrounding AV-related accidents.

Commonly referred to as a self-driving or driverless vehicle, an a AV is a type of vehicle that is capable of navigating and operating without human intervention. The Society of Automotive Engineers (SAE) defines six levels (0 through 5) of vehicle automation¹. The levels range from zero, in which a driver performs all driving tasks (no automation), to level 5, in which a suite of systems (combination of sensors and software) performs all driving tasks (full automation). A level 4 AV (highly automated) is defined as a vehicle capable of performing all driving functions under certain conditions while a level 5 AV (fully automated) is defined as a vehicle also capable of performing all driving functions under all conditions, including adverse weather conditions.

Legislation for Testing Vehicle Automation

As of October 2024, 27 states and the District of Columbia have passed legislation to allow the testing or operation of level 4 and level 5 vehicles². Additionally, on April 24, 2025, the U.S. Transportation Secretary, Sean P. Duffy, announced a new National Highway Traffic Safety Administration (NHTSA) Automated Vehicle (AV) Framework³ as part of the Department of Transportation’s (DOT) agenda to promote domestic innovation of AVs. The highest commercial use of AVs currently is in the robotaxi (ridesharing) industry⁴, operating mainly in large metropolitan areas including Phoenix, Los Angeles, Atlanta, Austin, and San Francisco.

The initial ridesharing markets were located within “fair weather” cities, however, situations have occurred in Phoenix (September 2025)⁵ and San Francisco (December 2025)⁶ where robotaxi operations were suspended because AVs became trapped on flooded roadways. More recently, 3,800 robotaxis after glitch allowed some vehicles to ‘drive into standing water’ in San Antonio, Texas (May 2026)⁷.

Although the robotaxi industry will surely adapt and improve the weather-sensing technology of its fleet based on weather-related incidents, current expansion plans include not so weather-friendly locations including Washington, D.C., Detroit, and London, England. Furthermore, plans to expand AV use to commercial trucking, even limited to localized metropolitan routes, will certainly lead to an increase in AV-related accidents.

Forensic Weather Forecasting

Large-scale severe weather events are foreseeable, or generally expected by the public, such as a hurricane approaching for days across the Atlantic or a winter storm outbreak across the Midwest. Small-scale weather events, for example afternoon thunderstorms, can be sudden, isolated, and unexpected. One minute there are clear skies, the next minute torrential downpours and high wind gusts cause rapid changes in traction and visibility. Sudden weather changes can also be an issue in the winter, with parts of the Northeast and Great Lakes accustomed to sudden snow squalls and lake-effect snow bands.

In meteorology, forecasting the impacts of large-scale weather systems is far easier than forecasting the impacts of small-scale weather systems such as severe thunderstorms, that can produce isolated high wind, hail, and flooding rains. A human driver can anticipate and react to rapidly changing weather conditions. When vehicle accidents occur, the weather conditions and their impact on roadway surfaces are compared to the driver’s reaction to those conditions to determine fault (foreseeable, avoidable). What will happen when there is no physical driver involved? This is key question that experts in forensic meteorology are asking.

Liability Exposure in AV-Related Accidents

Commercial AVs use both local (onboard technology and sensors) and remote technology (operations centers) to determine safe operating conditions during adverse weather⁸.  During a traditional commercial vehicle accident, liability typically falls upon the driver, the vehicle manufacturer, such as a mechanical failure (ex. brake failure) or software failure (ex. cruise control failing to disengage), and/or fleet operations. Similar to traditional accident investigations, the vehicle manufacturer will still have liability exposure in AV-related accidents; however, if a mechanical failure is found not to be at fault, the focus will shift to potential software-related failures related to either the sensory systems of the vehicle (cameras, radar, LiDAR) or the algorithm designed to safely operate the vehicle based on the sensory systems. If the black-box information retrieved from the AV shows the vehicle sensors and software performed as designed, liability claims involving an AV during inclement weather will likely be directed toward AV fleet operations.

Unlike a human driver who can relay “why” they may have made certain decisions during adverse weather conditions, the reason “why” an AV was allowed to operate during severe weather will likely be the focus of liability when the AV manufacturer is not at fault. To further complicate the issue, many manufacturers of AVs also provide fleet control systems to aid in determining where the vehicles can and should drive in environmental conditions. Understanding the weather-related fleet operation procedures and comparing those procedures to the specific weather conditions surrounding the accident will become paramount.

Understanding the Safety Operations

For example, when the National Weather Service (NWS) issues a warning, how is this warning applied to the AV operations? Hurricane Warnings and Winter Storm Warnings cover large geographic areas, are typically issued days in advance, and have a specific defined timeline of when the storm will begin, its related hazards, and when the storm will end. These situations are drastically different when compared to Severe Thunderstorm Warnings and Tornado Warnings that cover very isolated geographic areas, are issued shortly prior to, or in some situations after, hazardous conditions occur. Hazardous roadway conditions caused by the isolated severe weather (ex. ice, hail, dust) will also become factors in AV accident investigations. The decision to suspend AV fleet operations will further be complicated by the decision to resume operations during severe weather and related hazardous roadway conditions (the stop-go-stop paradox).

How Can the AV and Meteorology Industries Work Together?

The American Meteorological Society (AMS) has identified the need for increased collaboration between meteorology and the AV industry in both the testing of AV performance in adverse weather situations and development of regulations regarding the use of AVs in adverse weather situations. In June 2024, the AMS and the Transportation Research Board (TRB), a part of the National Academies of Sciences, Engineering, and Medicine, convened the second Automated Vehicles and Meteorology Summit at the University of Michigan in Ann Arbor, Michigan. The mission of the most recent summit was to “advance the future of autonomous transportation through weather intelligence.”⁹ 

Beyond the scientific and research interest surrounding AVs and inclement weather, there is also the broader philosophical perspective of Autonomous Machines (AMs) and the challenging questions regarding the designation of responsibility. This problem, commonly known as the “responsibility gap,” will test the notion that if an autonomous machine cannot be held responsible, who then is responsible (liable) for the consequences of an autonomous machine’s actions?¹⁰

As AVs become increasingly available, the operational use of commercial AVs will increase exponentially in the coming years and decades. Likewise, as commercial AVs become more widespread geographically, the number of AV-related accidents involving severe weather will increase. How these accidents are investigated and litigated will certainly steer future regulation, legislation, and growth of the AV industry.

Forensic Meteorology Expertise for Complex AV Claims and Litigation

As commercial AV use expands into new markets and weather conditions, the need for rigorous forensic weather investigation will only grow. Rimkus’ Meteorology Practice provides attorneys, claims adjusters, and insurance professionals with the scientific foundation they need to evaluate weather-related liability in AV accident cases, from reconstructing the precise weather conditions at the time of an incident to providing expert witness testimony in federal and state proceedings.

Our experts can utilize WeatherPath™, Rimkus’ proprietary weather data application, to develop detailed visual analyses of weather events and their impact on roadway conditions to give clients the evidentiary support to build or defend complex claims with confidence.

1. Source: SAE Levels of Driving Automation™ Refined for Clarity and International Audience ↩︎
2. Source: States Steer Autonomous Vehicle Legislation ↩︎
3. Source: Trump’s Transportation Secretary Sean P. Duffy Unveils New Automated Vehicle Framework as Part of Innovation Agenda ↩︎
4. Source: Autonomous Vehicle Market Size, Share & Industry Analysis, By Level (L1, L2, & L3 and L4 & L5), By Vehicle Type (Passenger Cars and Commercial Vehicles), and Regional Forecast, 2026-2034 ↩︎
5. Source: ‘They can’t see under the water’: Waymo vehicles back on the road after many became stuck in Arizona floodwaters ↩︎
6. Source: Waymo pauses San Francisco service amid severe weather ↩︎
7. Source: Waymo recalls 3,800 robotaxis after glitch allowed some vehicles to ‘drive into standing water’ ↩︎
8. Source: Self-Driving Vehicles and Bad Weather ↩︎
9. Source: Automated Vehicles and Meteorology Summit: Recap and Next Steps ↩︎
10. Source: Responsibility before freedom: closing the responsibility gaps for autonomous machines ↩︎