The Federal Motor Carrier Safety Administration (FMCSA) is exploring how artificial intelligence (AI) can help process Police Accident Reports (PARs) faster and more efficiently. In December 2024, FMCSA released a report detailing its research into AI-driven tools that could speed up the review of crash reports and improve accuracy.
This effort is part of the agency’s Crash Preventability Demonstration Program, which allows trucking companies and drivers to challenge certain types of crashes they believe were unavoidable.
The challenge of reviewing accident reports
Under the current system, trucking companies, drivers, and industry partners can submit crash review requests through FMCSA’s DataQs system. If a crash is determined to be unavoidable, it no longer negatively affects a carrier’s safety record. However, this process depends heavily on reviewing accident reports, which can vary widely in format, terminology, and detail depending on the state or jurisdiction where they were issued.
Since the program launched, nearly 70,000 crash review requests have been submitted, and recent expansions have increased the number of eligible crash types. As a result, processing times have slowed, meaning that non-preventable crashes continue to impact the safety scores of some fleets for longer than necessary. Reviewing these reports takes significant time and effort, delaying the process and making it harder to keep up with demand.
How AI can improve the process
To address these challenges, FMCSA partnered with the Virginia Tech Transportation Institute (VTTI) to study how AI could help speed up the review of accident reports. Researchers evaluated whether AI-powered tools could automate parts of the review process, reducing the time needed to analyze crash details and determine preventability.
What the study found: Initial results are promising. The results showed that AI could cut the time analysts spend on reviews by about 50%. With further refinements, that number could reach 65% to 70%. By using machine learning to quickly analyze large amounts of data, AI can help identify patterns and make preliminary assessments. This means faster results, more consistency, and less chance for human error.
What this means for trucking companies
Faster accident report processing means trucking companies could see non-preventable crashes removed from their records more quickly, improving their Compliance, Safety, and Accountability (CSA) scores. Lower CSA scores can lead to fewer inspections and lower insurance costs, which directly impacts a company’s bottom line.
Beyond crash reviews, AI could also assist with other regulatory processes. One potential application is analyzing public comments on new regulations, which could help make the rulemaking process more efficient. AI could also be used to speed up the publication of crash data, which currently takes years to finalize.
The future of AI in fleet safety
As FMCSA continues to refine these AI tools, their impact on fleet safety and regulatory processes will grow. More accurate and timely reviews mean resources can be directed where they’re needed most, helping law enforcement focus on identifying fleets with real safety risks. For trucking companies, these advancements could mean a fairer, more efficient system that better reflects their actual safety performance.
AI isn’t replacing human judgment, but it is making the process of reviewing accident reports faster and more reliable. As AI capabilities continue to advance, fleet managers can expect more transparency and efficiency in how safety data is analyzed and used.
For more industry insights from David O’Neal, check out the Motive blog.
