(Press-News.org) In a significant step towards improving road safety, Johns Hopkins University researchers have developed an A.I.-based based tool that can identify the risk factors contributing to car crashes across the United States and to accurately predict future incidents.
The tool, called SafeTraffic Copilot, aims to provide experts with both crash analyses and crash predictions to reduce the rising number of fatalities and injuries that happen on U.S. roads each year.
The work, led by Johns Hopkins University researchers, is published in Nature Communications.
“Car crashes in the U.S. continue to increase, despite decades of countermeasures, and these are complex events affected by numerous variables, like weather, traffic patterns, and driver behavior,” said senior author Hao (Frank) Yang, a professor of civil and systems engineering. “With SafeTraffic Copilot, our goal is to simplify this complexity and provide infrastructure designers and policymakers with data-based insights to mitigate crashes.”
The team uses a type of AI known as Large Language Models (LLMs) which are designed to process, understand, and learn from vast amounts of data. SafeTraffic Copilot was trained using text (i.e., descriptions of road conditions), numerical values (i.e., blood alcohol levels), satellite images and on-site photography. The team’s model also has the ability to evaluate both individual and combined risk factors, offering a more detailed understanding of how these elements interact to influence crashes.
By design, SafeTraffic Copilot incorporates a continuous learning loop so that prediction performance improves as more crash-related data is entered into the model, making it even more accurate over time. Even more importantly, by using LLMs, researchers can quantify the trustworthiness of the prediction—in other words, they can say a given prediction will be 70% accurate in a real-world scenario.
“By reframing crash prediction as a reasoning task and using LLMs to integrate written and visual data, the stakeholders can move from coarse, aggregate statistics, to a fine-tuned understanding of what causes specific crashes,” Yang said.
The model gives policymakers and transportation designers a trustworthy and interpretable tool to identify combinations of factors that elevate crash risk. The data can then be used to execute evidence-based interventions and more effective infrastructure planning to save lives and reduce injuries.
The researchers see the model as a copilot for human decision-making.
“Rather than replacing humans, LLMs should serve as copilots—processing information, identifying patterns, and quantifying risks—while humans remain the final decision-makers,” Yang said.
SafeTraffic Copilot has the potential to be a blueprint for responsibly integrating AI-based models into high-stakes fields, like public health and human safety. Because LLMs operate as large black-box models, users do not know how predictions are generated, deterring their use in high-risk decision-making scenarios.
The team plans to continue their research to better understand how AI models can be used responsibly in those settings.
“The central focus of our ongoing research is to find the best way to combine the strengths of humans and LLMs so that decisions in high-stakes domains are not only data-driven, but also transparent, accountable, and aligned with societal values,” he added.
Study authors include Hongru Du, assistant professor at the University of Virginia, and Johns Hopkins doctoral candidates Yang Zhao, Pu Wang, and Yibo Zhao.
CONTACT:
Doug Donovan
Director of Media Relations
Johns Hopkins University
443-462-2947 / dougdonovan@jhu.edu
jhunews@jhu.edu
###
Johns Hopkins University news releases are available online, as is information for reporters. To arrange interviews with Johns Hopkins experts, contact a media representative. Find more Johns Hopkins experts on the Experts Hub, and more Johns Hopkins stories on the Hub
END
A new drug combination could significantly delay the progression of a life-threatening form of prostate cancer in men with specific genetic mutations, finds a major international trial led by UCL researchers.
The Phase III AMPLITUDE trial, published in Nature Medicine, tested the addition of niraparib, a type of targeted cancer drug known as a PARP inhibitor1, to the standard treatment of abiraterone acetate and prednisone (AAP).2
The study focused on patients diagnosed with advanced prostate cancer where cells have spread to other parts of the body, who were starting their first ...
LOS ANGELES — Scientists at City of Hope®, one of the largest and most advanced cancer research and treatment organizations in the U.S., and a leading research center for diabetes and other life-threatening illnesses, have uncovered a gene called SMOC1 that plays a surprising role in the development of type 2 diabetes (T2D) by converting pancreatic cells that normally produce insulin into those that increase blood sugar.
The findings, published in Nature Communications, identify an important new therapeutic target for T2D and ...
A few years ago, researchers in Michal Lipson’s lab noticed something remarkable.
They were working on a project to improve LiDAR, a technology that uses lightwaves to measure distance. The lab was designing high-power chips that could produce brighter beams of light.
“As we sent more and more power through the chip, we noticed that it was creating what we call a frequency comb,” says Andres Gil-Molina, a former postdoctoral researcher in Lipson’s lab.
A frequency comb is a special type of light that contains many colors lined up next to each other in an orderly pattern, kind of like ...
Survey of 166 international experts highlights concerns about protecting human and wildlife health from environmental pollutants
Less than a third of industry scientists support including behavioural tests in chemical safety assessments, compared to 80 per cent of academics and 91 per cent of government scientists
Despite almost all scientists (97 per cent) agreeing that chemicals can affect wildlife behaviour, most testing is done by universities rather than chemical companies, leaving gaps ...
Cells all require the transport of materials to maintain their function. In nerve cells, a tiny motor made of protein called KIF1A is responsible for that. Mutations in this protein can lead to neurological disorders, including difficulties in walking, intellectual impairment and nerve degradation. It’s known that mutations in KIF1A also result in a weakened motor performance, but this has been difficult to measure so far. Researchers including those from the University of Tokyo and the National Institute of Information and Communications Technology (NICT) in Japan have measured changes in the force of KIF1A using a nanospring, a tiny, coiled structure, ...
A unique collaboration, the Elsevier Foundation and RIKEN, Japan’s premier research institute, have released “Envisioning Futures: Women’s Leadership and Gender Equity in Japanese Research,” a comprehensive report that shines a spotlight on the journeys, achievements and challenges faced by women research leaders in Japan. The report blends data-driven analysis with compelling personal narratives, offering a timely roadmap for institutional change and a call to action for the advancement of women in science and innovation.
The report shows that despite Japan’s global reputation for technological ...
Four-limbed vertebrates, known as tetrapods, have two enlarged areas in their spinal cords. The two enlargements have a correlation with the forelimbs and hind limbs, respectively. These enlargements are thought to be caused by the complex muscular system and the rich sensory networks supplying nerves to the limbs.
Meanwhile, it was long thought that fish had no enlarged areas in their spinal cords due to the absence of limbs. However, a recent study by scientists from Nagoya University in Japan has revealed that zebrafish, in fact, have enlarged areas in their spinal cords, although these areas are not visible ...
HALIFAX, Nova Scotia, CANADA, 7 October 2025 -- In a compelling Genomic Press Interview published today in Genomic Psychiatry, Dr. Martin Alda illuminates how decoding psychiatric heterogeneity holds the key to revolutionizing mental health treatment worldwide. The interview reveals how this internationally acclaimed researcher transformed our global understanding of bipolar disorder by proving that what appears as one condition actually represents multiple genetically distinct disorders, fundamentally ...
The most common first diagnosis of Alport syndrome in Japan is during the universal age-3 urine screening. In 60% of these children, the disease had already progressed far enough to qualify for treatment. Therefore, universal early-age urinalysis may be an apt means for both better prognoses and reduced costs of medical care.
Alport syndrome is a genetic disease that affects about one in 5,000 people. Patients cannot produce a certain type of collagen which leads to kidney failure, and may also lead to hearing loss and ...
A good night’s sleep has long been understood to help us consolidate new memories, but we don’t understand how. Associations with negative feelings like fear or stress can improve recall, but intentionally trying to remember can also be effective. But these two mechanisms are very different — one involuntary, one deliberate. Which influences memory most? To investigate, researchers asked participants to remember or forget words, some of which had negative emotional associations. They found that instructions improved recall ...
