Data Sharing Maps the Route to Fewer Traffic Injuries and Fatalities

Tens of thousands of people die in motor vehicle accidents each year in the United States. Many more are injured, resulting in $55 billion in medical and work-loss costs as well as untold amounts of suffering. Since human error is a factor in over 90 percent of these collisions, the introduction of automated systems in motor vehicles has the potential to reduce those numbers.

At MITRE, we’re working to maximize the benefits of these innovations, identify any limits to their effectiveness, and mitigate any unintended downsides. Our approach taps a powerful source of insight: large quantities of diverse data.

That data is at the foundation of the Partnership for Analytics Research in Transportation Safety (PARTS). Based on a model that has dramatically improved aviation safety, PARTS is a public-private partnership through which the National Highway Traffic Safety Administration and automobile manufacturers share safety data with MITRE for analysis to advance traffic safety.

"PARTS generates safety insights that individual organizations cannot," explains Jessica Lascara, who leads the PARTS project at MITRE. "We start by pooling large quantities of disparate data from multiple partners. Then, after we analyze the data, we report our findings to the PARTS partners for action within their own organizations."

Data Sharing Prototype Produces Insights

Begun as a pilot project in 2018, PARTS initially focused on determining how effective automatic emergency braking (AEB) systems were in reducing rear-end collisions in the real world. That study—the largest of its kind—confirmed that vehicles equipped with AEB were involved in significantly fewer rear-end strikes than other vehicles.

"During the prototype, PARTS partners discovered the extraordinary potential of data sharing and analysis to provide insights," says Chris Hill, chief engineer of MITRE’s Transportation Safety Division. "Now both government and industry partners are excited to expand this work—by studying today’s advanced safety technologies and laying the groundwork for researching the highly autonomous vehicles of the future."

Taking a Deeper Look at Advanced Driver Assistance Systems

In 2020, the partnership expanded to include additional automobile manufacturers, incorporate more data, and broaden the analytical focus. For this new phase of research, manufacturers provided feature information for approximately 49 million vehicles. MITRE linked this data to eight million police crash reports.

We’re now conducting analyses to gain insight into how advanced driver assistance systems (ADAS) perform in real-world scenarios. These systems include automatic emergency braking, forward collision warning, and lane management systems.

"We’re working to answer questions such as: What is the effectiveness of ADAS features to prevent crashes, and what factors influence that effectiveness?" Lascara says.

Estimates suggest that ADAS, if widely adopted, could prevent or mitigate about 40 percent of all crashes involving passenger vehicles, including 37 percent of the injuries and 29 percent of the fatalities they cause.

"To realize these breakthrough benefits, we must adopt approaches that accelerate industry-wide safety research, so we know what advanced technologies are most effective at saving lives and reducing harm," adds Amy Aukema, who serves as the partner engagement leader on PARTS.

"Only then can stakeholders make evidence-based decisions that make travel safer."

Laying the Foundation for Automated Driving Systems

PARTS is also looking to the future. Beyond ADAS, automated vehicles—those that can navigate without human intervention—have the potential to further improve traffic safety. They can also provide people currently unable to drive a car, such as those with disabilities, the opportunity to overcome accessibility limitations.

"Our current focus on ADAS lays the foundation for automated vehicles," Aukema says. "The vision is to expand the scope of the partnership. That will eventually allow us to use real-world data to study the safety effectiveness of vehicles operating at higher levels of automation as they enter the marketplace."

"As PARTS capabilities mature and we integrate new data sets into our work, we’ll be able to answer increasingly complex and nuanced research questions," Lascara adds.

"Ultimately, we want to collect data directly from the vehicles themselves, in real time. That would enable us to provide even more authoritative and accelerated safety insights to our government and industry partners.

“Armed with the data, we hope to make driving safer for everyone.”