High-risk drivers can now be rapidly identified while they're still endangering the rest of us on the road, thanks to new technology that predicts aggressive driving based on the risk level of drivers and keeps passengers safer from the dangerous effects of road rage.
A team of civil engineers at the University of Michigan created the system in partnership with Denso, an automotive parts manufacturer partly owned by Toyota. The technology is meant to monitor drivers and prevent car crashes due to aggressive driving, which is a traffic offense in many parts of the U.S. Aggressive driving, dangerous driving or other forms of unsafe driving are likewise illegal in the United Kingdom and many other countries around the world. The patent application was published by the U.S. Patent and Trademark Office on April 29.
While methods to spot aggressive drivers already exist, none have been able to predict behavior in real time. This instant, reliable feedback is key to a real-world application of the technology, as it can possibly prevent accidents and deaths due to risky drivers. The American Safety Council reports that 66% of traffic fatalities are caused by aggressive driving, and a striking 37% of aggressive driving incidents involve a firearm.
After a series of shootings during the late 1980s in the traffic-filled city of Los Angeles, the term road rage was coined to describe the angry behavior of drivers. Notably, road rage can lead to criminal charges. In the United States, the National Highway Traffic Safety Administration distinguishes it from aggressive driving, though both behaviors can end in fines and penalties for the individual responsible. As of 2017, 15 states have passed laws that target aggressive drivers.
"High-risk driving is not only detrimental to people's safety, but also costs billions of dollars in medical care and productivity loss of those involved in motor vehicle crashes. Even when aggressive drivers do not cause crashes, they make traffic flows unstable," explained Neda Masoud in an interview with The Academic Times. Masoud is a co-inventor of the innovation and an assistant professor at the University of Michigan.
For Masoud, a big part of the appeal in designing new and improved systems for the road is their potential to improve the economy. "A well-designed and -operated transportation system is central to economic prosperity in any society. The relationship between transportation systems and social mobility has been extensively documented," she stated.
Vehicles with automated and safer driving systems may benefit the natural environment, too, in a global economy that often leans on gas-powered vehicles to move people and goods. "Under the correct policies, these advancements can also reduce congestion and its externalities, including greenhouse gas emissions," Masoud said. Aggressive drivers, she explained, upset the stable flow of traffic, creating traffic jams that can increase fuel consumption — indeed, stop-and-go traffic decreases a car's gas mileage by 10% to 40%.
The team's invention relies on devices placed within the car and on traffic infrastructure such as streetlights to gather data on the driving patterns of an individual. The new technology can analyze data from these devices to predict the trajectory of a car, taking into account dynamic factors and patterns of driving that other high-risk models often miss. These highly personalized patterns are the key to spotting unsafe drivers, Masoud noted.
"Identifying aggressive driving cannot take place using simple rules, such as over-speeding," said Masoud. Over-speeding is a hallmark of aggressive drivers and can range from simply exceeding the speed limit to racing cars at 30 miles or more over the limit. "Rather, risky driving should be identified in relation to the rest of the traffic: when a vehicle engages in a behavior that is not expected by its peers, it will expose its own occupants as well as its surrounding traffic to safety risks."
The engineers' system takes spatial and temporal data from the sensors to compute the trajectory of the vehicle. The system then uses the car's trajectory to create a dynamic time warping-based score, which, in this innovation, compares objects moving at two different speeds based on the sound waves they produce. This driver score is compared to the scores of other drivers to find any anomalies, which could indicate the driver is making unexpected moves or turns and is therefore riskier than others on the road.
Masoud and her co-inventors have a clear view of how their technology can be implemented by original equipment manufacturers, such as Motorcraft, a company that makes parts for new Ford vehicles. Vehicles equipped with the team's sensors can assess the level of aggressive driving in the car and in the car's immediate surroundings. They can also issue alerts to keep drivers and passengers safer on the road.
Transportation agencies could place the team's devices on roadside units to better serve drivers and improve infrastructure, such as in a cabinet that controls traffic lights at an intersection. "For example, if the model shows that the aggressiveness score of a high portion of drivers increases as they approach an intersection, it is likely that the network design or the traffic light phasing and timing at that intersection needs to be improved," said Masoud.
One possible concern is cities or municipalities abusing the technology, just as some currently do with red-light cameras. For example, Los Angeles discontinued its red-light camera program in 2011, with a local official citing an 80% increase in rear-end collisions after red-light cameras were installed. However, some jurisdictions may not want to reverse such a lucrative law; near the heart of Los Angeles, the city of Beverly Hills continued to issue about 1,200 tickets a month that raked in $1.6 million a year as of 2014.
Masoud acknowledges the potential for false alarms with the team's invention, though she thinks security trumps a few incorrect readings. "False alarms are a small price to pay for the enhanced level of safety that the technology can provide. That said, there should always be an opportunity for the wrongly accused to defend themselves against any false claims of aggressive driving," she said.
The Michigan engineers will continue to work on their technology to incorporate more dynamic features into the system. They hope to provide a safer driving environment for everyone on the road.
"Identifying and mitigating aggressive driving is important to both save lives and direct and indirect costs incurred by individual travelers," said Masoud.
The application for the patent, "System for predicting aggressive driving," was filed Aug. 28, 2020 with the U.S. Patent and Trademark Office. It was published April 29, 2021 with the application number 17/006470. The inventors of the pending patent are Ethan Zhang, Neda Masoud, Wei Zhang and Rajesh Kumar Malhan, University of Michigan.
Parola Analytics provided technical research for this story.