Driver's assistance tech could increase your risk of crashing, UT researcher says
AUSTIN (KXAN) — New research out of the University of Texas at Austin is revealing that the technology meant to make driving safer could be putting you at risk.
In a new paper published in Production and Operations Management, UT Austin Professor Ashish Agarwal explored how human drivers react to alerts in their cars meant to prevent accidents.
The researchers focused on two types of warning systems in about 200,000 cars. Warning systems broke down into two categories: non-urgent and urgent warnings.
Non-urgent systems include blind spot detection sensors that alert drivers to a car in their blind spot, while urgent systems include collision prevention sensors. These sensors alert a driver that they are about to hit a vehicle.
The researchers found when non-urgent warning systems were in use, drivers had better speed control and didn't slam on their brakes as often. While urgent systems led to more speeding and more hard braking.
"If you're looking at urgent warnings, it's the opposite, which is counterintuitive, right?" Agarwal said. "Your speeding behavior gets worse, and hard braking behavior also gets worse with these warnings, and the behavior change it degrades over time."
Based on their research, the team found:
- Blind spot detection reduced the daily number of hard braking instances by 6.76% and speeding events by 9.34%, compared with cars without advanced driver's assistance systems.
- While forward collision warning systems led to 5.65% more hard braking and 5.34% more speeding.
"On average, these features help you prevent collisions, but the effectiveness (over time) increases for non-urgent warnings like blind spot detection, whereas for urgent warnings, the effectiveness reduces. The driver behavior becomes worse," Agarwal said.
According to Agarwal, the way we process this information is partially to blame. Urgent warnings occur suddenly and cause "reactive thinking," while non-urgent warnings have "a chance for reflection." This constant stimulus, like say every time a car drives by you, leads to learning that improves behavior.
Active Driving Assistance Systems
These types of driving assistance systems have been in cars for years.
"Advanced driver assistance systems could prevent about 37 million crashes, 14 million injuries and nearly 250,000 deaths over the next 30 years," said Doug Shupe, a spokesperson with AAA.
ADAS include forward collision warning detection, adaptive cruise control, lane departure warnings and blind spot monitoring.
Despite the benefits, these systems aren't without their issues. A 2020 study conducted by AAA found that over the course of 4,000 miles of driving, the systems have issues every eight miles. They also disengage with little notice.
The systems are also not perfect. A 2022 study found that at 30 mph, automated emergency braking succeeds seventeen out of twenty times. The success rate drops at speeds higher. At 40 mph, the system works six times out of twenty.
"There is no substitute for an active, focused driver out there on the roadway," Shupe said.
Agarwal's research, which was done on behalf of an unnamed large automaker, began in 2018. The team, made up of researchers from the City University of Hong Kong and the University of Maryland, looked at vehicles produced between 2018 and 2019.
Agarwal works in UT's Department of Information, Risk and Operations Management. His studies focus on the design of business technologies and how humans behave around them.
"When you're designing business technologies, you need to incorporate understanding of the user behavior to better design the products, or better design business technologies," Agarwal said.