Gridlock Guy: How smartphones and AI are helping make roads safer
For all the debate that the use of artificial intelligence sparks, AI’s ability to aggregate data and map trends is undeniable. How humans read and implement that data is what makes this technology effective.
Enter Cambridge Mobile Telematics and its new portal StreetVision. The Massachusetts firm gathers reams of traffic information from willing drivers’ phones, anonymizes it, and then maps out trends and stats. They get a large part of these numbers from motorists who opt in for safe-driving discounts on their auto insurance plans.
Several months ago, we discussed CMT’s tracking of distracted driving in Georgia and how the Peach State lags behind other states that have similar hands-free laws.
“The sensors on the smartphone allow us to create a picture of risk,” CMT senior vice president of strategy, Ryan McMahon, told 11Alive and the AJC. “We can look at things like near-miss crashes through braking, harsh acceleration, (and) high-G events. We can look at the inverse with hard acceleration events. You can look at speeding, you can look at speeding relative to the speed limit, to the type of the road, to the prevailing traffic.”
This is a long list of different maneuvers, but the basic point is that satellites can track any phone movement, and companies like CMT can get a very clear picture of trends based on them.
“About a trillion data points are gathered every day in these systems. It’s a ton of information,” McMahon said. “We’ve created a platform that allows road safety engineers and highway safety offices to log in to a website to identify and understand where those hot spots are and how they’re changing every day.”
One such spot in metro Atlanta that McMahon pointed out to me is a short East Point off-ramp from Highway 166/Langford Parkway EB to Main Street. Drivers plug along at freeway speeds and then have just a few seconds to decelerate to a stop sign. They also have to negotiate a curve while doing so.
StreetVision’s data map showed an unusually high amount of hard braking on that ramp, along with plenty of phone motion. AI collates that data, and experts on CMT’s team analyze it and can predict a high likelihood of crashes in that spot. “The more that roadway safety experts can identify and understand the conditions that cause crashes, the more they can implement solutions that stop those before they turn into tragedy.”
StreetVision compares these movement events to all other data points and shows a heat map. This particular ramp, which the program divides into 12-foot road sections, measures “high” right in the apex of the ramp’s curve.
The increased phone motion on the Hwy. 166 EB ramp to Main Street also shows that drivers are very likely distracted, which makes the crash risk even higher, McMahon said.
CMT said StreetVision is being used in 32 states and in 10 countries. They were demonstrating it in Washington, D.C., recently, McMahon said, when he noticed a nearby intersection that had similar data points to that Highway 166 EB off ramp. McMahon and the team drove over there and found a stop sign partially obstructed by an overgrown tree. People were jumping on the brakes harshly, very likely because they were seeing the stop sign at the last second.
McMahon noted that telematics — data collected via telecommunications — has also shown the strikingly high numbers of distracted driving nationwide. “We realized that 34% of (crashes CMT monitored), a driver’s phone is in their hand a minute prior to the impact.”
Combine erratic driver behavior with flaws in road design and driver distraction: It’s a nasty concoction.
But McMahon hopes that making better policy and design decisions can predict and then account for these bad moves. “And I think that we’re gonna see a lot more work at the base of the pyramid, so to speak, looking at the risk behaviors and then building up to solve at the top, which that last link in the chain is a crash that turns into a tragedy. And the more that it’s in the hands of people that can solve the problems, the better we all are.”
GDOT has worked with CMT but told me it uses a different company that offers similar services.
Ultimately, drivers still hold the most responsibility for safety in the commuting environment. But big data and artificial intelligence are helping decision-makers set motorists and pedestrians up for better success.
Doug Turnbull covers the traffic/transportation beat for WXIA-TV (11Alive). His reports appear on the 11Alive Morning News from 6 a.m. to 9 a.m. and on 11Alive.com. Email Doug at dturnbull@11alive.com.
About the Author
