A poster in the elevator at the Las Vegas Traffic Management Center warns Nevada Highway Patrol (NHP) officers to drive safely as they respond to calls, but a new solution — in pilot since September — is helping them to not only arrive faster but clear accidents more quickly as well.
Stood up in 2005 and 2006, the center already displayed more than a dozen live feeds of Freeway and Arterial System of Transportation (FAST) footage from more than 700 cameras across southern Nevada.
But dispatchers and 911 callers were once its key sources of outside information. And troopers were anxious to find a way to cut response times and clear accidents faster, knowing that the probability of secondary collisions rise more than 2.5 percent for every minute a travel lane is blocked.
“You can sit and stare at these cameras for so long. You know you’re missing something,” said David D. Crisler, Regional Transportation Commission of Southern Nevada (RTC) traffic operations supervisor at the center.
In its first U.S. pilot, Israeli startup Waycare is training artificial intelligence on the region’s traffic woes, closely scrutinizing two corridors in hopes of helping highway patrol officers work smarter.
Coordinating with NHP, the Nevada Department of Transportation and RTC, Waycare has deployed a cloud-based platform aimed at optimizing traffic management systems and emergency response, CEO Noam Maital told members of a tour group on Jan. 10.
Many in the tour, which included at least one WayCare investor and representatives of Geotab and Volvo, were in Las Vegas for the annual Consumer Electronics Show (CES).
The Waycare platform, which draws on everything from social media to crowdsourcing apps like Waze, refines and synthesizes information including speed, braking and acceleration data, then puts it to work predicting potential highway trouble spots.
“When you’re identifying hard brakes or acceleration within the vehicle, that’s stuff that’s used today for insurance. It’s used today for fleet management. But it isn’t used by cities to identify accidents or near accidents or congestion events or any event that happens, a spill on the road,” Maital told the group.
Combining data sets, city infrastructure and ecosystems — everything from weather patterns to road construction and special events — lets the platform “turn decisions into a proactive method with predictive analytics,” the CEO said.
“What we’re doing is we’re identifying the incidents by looking at past behavior of that road, identifying abnormalities that are happening right now on that road in real time and alerting the operators,” Maital added.
As traffic engineering technicians work, the platform suggests potential problem areas via pop-up windows, based on driver behavior mined from all that data. Center command staff has always known their officers’ approximate locations, but through new mobile data terminals in cruisers, they can now see their exact locations in real-time and redeploy them more quickly.
Knowing where incidents are, or could, develop lets technicians use those hundreds of cameras more effectively, zooming out as far as a mile — then sending confirmed incident reports with geotags and 30-second video clips to officers en route.
“We were able to start discerning right away which incidents were going to have an impact on traffic,” Theresa Gaisser, an RTC traffic engineering technician, said in an interview. “We look at it and use that, keep it in the forefront of our minds.”
"Now, the troopers don’t have to ask dispatch [where to go], because they can just pull it up,” said Nevada Department of Public Safety (DPS) Lt. Col. Dan Solow.
Since the system went live around late-September, officials have charted a 12 percent improvement in NHP response times, and a 23 percent drop in secondary collisions — often more serious than primary collisions — all because accidents are being cleared faster.
Brian Hoeft, RTC director of traffic management, called the achievement “so significant,” and said being a newer, collaborative city has made Vegas “ripe” for the pilot.
NHP Capt. Tom Ely said the improvements are “pretty slick stuff,” and praised the platform’s predictive nature for allowing deployment of officers to areas where incidents haven’t happened yet, but may soon. “As a state agency, we don’t always get access to the latest and greatest technology, and this is the latest and greatest,” Ely said.
See the big picture of how government agencies are utilizing artificial intelligence by exploring our Government Technology editorial database geographically visualized by location and date.