Consistently ranked as one of the worst cities for traffic in the United States, Los Angeles’ roadway congestion has continued to climb back to pre-pandemic levels. To combat inefficiencies that cause increased levels of traffic, cloud-based software platform Lyt has employed artificial intelligence and machine learning technologies to help improve the flow of vehicles.
“There are so many different flavors of AI and machine learning, and this is the future,” said Bobby Lee, director of marketing at Lyt. “What is really exciting about what we’re doing here at Lyt is that we’re getting at the central nerve of what’s going to improve mobility, reduce congestion and get people where they want to go.”
To achieve their mission, Lyt uses AI and machine learning to optimize traffic light signals. The technology works by installing a single edge device into a city’s Traffic Management Center that allows vehicles to communicate to every networked traffic signal on the Lyt platform.
For example, an emergency vehicle that needs to arrive at a destination across its city would be able to have a clear route of travel as the city’s cloud network of traffic signals could clear a path. Lee explained that the technology for emergency vehicle preemption and transit signal priority originated in the 1950s as a hardware-based solution, but that speed limits and commuting patterns have changed in the past 70 years.
“We’re taking a software, cloud-based approach to solving this issue,” Lee said. “Instead of putting in all these pieces of hardware, hoping we can close up the intersection, waiting for that light to go green…we’re saying, ‘Wait a second, let’s step back here. Let’s take a 10,000-foot view of the city. Let’s look at that entire corridor. Let’s look at all the things that are happening in context, and let’s apply machine learning to what’s happening.’
“With emergency vehicles, your fire station could be across town. It could be four or five miles to get to a heart attack, or to get to a three-alarm fire. We want to make sure that those vehicles can get there straight away, and that means clearing the entire path of travel. … We are the cloud system, the software, the machine learning that sits over the city and sees what’s going on, and we can provide the information back to the city and their signals so they can make the right decisions on how to optimize the traffic lights.”
In Sacramento’s Rancho Cordova suburb, Lyt deployed their emergency vehicle preemption solution to address the traffic challenges posed by the community’s location above the gore point of the I-80 and Highway 50, which comes out of Lake Tahoe. Traffic in Rancho Cordova had gotten so bad that engineers from the city had to manually flip traffic lights to green over the course of an afternoon to help emergency vehicles pass through the roadway.
After deployment, Lyt’s system was able to increase the speed of firetrucks by 69% and response times were reduced an average of 42 seconds.
“With cardiac incidents like a heart attack, every minute reduces your survival chances by 10%, so 42 seconds is significant,” Lee said. “We were able to get started in just a couple of weeks, and all we needed was the green light to go ahead from within the city. So for this software, the advantage of it is that once the city signs off on it, we’re in there really, really quickly, and we can provide those immediate benefits to the community. And we have heard it anecdotally that residents have called in, when they were testing having the system on and off, and they’ve noticed the difference.”
Lee said that the same model works for transit vehicles, such as buses, by studying ridership patterns and routes. A signal could be sent from a bus to inform the nearest traffic light on the route of its presence.
“(Bus riders) want reliability,” Lee explained. “They want frequency, so good service every 10 minutes, every 15 minutes. And they want to know if it’s going to get them to where they want to go when they need to get there. With a system like this, a solution like this, it’s going to really help the agency sell that back to the riders.”
Lee raised the example of Lyt’s work in a “mixed-use, transit-dependent” neighborhood in San Jose, where the community relies heavily on travel down Route 77, a north-south corridor. Lyt’s team set up their AI-powered priority signals at 17 intersections and built a web portal that showed the real-team location and activity of Santa Clara Valley Transportation Authority buses, including their routes, speeds and upcoming stops.
After the installation, residents’ travel time decreased by 20% and the Santa Clara Valley Transportation Authority’s diesel usage was reduced by 14%.
“You’ve got a win-win here, which is shorter travel times when getting folks between work, home and school, the transit agency saving on diesel and a community breathing in less of that harmful particulate,” Lee said. “We also deployed one of our big deployments for transit, actually, up in Portland. We’re on a new express route…I believe it’s about 60 intersections. They go from Downtown Portland all the way up in the suburbs, the eastern suburbs. (It’s) about 10,000 riders a day. … From us being on the ground, and from our partners, our agency partners, we have seen a dramatic improvement in speeds and reliability and schedule adherence up there.”
Lee insisted that Lyt’s model would succeed in Downtown LA, as dense urban centers contain a large ridership base and require a high frequency of connectivity between riders and their destinations. By encouraging DTLA residents to travel via public transport, Lyt could reduce traffic, as less vehicles would be out on the roads.
“If you’re able to demonstrate to those riders that each and every one of those lines going through Downtown, through the connected corridors, that they’re receiving these priority green lights and they’re receiving great travel times and great speeds, and getting them into their home, work or school as quickly as possible, they are going to want to mode shift, which is switch out of their cars or switch out of other modes of transportation and into the buses.
“Every transit agency across the country has experienced a decline in ridership, and we have seen that opposite effect in private vehicle usage on the roadway, traffic going back to 95 to 100% of what it was pre-COVID. So the only way we’re going to get ourselves out of this congestion situation is to provide an attractive alternative, and that’s to encourage people to get back onto transit.”
According to reports, the average U.S. driver spent over 50 hours, nearly an hour per week, in traffic in 2022. Though Lyt’s solutions are currently focusing on emergency and public transit vehicles, Lee said that the technology is applicable across all vehicle classes.
“It is a really fascinating approach that this industry, who’s used to expensive hardware…they have a different approach,” he explained. “We think that a lot of communities here in California and beyond can really be helped by Lyt and our solutions.”
The core of Lyt’s mission is to deliver quicker, more reliable commutes born from more efficient and environmentally friendly travel methods without the need to build infrastructure or install hardware replacements. Lee explained that the traffic light is the “only controlled device on the roadway,” which is why addressing its signal is the best place to start in finding solutions to congestion that will impact the entire mobility network of a city.
“This is technology that, of course, consumers don’t buy into directly, but that every community ought to be looking at,” Lee described. “The power of cloud computing, the power of AI and machine learning can be applied in a lot of different ways, and this is just one solid example of how it can really change the way that we get around.
“We think that this is what is going to really move the needle for the next 10, 20-plus years in helping us to get around. … Really, the sky’s the limit.”