Mobility intelligence in action: Tackling data challenges in the public sector
Arity’s mobile intelligence products help Vision Zero advocates, transportation planners, and engineers reduce road risks and provide proactive and accelerated insights to support effective and time-sensitive infrastructure planning.
Bill Kotowski, Grants Officer for the Idaho Transportation Department’s Office of Highway Safety, piloted the use of Arity’s mobile intelligence products to support decision-making on some key initiatives. Arity’s Kamron Clifford, Director of Product for Mobility Intelligence, sat down with Bill to highlight how an understanding of driving behavior at scale can be a leading indicator of road safety and risk levels.
This Q&A has been edited for length and clarity.
What are some of the data challenges you encounter in the public sector?
Complicated procurement procedures
One of the big challenges is that there are so many different contracting mechanisms and requirements for competitive proposals and bids. And when you deal with some of these newer data sets, there are very few providers, so it can put up some roadblocks sometimes. Beyond that, sometimes when we talk about data, the IT perspective from a state government can add some extra bureaucratic red tape in contracting.
Government procurement as a whole can be a challenge. A lot of that’s just to make sure we’re being responsible stewards of taxpayer dollars, but we need to figure out ways to move faster to help people get to wherever they’re going as safely as they can.
How to use the data that is collected
One of the other challenges with all these new data sources and partners is how do we use what we find? How do we organize ourselves as an industry to take what we know, implement things that will make the road safer, and then continue to be better public servants?
Timely data
A lot of the data we use is from our crashes, and we use crashes to indicate risk. But it doesn’t capture risky behaviors all the time. There’s a lot of near misses that we don’t see. For example, a crash happens, a police officer takes a report and later submits that report to our office, and the timelines may vary depending on the agency. Then we have a team that will look through the report, input the data into our platform, and then we try to analyze it and collect those insights. There’s always a lag.
If there are ways to enhance the timeliness of the data, I think it is really important.
Can you share a couple of examples of how mobility intelligence data helped you to understand and improve safety on Idaho’s roads?
Distracted driving
Idaho’s hands-free driving law, Title 49, passed in 2020. Just to give a little bit of context for distracted driving in Idaho, 241 people were killed between 2014 and 2018 and it’s a factor in nearly one in five crashes. One of our concerns is, is that an underreported behavior? How do we capture it? When we look at distracted driving, when there’s a citation from a law enforcement officer, how do you prove that they were looking at their phone? If it’s someone that gets involved with a rear end crash because they were looking at their phone and not at the roadway, more often than not, they’re going to be cited for either inattention or for following too closely, failure to yield — some of those more aggressive driving behaviors. And we’re not actually getting a good picture of what’s happening in terms of distraction.
So, what we wanted to do as part of this study is compare handheld phone use events with hard braking events. When we look at the hard braking events, what we’re looking for is a sub deceleration of around 7.1 miles per hour or more within a certain timeframe from the phone’s telematics.
What you can see is after the hands-free law, there was a pretty sizable decrease in some of those hard braking events. And what we were able to do with this comparative data is look at these areas and determine, okay, are they really speeding and following too close and just rear-ending each other or are they distracted and looking up and slamming on the brakes? The good news is harsh braking doesn’t always result in a crash, and so we can do something about this and be more proactive.
We then identified several hotspots throughout the county and looked at some of the contextual elements of the roadway. Is there an exit nearby? Is there a realignment that recently happened? It was very insightful in helping us look at the roadway with a more holistic approach.
And once we compared the data, we were able to extract some insights in terms of the severity of the hard braking events. There was a pretty significant decrease in harsh braking events after the hands-free bill was passed and implemented. And the severity of the different hard braking events in the area where the hands-free law was more observed actually improved. They were less harsh on those braking events. And what that means from a road user’s perspective is when those hard braking events aren’t as hard, the survivability of a crash – if you’re walking down the street and someone hits you – it’s exponentially better. If you’re in another car and you’re hit by someone who’s able to stop and decelerate at a better rate, survivability increases significantly. And so, we were very excited to see that there was a correlation at least from this part of our dataset that hands-free and harsh braking events can both go down.
“It can take literally years to have enough of a data set from crashes alone to see a significant difference. This data from Arity allowed us to see much more real time: okay, well this roundabout was installed, these behaviors changed, and here’s how people are safer now.”
Road safety
We wanted to look at a before and after comparison of a roundabout. So, we took this intersection and looked at a timeframe before the roundabout construction began and after it was completed.
Before, there was a two-way stop for east and west traffic. North and south was free flowing. I believe it was a 45-mph road. And what you would see is there’s a little bit of hard braking, especially northbound going into this intersection. But after the roundabout was installed, the braking started and was much more gradual. It was exactly what we wanted to see out of a roundabout, people slowing down and accelerating at a more gradual rate instead of just coming to the intersection and slamming on the brakes.
We also looked at the acceleration afterwards and did a speed analysis. Before the roundabout, on the south to north side it was 45 mph, so people were going 45 or higher. So, if you’re crossing traffic, you’ve got to wait your turn and maybe try to jump out and get through. And that’s when crashes happen. After the roundabout, we saw slower, more gradual deceleration and acceleration. And I think that moved the risk of a high-speed crash away from the intersection, which was exactly what this roundabout was designed to do.
A lot of times when we look at infrastructure enhancements and we try to do a before-and-after comparison using just the crash data, it can take years. It can take literally years to have enough of a data set from crashes alone to see a significant difference. This data from Arity allowed us to see much more real time: okay, well this roundabout was installed, these behaviors changed, and here’s how people are safer now. And when we go back to talking to the public about why we install roundabouts, these are great examples to show them people are safer now.
Mobility data is a game-changer
A few years ago, we were giving a presentation to a high school about distracted driving and one of the students said, “I don’t need you to preach to me. Just give me the data, tell me what’s happening, and I can make smart choices.” And I think that’s what this is really about, right? Opening up our blind spots, because what we’re doing needs to get better.
The world has changed a lot. Whether it’s from the infrastructure or the behavioral side of safety, a lot of our countermeasures are outdated. If we want to innovate and find ways to meet today’s needs, we need to have better and more complete data.
Everything that we do as a public agency is funded with taxpayer dollars, and we can’t just throw money into a problem without having data to support it. And so, what does that data look like? How do we find those gaps and how do we bridge them?
I’ve been really excited about working with your team because it opens up some of those opportunities. If we know these things and we have a data source that’s valid, we can justify the planning and programming that we want to create.