Editor’s note: This article originally appeared in Police1’s digital edition, “Guide to traffic enforcement.” Click here to download this free publication.
Amid the national debate on police reform, traffic stops have emerged as a focal point of scrutiny, raising pivotal questions about the future of traffic enforcement. John Leibovitz is the founder and CEO of Passage Safety based in Washington, DC. Passage Safety helps cities examine, enforce and evolve traffic behavior for safer, better and more equitable streets. Leibovitz’s innovative approach, encapsulated in his white paper “From Speed Traps to Safety Zones,” leverages network sensors and insights from behavioral science to create a more effective and equitable automated traffic enforcement system.
I recently spoke with John about this paper for an episode of the Policing Matters podcast. We discussed not only the technological capabilities of Passage Safety but also how such advancements could redefine the landscape of traffic enforcement in a society increasingly focused on police reform and community safety. This is an excerpt from the episode. Click here to listen to the complete interview.
Jim Dudley: What got you into traffic research?
John Leibovitz: What motivated me to start this was my children. I started to notice an uptick in extremely aggressive and bad traffic behaviors, especially coming out of the pandemic. I have a background in technology, business and regulation, and this area covers all three of those things. I sat down with some friends who are more technically minded, and we started thinking about traffic enforcement and specifically automated enforcement and came up with a thesis about how traffic enforcement could evolve to a new generation.
Jim Dudley: You write about the transition from speed traps to safety zones and how that shift could alleviate law enforcement officers from doing traffic enforcement. Can you elaborate?
John Leibovitz: Automated traffic enforcement such as speed cameras is not a new technology. It’s been around in some form since the late 1960s. What’s interesting about automated enforcement is that it works. Studies show that when there is a camera at a particular place, people do slow down because they learn that the camera is there, especially repeat drivers such as commuters.
The challenge is that in most jurisdictions, because of the traditionally high cost of the systems and the politics around them, you have a very limited number of sensors. What ends up happening is you have certain hotspots where there are cameras and people slow down in those locations, but then they speed up again. Behavior changes locally but not globally. That is the core issue we’re trying to address with Passage Safety.
The technology has come a long way since the first digital cameras. You can build a speed camera with off-the-shelf parts at a fraction of the cost of traditional cameras, so cameras can be a lot more widely deployed for a lot lower cost. Our thought is instead of having one expensive camera at one place where people slow down but then speed up again, why not expand that coverage to create a zone? It could be a corridor, it could be a neighborhood, it could be a set of blocks, it could be all around a school.
And that changes the way you might think about enforcement. Instead of having one big fine for blowing through a single camera, maybe you get several smaller fines. What ends up happening is people are getting more constant enforcement and potentially real-time notifications through their phone in a non-distracting way to alert them that they’re speeding. This changes the paradigm of how automated enforcement works.
From behavioral science, we know that the frequency and rapidity of feedback and enforcement does change behavior. What we want to do is try to bring in some of that literature and those findings from behavioral science. Couple that with advances in technology to make sensors a lot cheaper, and you’re generating a lot of speed data from anonymous drivers. Looking at that data, you can look at how to design programs that have the most effect in terms of making the streets safer.
We’ve been working with academics to think of it in an analytically driven data-driven way to get at the core problem, which is how do you maximize safety from these systems given a lot of the constraints that you have around politics, budget, revenue and equity, all the other things that are wrapped up in it.
The fact that we are talking about machines and not humans has a lot of implications for equity and policing in general.
Jim Dudley: When we’re talking technology seeing the violation, all the accusations of race or gender bias or anything else, economic status goes out the window, right?
John Leibovitz: When you think about equity issues, there’s a lot going on with law enforcement, traffic stops, speed enforcement, etc.
The first and most obvious thing is that a lot of people think of automated enforcement as a less biased mechanism for enforcement than human enforcement because it’s a machine, and like you said, a machine just observes a certain input and gives you an output. In a lot of jurisdictions, they have cut back on police enforcement of traffic laws and traffic stops due to equity concerns. And in a lot of places, they’re now looking to fill the gap because they’ve seen upticks in traffic safety problems. It’s a real epidemic. It’s one of the leading causes of death of people under 50 in the US. It’s a real problem. So, just taking something that’s human enforcement and making it automated has some potential equity benefits.
Then you get into the question of where these cameras are deployed. I think geography matters. In most cities, the neighborhoods are stratified in terms of income distribution and other demographic factors. And so, if you were to put all the cameras in one part of the city and not in another part, you’re going to be affecting certain populations more than others.
One thing that we say in the white paper, which I think is important to understand, is that the way cameras are deployed today there is a scarcity mindset where you have a very small pool of cameras to deploy in the city. For example, in Washington DC, there are presently about 130 speed cameras. My understanding is they’re about to double that. Let’s say they get to 300 speed cameras with the new legislation that passed this year. There are 15,000 city blocks in Washington DC. That means you’re enforcing on about 1% of city streets.
It forces you as a policymaker to make some hard choices about where you position the cameras. What neighborhoods do they go in? And that creates winners and losers on both sides – people who get more fines, but also people who get more safety.
If you have much more widespread camera coverage, it evens that out in a lot of ways. And I think that just having more enforcement in some ways can help with the risk side of the equation. And then lowering the fines and giving people notifications and opportunities to modulate their behavior, reminders, so that they in real time know that they’re getting fined and maybe they get a warning before they get fined. Maybe you say, we caught you speeding on the loss block. Here’s your second warning, slow down, or you’re going to get a warning. Maybe the fines escalates. Maybe it’s not fines, maybe it’s points, which is something that’s a real live discussion here in DC.
Those are all policy questions, but the fundamental point is that if you have more technology, more data, more insights and more behavioral science expertise, you have many, many more tools to try to address those trade-offs than you do with the traditional system of infrequent enforcement with very high fines.
Jim Dudley: You mentioned the possibility of real-time feedback to drivers through the network system. How can law enforcement agencies utilize this immediate feedback mechanism to foster safer driving habits within their communities? We have automated mapping systems like Waze and Google, and sometimes they tell us if there’s an accident in real-time. Can law enforcement officers be alerted if there is a high time for people speeding or running stoplights?
John Leibovitz: One of the interesting things about being in Washington DC is that it has some very innovative people and innovation programs. We were fortunate to participate in one of those programs earlier this year.
We deployed several of our inexpensive sensors on different streets in the district. They were there from the coldest part of the winter to the hottest part of the summer. What we found was that these sensors generated an enormous amount of data in real-time.
Between those three sensors over the course of a month, there was 350,000 rows of data, with each row representing a data point in terms of a vehicle that was traversing the street at a certain speed and some other data associated with that. Only the cars that were above a speed threshold, which in this case was 11 miles per hour over the speed limit, would actually have any vehicular identifying information associated with it.
The point is that the data that you get from that in real-time can be sliced and diced in all kinds of different ways to give insights to law enforcement, to Department of Transportation planners, engineers, people designing roads to help prioritize resources. Even without changing the way enforcement and citations happen, just having the data of how fast vehicles are going and which places are the worst is valuable.
For example, one of these locations was a school zone, where we found no one was obeying the 15-mile-an-hour limit. That forces you to ask questions like should we have the 15-mile-an-hour limit? Or if we do, is it purely symbolic or do we want to take it seriously? And if we do, do we want to put speed bumps there? Do we want to do more enforcement there? Do we want to station a police vehicle there at certain times? Those are insights that are actionable for the city. Having lots of data in real-time is very valuable.
Jim Dudley: Yeah. So empirical data means something as well, and oftentimes we see this policy pushed out and law enforcement doesn’t have a seat at the table as a stakeholder. Is there going to be room for, is there a chair set up for law enforcement to come in and say, okay, that’s a great idea.
John Leibovitz: My experience in government was that the best solutions to really hard problems happen when you have a truly multi-disciplinary, multi-agency approach where people are really collaborating and trying to problem solve together.
When it comes to specific individual design decisions about intersections and streets, I’d like to think that there would be some way of corralling all the insights and observations and ideas from different agencies recognizing that there are transportation planners and these DoTs that are churning out intersections because there’s a heavy workload. So I don’t know exactly how that works and what the best practices are. I do know that the way our system works, we can collect this data and create a dashboard and present that dashboard to all kinds of different owners within the city government, across departments, and multiple departments can have views of the same data. And then hopefully that’s a catalyst for them to talk about it and figure out what to do about problem areas.
Jim Dudley: Can you share any pilot projects or real-world implementations where you’ve outlined in your white paper and they’re being used? Is that happening?
John Leibovitz: Yeah, so I mentioned the one we did in DC, which was a fairly involved effort working with the State Department of Transportation. We learned a ton. We hope they learned some things too, but we learned a ton about what it takes to work with the city government in a big city like DC where you’ve got not just overlapping agencies, but even within a single agency, they’re complex agencies with a lot of different equities in terms of, some people were worried about safety, some people were worried about the infrastructure and how do you work with them. And so we were able to do that, that we think that was a successful pilot.
We had a previous pilot in another local area. We’ve had some discussions with other areas, cities in the Washington DC, greater DC areas, including Maryland and Virginia. And I should say there is an enormous amount of federal money out there from the IIJA, the infrastructure law that comes up once a year at different intervals for different programs that can support both experimentation and deployment of these types of technologies.
So if any of your listeners are out there and interested in exploring this, we’re very interested in expanding doing some pilots and deployment projects in other places. We can customize the technology to different needs to have different requirements to different constraints, but also potentially with an eye towards, once it’s proved out or even people are comfortable with it, there can be a leverage effect of getting access to some of these federal monies in order to really scale up and prove it out at a bigger scale. And so we obviously track those things. We have some ideas on how it slots, and we have some university partners who are very interested in working with us and helping to do the analysis and manage the program to take the burden off the city.
So anyone out there who’s interested, please reach out to us. I think you’ll link our information, but you can just send an email to info@passage.city and it’ll get to me.
