{"version":1,"type":"rich","provider_name":"Libsyn","provider_url":"https:\/\/www.libsyn.com","height":90,"width":600,"title":"StreetSmart Episode 13: Marty Beard, CEO of Hayden AI","description":"On this episode of StreetSmart, Streetsblog California editor Damien Newton talks with Marty Beard, CEO of Hayden.AI, about the rapid expansion of automated camera enforcement on buses and city vehicles across California and beyond. Hayden.AI\u2019s technology uses forward-facing cameras mounted on transit buses and parking enforcement vehicles to identify cars blocking bus lanes, bus stops, and bike lanes. Beard explains how the data has repeatedly surprised cities, revealing widespread violations \u2014 and how enforcement has led to faster bus speeds, fewer collisions, and more reliable transit service. The conversation also explores how camera-based enforcement fits into post-2020 efforts to reduce police traffic stops, addresses common concerns about data privacy and surveillance, and examines why cities see these programs as performance tools rather than revenue generators. Beard also previews emerging uses for the technology, including identifying unpermitted construction that disrupts transit operations. During the podcast Damien references coverage of Hayden AI at Streetsblog and Santa Monica Next several times. Since Next syndicates Streetsblog's coverage of related issues, you can see all of both publications coverage at Santa Monica Next's page for Hayden AI. A transcript of this podcast can be found below. It has been lightly edited for readibility and clarity. Damien Newton So we\u2019re recording this podcast remotely on Zencastr. I\u2019m Damien Newton, and I\u2019m joined today by Marty Beard, CEO of Hayden.AI. Thanks so much for being here. Marty Beard Thank you very much for having me. Damien Newton I\u2019ll be honest with listeners: in the pre-show I told Marty that Hayden has a lot going on in California right now. Rather than firing off a bunch of narrow questions, I figured it made more sense to let him lay it all out. We\u2019ve covered some of this work in San Francisco, Los Angeles, and Santa Monica on our local Streetsblog sites, but not as much yet on Streetsblog California. So if you\u2019re not following those city sites, you might not have the full picture. Marty, why don\u2019t you start by giving us a short overview of what Hayden.AI is and what you\u2019re working on right now? Marty Beard That sounds great. At our core, we\u2019re a technology company \u2014 you could call us an AI company, and that\u2019s true \u2014 but more importantly, we\u2019re a public transit company. Everything we do is focused 100 percent on improving public transit. The way we do that is by installing cameras on transit buses, parking enforcement vehicles, and similar fleets, and pairing that hardware with our software. The goal is simple: keep bus lanes clear, keep bike lanes clear, and allow public transit to do what it\u2019s supposed to do. That\u2019s our entire mission. We don\u2019t do anything outside of that. We operate across the U.S., internationally, and of course here in California. In California specifically, we work with LA Metro, AC Transit, and cities like Sacramento, Culver City, and Santa Monica. While the locations vary, the common thread is always the same: how can technology help improve bus speed, reduce collisions, and ensure bike lanes are usable? Damien Newton Most of the coverage we\u2019ve done has focused on your cameras being installed on buses \u2014 and now sometimes on parking enforcement vehicles \u2014 to help cities enforce bus lane and bike lane laws without relying on traditional police traffic stops. Is that a fair way to describe it? Marty Beard Yeah, exactly. Agencies bring us in to do just that. The cameras are installed inside the vehicle \u2014 usually a bus \u2014 and they\u2019re designed to do one thing only: look ahead at bus lanes, bus stops, and bike lanes, and identify vehicles that are illegally parked or blocking access. The system does not identify people. It doesn\u2019t analyze broader traffic patterns. It\u2019s optimized for a very narrow task: identifying a vehicle obstructing transit infrastructure. When a violation is detected, an image or short clip of the vehicle is captured. That information is then reviewed by the appropriate enforcement agency, which makes the final decision about whether a citation is issued. The benefit is that it\u2019s extremely efficient, very accurate, and \u2014 most importantly \u2014 it works. Damien Newton I covered the Santa Monica pilot when that report came out, and we used the phrase \u201can epidemic of scofflaws\u201d in the headline because the numbers were pretty staggering. This was along just seven bus routes, over a short pilot period, and the number of vehicles blocking bus and bike lanes was astronomical. Are you seeing similar results elsewhere \u2014 that moment of \u201cwow, this is happening all the time\u201d? Marty Beard One hundred percent. There are two things that happen almost universally. First, agencies are surprised by the sheer volume of violations. They know it\u2019s a problem, but once they start seeing daily, route-by-route data, the scale becomes undeniable. The second thing is what happens after the data starts coming in over time. Agencies can look at trends and ask: are we changing behavior? And the answer is yes. Bus speeds improve, collisions go down, on-time performance gets better. In some cases, the improvements are dramatic \u2014 we\u2019ve seen 20 percent or more increases in bus speeds on certain routes. That network effect is huge. So first it\u2019s \u201cwow, this problem is worse than we thought,\u201d and then it\u2019s \u201cwow, this is actually working.\u201d Damien Newton I imagine timing plays a role here too. After 2020 and the George Floyd protests, there was a push to reduce police interactions for minor infractions, including traffic enforcement. A lot of these so-called nuisance laws just weren\u2019t being enforced anymore. So now you have a way to enforce them without those interactions \u2014 and maybe also correct some bad habits people picked up along the way. Does that sound right? Marty Beard I think you nailed it. It\u2019s also safer for enforcement staff. Parking enforcement is a tough job \u2014 you\u2019re not exactly the most popular person in the neighborhood. Technology helps because it\u2019s consistent and focused. There\u2019s this perception sometimes that cameras are spying on everything, but that\u2019s really not what this is. The camera is optimized for one specific task: is a vehicle where it shouldn\u2019t be? If there\u2019s a legitimate reason for that vehicle to be there, the citation won\u2019t be enforced. But if someone blocks a bus lane to grab a latte and 45 people can\u2019t board the bus, that\u2019s a real problem. This helps address that. Damien Newton Last week, our Streetsblog Los Angeles editor noticed something interesting during a SCAG presentation. LA Metro quietly announced plans to expand its AI camera program from 100 cameras to 400. No details beyond a slide. Can you tell us anything about that, or do I need to bug Metro\u2019s PR team? Marty Beard You\u2019ll need to ask LA Metro directly. What I can say is that we love working with them, and the results speak for themselves. But it\u2019s best for them to talk about their plans. Damien Newton They were one of your first major transit agency partners, right? Marty Beard Yes, absolutely. Along with places like New York, Chicago, and Washington, D.C., LA Metro has been an anchor customer for us. Damien Newton I just want to note for listeners: those are also all cities with Streetsblog sites \u2014 purely coincidental, I\u2019m sure. Are you in Boston too? Marty Beard I can\u2019t comment on that one. But yes, we do follow Streetsblog very closely \u2014 clearly our expansion strategy. Damien Newton We\u2019re eyeing San Diego for a Streetsblog site in the next year or two, so keep that in mind. Marty Beard That\u2019s actually where I live. Damien Newton Well, there you go \u2014 I had no idea. Marty Beard The company\u2019s headquartered in the Bay Area, but we\u2019re spread across California and the East Coast. Damien Newton Welcome to 2026 \u2014 we don\u2019t all have to be in the same city anymore. Are there any other expansions or developments you can talk about? Marty Beard What I can say is that we\u2019ve passed 2,100 installations, and every market we\u2019re in is expanding. We\u2019re also seeing growing interest beyond bus lanes \u2014 particularly bike lanes and parking enforcement vehicles, like in Santa Monica. And we\u2019re starting to look at new use cases: where else can this kind of focused, privacy-respecting technology help public transit? Damien Newton Cities do generate some revenue from this, but as I understand it, that\u2019s not the primary goal. The goal is improving bus speed and bike lane reliability. I\u2019d guess transit riders and cyclists are overwhelmingly supportive, while drivers are more skeptical. Marty Beard Our biggest supporters are transit riders and cyclists, by far. What surprised me when I entered this space is how little agencies focus on revenue. What they care about is performance: speed, reliability, safety. As a vendor, that means you have to prove it works \u2014 and show the data. Damien Newton I don\u2019t bike as much anymore, but just walking or running along corridors in Santa Monica where the cameras are installed, things feel noticeably calmer. Less honking, fewer blocked lanes. It\u2019s tangible. Marty Beard We see that reflected in the data as well. Damien Newton One concern that always comes up with camera technology is data privacy \u2014 especially with fears about data being shared beyond its original purpose. So what happens to the data you collect? Who owns it? Who can access it? Marty Beard It\u2019s a completely valid concern. Hayden does not own the data. The transit agencies do. We only collect the violation itself \u2014 typically a short video clip or still image of a vehicle obstructing a bus lane, bus stop, or bike lane. No facial recognition. No human identification. Nothing beyond that. The data is captured on the vehicle and sent to the enforcement agency, which makes the final decision. We don\u2019t issue tickets. If a government agency asked us for broad location data, we wouldn\u2019t even have it. Damien Newton Before we wrap up \u2014 is there anything we didn\u2019t cover that you want to mention? Marty Beard One emerging area is road construction. Cities want to know: is construction permitted? Is it happening where it\u2019s supposed to? Is it unexpectedly blocking transit? Our cameras are starting to help identify unpermitted or unplanned obstructions so agencies can respond more quickly. Damien Newton So the cameras are catching unpermitted construction? Marty Beard Exactly. It\u2019s a newer area, but one that could really benefit transit riders. Damien Newton That\u2019s fascinating \u2014 not something I\u2019d even thought of. Marty, thanks so much for your time. Next time I\u2019m in San Diego, I\u2019ll reach out and we\u2019ll grab coffee. Marty Beard Sounds great. Thanks again. ","author_name":"StreetSmart","author_url":"https:\/\/sites.libsyn.com\/StreetSmart","html":"