Historic, Dense, and Getting Busier: How New York Is Rethinking Utility Damage Prevention

New York's underground holds some of the oldest, densest infrastructure in the country — and it's about to get a lot more crowded. With $1.95 trillion in infrastructure spending set to hit the ground in the US between 2026 and 2030, the strain on an already-overloaded damage-prevention system is only going to grow.

So on Wednesday, June 24, we brought the people who work in that system into one room in Brooklyn for the second edition of our Build Better Together Roadshow. Utility owners, transportation agencies, engineering firms, contractors, utility coordinators, and damage prevention professionals — in person, working through how we Build Better Together by aligning people, data, and technology to get the job done. The conversations were the kind you rarely hear from an industry conference stage: honest about what's not working, direct about what it costs, and focused on what to do next.

Here's what happened in that room.

Rethinking Damage Prevention: The Challenge of Ought vs. Is 

Sarah Magruder Lyle, president and CEO of the Common Ground Alliance, led the opening keynote by asking the room to take off their company hats and put on a damage prevention hat, and to trade the word “responsibility” for “accountability.” For the past 26 years, the CGA has worked to protect underground infrastructure and to help everyone in the industry understand they have a role to play in preventing damages, and pretending otherwise is how damage keeps climbing. 

And they are climbing: damages rose from 189,052 in 2024 to 213,985 in 2025. That’s despite CGA’s 50 in 5 challenge, setting an ambitious bar for lowering the damage rate. But the reason those damages continue to rise isn’t a mystery, she noted. We’re still running a system built for a world of a million calls a year, and it’s straining under a construction boom on top of congested, unmapped, aging ground. 

This is clear when looking at what’s actually causing the damage — it’s not gas hitting gas, but everything happening around the gas line, like construction, DOT work, fiber, and water. Telecom crews are increasingly cutting telecom, with those damages rising from 26% to 30%. But pipelines are the only facilities subject to federal oversight, so when something goes wrong, the gas utility pays the fine. At the same time, fiber, water, and power rarely face the same consequences. 

The stakes of this high and rising damage rate aren’t abstract. Magruder Lyle walked through the list of catastrophes that cost the lives of people and the well-being of communities: an excavator outside Dallas Love and Dallas FTW airports that hit Frontier fiber lines that support the FAA’s air traffic control radars, knocking out computers, phones, and local radar with planes on approach and stranding 100,000 passengers. A Missouri home explosion killed a five-year-old. A strike that killed a farmer and his son on an Illinois high-pressure line, where the farmer had called 811 at least ten times that year but skipped it that time because he was sure he knew where it was. 

So why do these strikes keep happening when people know what they ought to do? It’s not a new process to call 811, wait the required time, confirm the marks, and dig with care. But it’s not always followed. Magruder Lyle urged the room to stop asking why a strike happened and start asking how the process failed the people working inside it — and to redesign it so the right thing to do is the easiest thing to do. 

Her prescription was self-regulation on the industry’s own terms — standardization across states instead of the status quo, where every state is a snowflake; exemption reform; and balanced enforcement — before those terms are written by someone else. The technology to do it, she noted, already exists. Adopting it costs less than a dig-in.  

The Future of 811 — Smarter Alignment, Fewer Damages, Safer New York 

Moderated by Bob Terjesen of National Grid, this panel picked up where the keynote left off: if the old scorecard was damages per 1,000, what should we measure now? 

Massoud Tahamtani, who spent 45 years with Virginia and PHMSA, made the case for leading indicators over lagging ones — Virginia drove damages down from 5 to .98 per 1,000 tickets in part by building the country’s first positive response system. Pete Corredor of Con Edison tracks raw incident counts, not just rates, to see where to direct resources. Andrew Melnyck of NYC DDC measures prevention in the design phase, catching conflicts before a shovel moves. And Kevin Hopper of UDIG New York named an interesting data gap: 75% of New York’s DIRT damages are reported by gas operators alone. You can’t measure the system when only one player is in the sandbox and new broadband is being installed with little to no mapping standards. 

The root causes, Terjesen noted, never change: on Long Island, close to 40% of damages come from excavators who never called 811. It’s not an awareness problem, as the offenders know and have previously used 811. But when the system is slow or unreliable, economics and efficiency take precedence, and people weigh the risk and often decide to dig. 

The panel’s answers pointed in one direction: make the one-call center the hub. Hopper described a future where API connections let the center coordinate everyone’s data without forcing anyone to hand it over. Roger Sampson of New York 811 pushed for the enforcement and legislative change the state badly needs — New York’s law hasn’t been meaningfully updated in 32 years, although UDIG and New York 811 have now submitted new language. And Tahamtani reminded the room that Virginia’s law became the national model because all the stakeholders sat in a room and rewrote it together. 

Driving Innovation — Optimizing Infrastructure Delivery Through Partnerships 

Brian Barkwill of National Grid’s gas innovation team framed the stakes plainly: when you’re asked to go faster, cheaper, and safer with fewer resources, innovation isn’t optional. 

Alex Miguel of Bancker Construction, whose crews install gas, water, electric, and telecom, made the case for coordinated digs: instead of each utility tearing up the same street on its own schedule, one project lays several systems at once, with the crews agreeing up front on who owns which lane and depth. That saves money while preventing strikes. 

Kenneth Murray of The Hallen Construction Co argued that pilots succeed when they’re premised on two key factors: operational learning involving the people closest to the work, and a real measurement system. When Hallen piloted speed bumps to protect crews, the team didn’t wait for a drop in struck workers to call it a win — that’s rare enough that measuring a decrease takes too long. Instead, they measured whether vehicles slowed down, a key safety indicator, so that they could judge their success quickly and effectively. 

Joe Forline, drawing on decades at PG&E and PSE&G, put it bluntly: tie the damage rate to executive pay and the whole company aligns around it. Roderick Fiske of Bond also offered the filter he runs every potential platform and innovation through, echoing what Magruder Lyle said in the keynote: it has to be easy, it has to be cost-effective, and it has to actually solve a problem. “It’s shiny. It’s nice,” he said of anything that misses. “But is it actually going to check off the other three things that I need it to be?” 

A recurring theme in the panel discussion was the damage done by the silo — not between safety and operations, Murray said, but between the field and management. The fix is for leaders to show up regularly where the work happens. 

Miguel put the frontline case bluntly. “We want them to be live gas professionals,” he said. “It’s what they do best.” Pile on tasks that involve simply checking a box with no real purpose or need for a human, and you pull a crew off that work they do best — so to him, the highest use of AI is automating the paperwork that has to happen and keeping those field crews in the field. 

Fiske noted that design-build jobs are where 4M's utility intelligence really shines — quickly surfacing utility information his teams wouldn't otherwise have.
Then an excavator in the audience described a Long Island municipality that requires 4M on every job. The room heard the cleanest argument for adoption there is: customers requiring it.

Harnessing AI — Navigating the Utility Transformation 

The day’s final panel, moderated by Kyler Van Gulden of National Grid, started with two clarifying points: not everything is AI (sometimes it’s just basic automation and digital transformation), and “just AI it” isn’t a strategy. The line, said National Grid’s Joe Kearns, is real insight at scale — one person can’t review 250,000 images in minutes, and that’s where the technology earns its place. National Grid is already using machine learning to predict where and when reactive gas works will be needed, down to an area and a time window, so crews can be planned rather than scrambling to get them to where they need to go. 

The use cases stacked up fast. Max Tuttman of the Ad Hoc Group described startups gaining traction in risk reduction and efficiency, such as wildfire detection or satellite-driven vegetation management. But they soon learn that utilities run on a very different timeline than startups that need to provide their investors with quarterly updates — one utility, he noted, spent a year deciding whether goats grazing a solar farm counted as capital or O&M. 

Terry Young of Arcadis made the case for predictive maintenance that recalibrates as a living plan. Ed Shappell of WSB pointed to speed — AI paired with LiDAR, imagery, and GPR is accelerating permitting and construction — with QA/QC wrapped around every output.

Then Itzik Malka, 4M's CEO and co-founder, reframed the whole problem. 4M isn't solving "mapping," he said. It's connecting what was designed to what was actually built — closing the gap where fidelity gets lost across fragmented PDFs, CAD, GIS, and permit data. The hardest and most important piece is updatability: the moment you leave the site, your data becomes a record. Real AI in this space isn't guessing the next word in a sentence; it's visual reasoning about physical reality, built on a context layer and verified by the people who carry the knowledge — the translators, as Malka called them, the Bob Terjesens and the Massoud Tahamtanis. 

The cultural work is just as real. Kearns's answer to adoption was three words — change management, change management, change management — and a champions network that lets ideas come back as the team's own. Van Gulden put the stakes in plain terms: adopting AI is the shovel-to-the-jackhammer moment. You can still dig a hole with a shovel, but the crews that reach for the jackhammer open the road faster, and the ones who don't will slowly fade. The real task, he said, is building the next generation of professionals who pick up the better tool. Shappell and Young pushed data standards and "trust but verify." 

Malka argued that sharing data saves more than it risks, and that the reluctance often traces to a hard truth: many owners simply don't know where their facilities are. Records decades old, drawn without GPS, are like perfume — meant to be smelled and sampled, not swallowed whole, because they’re not ready to use. 

The proof was in the numbers. In its pilot with National Grid, using public data alone, 4M placed lines with 89% accuracy within three feet — on the right side of the street. With utility data added, that climbs toward 95%, with the potential to prevent an estimated 40% of damages. As Malka put it, one plus one can equal five.

Live On-Site Activation — Visualizing the Underground

The day's ideas got a field test on the street outside, where 4M and Trimble ran a live constructability walkthrough on Sutton Street, right outside the venue, with a real scenario: The city needs to replace a 100-year-old water main, but a National Grid gas main sits right where the new line has to go.

4M streamed its data into Trimble Connect, and a field crew visualized it on the real street through Trimble's SiteVision augmented reality tool: gas lines, conduit, and faint road marks pulled from years of historic imagery, overlaid in place. Crews measured a vault without standing over it and flagged an electric line that would need coordination; in the model, WSB shifted the gas main to the far side of the street, clearing every crossing. Field validation was happening in real time while the audience watched from inside. Office to field in minutes. One connected workflow involving people, data, and technology on a real street in a real-life scenario.

That is what Build Better Together actually looks like. Bringing people, data, and technology together.

What’s next 

The New York event built on the momentum and the discussions from the first Build Better Together Roadshow in Georgia. And that momentum was built by once again bringing everyone in the room together to talk honestly about what should be and what actually is right now, and where we go from here. 

As Sarah Magruder Lyle put it in her keynote, "The technology’s here. We just have to adopt it — we can find it, we can do it, we can use it."

The next stop is coming. If you want to be in the room where these conversations happen, now is the time to get on the list.

👉 Sign up for the next Build Better Together Roadshow event.