In my experience, annual drive-by inspections like the ones described in the Cox case fall far below what safety management considers a reliable method for identifying vault-lid hazards. A technician checking a cover from inside a vehicle cannot detect looseness, instability, or structural fatigue, and that's exactly the type of failure seen in incidents where pedestrians have stepped onto a lid that shifts or collapses. Best practice is a scheduled, hands-on walking inspection where the cover is physically tested for movement, seating, and surface condition, backed by a clear timetable that requires more frequent checks than once a year. Cities also face real enforcement challenges when a defective vault appears. Ownership of older telecom and utility infrastructure isn't always clearly documented, and operators sometimes dispute responsibility, which delays both corrective work and accountability. When inspection methods are limited and responsibility is unclear, a utility can technically meet the minimum requirements on paper while a sidewalk hazard remains in place. That combination is what makes these incidents far more preventable than they appear.
A sidewalk vault that collapses under someone's feet isn't an accident, it's a predictable failure caused by weak inspection habits and outdated standards. My practice focuses on criminal defense, but I deal with questions of responsibility every day. When a utility relies on quick visual checks or annual "drive-bys," that's not real oversight. A vault cover can appear fine from behind a windshield and still be one bad step away from giving out. No investigator would ever sign off on that level of review when the risk involves a public walkway. Cities deal with another recurring problem. After someone is hurt, the utility may deny ownership of the vault. That leaves municipalities chasing answers instead of enforcing maintenance. In any field, when no one accepts responsibility, the risk grows. Infrastructure is no different. If the party in control isn't clearly identified, safety falls through the cracks. Real safety management requires hands-on inspections, documented maintenance, and a clear chain of responsibility. Municipalities can only enforce standards when they know exactly who owns the structure. And utilities cannot rely on minimum compliance when those rules were written long before modern traffic, foot patterns, and load demands. Major injuries often reveal what has been ignored for years. Strong oversight prevents these failures long before a lawsuit forces everyone to pay attention.
In any large system, from a cloud platform to a city's infrastructure, there's a huge difference between designing for compliance and designing for safety. Most utility inspection practices we see today, like annual drive-by checks, are built for compliance. They produce a simple, auditable record that a box was checked on a certain date, which is enough to satisfy a regulator. The problem is that this approach treats infrastructure as if it's static. But it's not. It's a dynamic system that is constantly breaking down under the stress of traffic, weather, and time. The real issue is a failure of imagination in how we think about risk. A drive-by check is just passive data collection. It can only spot a problem that has already become obvious. A modern safety system, on the other hand, should be predictive. It would model the lifecycle of these assets, using data on materials, installation dates, repair history, and environmental stress to identify which specific part is most likely to fail next. So when utilities deny ownership of a piece of equipment, they aren't just dodging financial responsibility. They are creating critical gaps in this data system. What this means is that a true, city-wide risk assessment becomes impossible. I remember mentoring a young engineer who built a health dashboard for a critical software system. It showed a beautiful sea of green "OK" lights. He was proud, but I had to explain that "running" is not the same as "healthy." We weren't measuring the subtle signs of system load and memory fatigue that come right before a crash. We were only set up to report a failure after it happened. Municipal oversight of public utilities often operates on this same principle. They are managing a dashboard of green lights, mistaking the absence of a visible crisis for the presence of safety. But safety isn't a status you just check. It's a condition you must continuously create.
I've spent years dealing with infrastructure challenges on federal projects across all 50 states, and one theme shows up every time: inspection practices often lag behind what real safety requires. Many utilities rely on quick visual checks or drive-bys, which do little more than confirm the cover is still there. What they don't do is identify structural fatigue, shifting soils, or load issues, problems that only show up through hands-on inspections and proper testing. Cities face another hurdle when utilities push back on ownership. I've seen projects stall for weeks because no one would claim responsibility for a damaged vault or cover. That gridlock leaves the public exposed. Until cities have clearer authority and faster escalation processes, dangerous conditions will sit untouched. Most municipal teams want to enforce standards, but they're stretched thin. A single inspector might cover hundreds of assets. The result is a constant gap between what should happen and what can realistically get done. That's where better technology and clearer accountability would make an immediate impact. AI can help close the gap. We've already seen how automation simplifies complex material-sourcing decisions. The same principles apply here. Instead of waiting for annual checks, systems could flag hazards earlier and route maintenance faster. Preventing catastrophic injuries requires more than checking boxes. It requires modern tools, clear ownership, and inspection practices built around how infrastructure actually behaves, not outdated minimums.