Every carrier has the story. A driver follows the route, the route goes under a bridge that a 13’6″ trailer does not fit under, and the day ends with police, a closed road, and a phone call nobody wants to make.

The reflex is to blame the driver. Sometimes that’s fair. Often the driver did exactly what the screen said, and the screen was working from data that was wrong, missing, or never meant for a truck at all.

This post is about where that data comes from, why it fails, and what to ask a vendor before you trust it with an 80,000-pound vehicle.

Key Takeaways

• Nobody keeps a national count of bridge strikes. New York alone had 350 in 2024, and Pennsylvania more than 600 between 2013 and 2023, figures cited by members of Congress (FreightWaves, December 2025).
• Texas spent about $6.7 million a year repairing struck bridges (Stateline, 2019).
• Consumer navigation has no concept of trailer height. It isn’t a setting you forgot to turn on.

How Often Does This Actually Happen?

Nobody keeps a national count. The federal trucking regulator doesn’t track bridge strikes (Stateline, Trucks vs. Bridges, 2019), and the Federal Highway Administration’s 2018 report on them said only that there are thousands a year (as cited by Trimble). State numbers show the scale: New York had 350 bridge strikes in 2024 alone, and Pennsylvania more than 600 between 2013 and 2023 (FreightWaves, December 2025).

That’s not a rare catastrophe. It’s a routine, expensive failure mode.

The repair bill is the visible part. The rest of the cost is the closed road, the load that didn’t deliver, the insurance consequence, and the driver who may not be driving for you afterward.

Why Doesn’t Consumer Navigation Work for Trucks?

Because it was never trying to. Waze and Google Maps route a car, and a car has no height, weight, axle count, or hazmat class. There’s no truck mode hiding in the settings.

The members of Congress behind a December 2025 bill on truck-bridge crashes named trucks following car GPS as part of the problem (FreightWaves, New legislation tackles truck-bridge crashes). The mechanism is not in dispute: a routing engine that doesn’t model the vehicle can’t avoid a constraint it doesn’t know exists.

Here’s the part that catches experienced people out. A driver using consumer navigation isn’t being careless. The route looks plausible, the road is a real road, and nothing warns them. The failure is silent right up until it isn’t.

Truck navigation isn’t consumer navigation with a truck icon. It’s a different problem: the map has to carry clearance, weight limit, bridge posting, and restriction data per segment, and that data has to be right.

Where Does Truck Restriction Data Actually Come From?

This is the question almost nobody asks their vendor, and it’s the one that determines whether the routing works.

Restriction data is assembled from public sources: the National Bridge Inventory, the National Transportation Atlas Database, and state DOT restriction files, which each state publishes in its own format on its own schedule. Somebody has to ingest all of it, reconcile it, and match it to road segments.

Most trucking platforms don’t do that work. They license a finished routing product, which means the restriction data arrives as a black box. When it’s wrong, the vendor can’t tell you why, and neither can you.

We took the other path and built the pipeline ourselves, across all 50 states. I’d like to tell you that made everything clean. It did the opposite, and that’s the actual argument for owning it.

What owning the parser found: in one state’s restriction file, “no restriction recorded” was encoded with a sentinel value that a straightforward reader interprets as a hard block. Consumed as-is, it would have routed trucks around 1,498 interstate segments that are perfectly legal. Our record-count checks passed the whole time, because the rows were all present. They were just inverted. A licensed feed would have shipped that to us as gospel.

That’s the case for provenance in one paragraph. Not that our data is flawless, but that when it’s wrong we can find out and fix it.

What Should You Ask a Routing Vendor?

Four questions separate real truck routing from a truck icon on a car map. None of them requires you to be technical.

1. Where does your restriction data come from, and how often is it refreshed? “We license it” is an acceptable answer if they can name the source and the cadence. “It’s proprietary” means nobody there knows.
2. What happens when a restriction is wrong? Ask about the actual process. If the answer routes through a third party’s support queue, your correction takes months or never happens.
3. Does the route account for the trailer I’m actually pulling? Height, weight, length, axle configuration, and hazmat should be vehicle profile settings that change the route, not a disclaimer.
4. Can the driver see why a road was avoided? A route that says “avoided: 12’6″ clearance on US-1” builds trust. A route that silently detours 40 miles gets overridden by a driver who thinks it’s being stupid.

That last one comes from running trucks. A driver who doesn’t understand a detour will straighten it out themselves, and then the routing engine’s correctness stops mattering.

Is Truck Routing Enough on Its Own?

No, and I want to be careful here, because this is where the industry oversells.

No routing engine can promise a legal route everywhere. Restriction data is only as complete as what states publish, temporary postings change faster than any file updates, and construction detours appear overnight. What good routing does is route on every restriction in the public record and tell you what it knows.

Anyone selling you a guarantee is selling you something they can’t deliver. What you should expect is auditable coverage, clear reasons for detours, and a vendor who can fix an error when you find one.

The driver remains the last line. A routing engine that encourages a driver to stop checking signs has made things worse, not better.

Where This Fits in a Small Fleet

Routing sits oddly in most software stacks. Dispatch lives in the TMS, navigation lives on the driver’s phone in a separate app, and the two don’t talk. The dispatcher plans a route the driver never sees, and the driver follows a route the dispatcher can’t inspect.

That split is why a load can be planned on one path and driven on another, and why nobody notices until the miles don’t reconcile.

That’s the gap we’re closing in Oculus Freight. Our routing engine already powers its fuel planning, and truck navigation on the same restriction data is coming to the driver app, so the route planned will be the route driven. Until then, the app opens each stop in Trimble CoPilot or Google Maps.

For a five-truck fleet, is this the deciding factor? Honestly, maybe not on day one. One bridge strike, with the repair bill, the lost load and the insurance consequence, changes that math quickly.

Frequently Asked Questions

Can I just use Google Maps with a truck setting?

There isn’t one. Google Maps and Waze route passenger vehicles and have no model of trailer height, weight, or bridge clearance. Lawmakers pushing for a federal fix name car GPS in trucks as part of the problem.

How accurate is truck routing data?

It depends entirely on the source and how recently it was refreshed. Restriction data comes from the National Bridge Inventory, NTAD, and state DOT files, and each state publishes on its own schedule. Ask any vendor to name their sources and refresh cadence.

What does a bridge strike actually cost?

There’s no national average. Texas spent about $6.7 million a year repairing struck bridges and recorded about a hundred strikes in 2018 (Stateline, 2019). That excludes the truck, the load, road closure costs and insurance consequences.

Does Oculus Freight guarantee a legal route?

No, and be suspicious of anyone who does. Our routing engine uses the restrictions in the public record, from data we maintain ourselves, and the truck navigation coming to the Oculus Freight driver app will show why a road was avoided. Temporary postings and unpublished local restrictions exist. The driver is still the last check.

Why does route provenance matter if the route looks fine?

Because routing errors are silent. A wrong restriction shows up as an unexplained detour or a missing warning, never as an error message. If your vendor can’t inspect the data, nobody can tell a real restriction from a data artifact.

The Short Version

Bridge strikes aren’t rare, they’re expensive, and most trace back to a routing engine that had no idea how tall the vehicle was.

Ask where the restriction data comes from. Ask what happens when it’s wrong. Ask whether the driver can see why a road was avoided. Those three answers tell you more than any feature list.

To see what Oculus Freight does today, see the founding-carrier offer or tell us what you’re running today. For the wider platform question, see our comparison of TMS platforms for small carriers, and for the money side, how to actually get paid for detention.

---

Sources

• Stateline (Pew Charitable Trusts), Trucks vs. Bridges: Dangerous and Costly, December 2019, republished by Gov1. Retrieved 2026-10-17. https://www.gov1.com/infrastructure/articles/trucks-vs-bridges-dangerous-and-costly-nwsqy4YZB8dUzctJ/
• FreightWaves, New legislation tackles truck-bridge crashes, December 16, 2025. Retrieved 2026-10-17. https://www.freightwaves.com/news/new-legislation-tackles-truck-bridge-crashes
• Trimble Transportation, Avoid Bridge Strikes with Commercial Vehicle Specific Routing. Retrieved 2026-09-02. https://transportation.trimble.com/blog/avoid-bridge-strikes-with-commercial-vehicle-specific-routing
• Federal Highway Administration, National Bridge Inventory. https://www.fhwa.dot.gov/bridge/nbi.cfm
• Bureau of Transportation Statistics, National Transportation Atlas Database. https://www.bts.gov/ntad