Charging for Damaged and Lost Traffic Control Devices

Every traffic control company has an inventory count that comes up short and a stack of drums nobody can quite account for. The devices went out, some of them did not come back, and the specific job that consumed them is unknowable by the time anyone notices.

That is not a small cost. Channelizing devices, signs, stands, lights, and barricades represent real capital that depreciates through use and disappears through attrition, and in most companies the entire loss is absorbed as overhead — priced into nothing, recovered from nobody, and invisible on any job's margin.

This is about making it visible and, where the contract allows, recoverable.

Three different problems that get called one thing

"Lost equipment" covers at least three situations with different causes and different remedies.

Destroyed on the job. A vehicle takes out a taper. Wind flattens an array. A plow removes forty cones in one pass. The devices are gone and there is a specific, datable event behind it.

Left behind. A partial pickup, devices staged out of sight, a crew that ran out of daylight. These are frequently recoverable if anyone notices within a day.

Walked off. Theft, or a well-meaning contractor who "borrowed" your barricades for their own scope. Common, rarely reported, almost never billed.

Only the first has an obvious third party attached. The other two are internal control failures — which is precisely why lumping them together as one attrition number prevents you from fixing either.

The manifest is the whole mechanism

The tracking problem has a boring solution: know what left the yard for a specific job, and know what came back.

That means a device manifest attached to the job — counts by type, ideally with identifiers on higher-value items like arrow boards and message signs — checked out at load and checked in at return. The delta is the attrition for that job, computed automatically rather than discovered at an annual count.

Two practical notes from companies that have tried this and had it fail.

Check-out has to be faster than not checking out. If it is a paper form or a desktop entry, it will be skipped on the mornings that matter — which are the busy ones, which are the ones with the most attrition. Scan or tap on a phone, or it will not survive contact with a real yard.

Emergency and after-hours loads are where the control breaks. Devices grabbed at midnight for a callout do not go through the daylight process. This is where most unexplained shortfall originates, and it is worth a deliberate exception path rather than pretending the normal process covers it. Our post on emergency callout dispatch covers the broader shape of that problem.

The location side of this — knowing which job site devices are currently sitting on — is covered in tracking traffic control devices by job site. Location and attrition are different questions answered by the same manifest.

Evidence has to be captured at the site

This is the part that determines whether damage is recoverable, and it is the step almost universally skipped.

Damaged devices photographed at the yard prove nothing. They are damaged devices in a yard. Photographed on the road, geotagged and timestamped, with the job context visible, they are evidence of what happened on a specific job on a specific night.

The sequence that works: when the crew finds damage, photograph it in place before loading, note the count and type on the ticket, and have the customer's representative acknowledge it at signature. That third step is what converts a claim into a settled item, and it costs about ninety seconds.

If a vehicle caused the damage, that event is also an intrusion, and it belongs in your safety reporting whether or not anyone was hurt — see work zone intrusion prevention and response. Property-damage-only events appear in no national dataset, which means your own record is the only one that will ever exist.

The frequency is not marginal. The Associated General Contractors of America, reporting its annual survey with HCSS, states that "Sixty percent of highway construction firms report having at least one car crash into their work zones during the past year". Most of those crashes destroyed something before they were counted as a crash.

What the contract actually says

Recovery is a contract question, and this is where general advice stops being useful.

Many traffic control and equipment rental agreements place devices in the customer's care, custody, and control while deployed on their site, with the customer responsible for loss or damage other than ordinary wear. Many do not. Some carve out third-party vehicle damage specifically; some are silent, which is worse.

Three things worth doing, none of which are legal advice:

Read your own agreement and know which category it is in. A surprising number of companies do not know, because the template was written by someone who left.

Have counsel review the damage and loss language once. It is a small, one-time expense against a recurring cost line.

Define replacement values in advance. A schedule of device replacement costs, attached to the agreement, removes the negotiation from the moment of loss. Without it, every damage claim becomes a debate about what a used drum is worth.

Standing caveat: contract terms, and their enforceability, vary by jurisdiction and by the specific agreement. Nothing here is legal advice or an opinion about whether a particular clause applies to your situation. Have a lawyer look at your actual paperwork.

Pricing attrition when you cannot recover it

Sometimes recovery is not available — the contract does not support it, the customer relationship does not survive it, or the loss category is internal. That does not mean the cost disappears; it means it has to be priced.

The right home for unrecoverable attrition is in the rental rate itself, computed from measured loss rates by job type rather than from a guess. Our post on device rental rates and pricing covers rate construction; attrition is one of its inputs, and it is the input most often set to zero by default.

Two categories usually justify different rates once you measure them. Devices deployed on high-speed, high-truck corridors attrit faster — the crash data is unambiguous that heavy vehicles are overrepresented, with the Work Zone Safety Information Clearinghouse recording 233 truck-involved fatal work zone crashes in 2024 accounting for 282 deaths. And long-duration unattended deployments attrit faster than attended short-duration work, for obvious reasons.

If your rate card charges the same for a two-hour attended closure and a three-week unattended one, you are subsidizing the second with the first.

The reporting that makes this actionable

Once manifests exist, four numbers become available and all four are useful.

Attrition rate by job type. Which work actually consumes your inventory.

Attrition rate by customer. Occasionally illuminating, occasionally the beginning of a difficult conversation.

Recovery rate. Of the damage you documented, how much was actually billed and collected. A low number usually means the evidence was captured too late, not that the contract was weak.

Replacement spend against rental revenue. The single ratio that tells you whether the rate card is working.

None of these require a data project. They require that check-out, check-in, and damage capture happen on the job, in a system where the job is the organizing record. That is how Traffic OS handles equipment and rentals, and the pricing page lays out what is at each tier. You can book a walkthrough to see the manifest and damage-capture flow on real screens.

The devices that disappear are not the expensive problem. The expensive problem is that they disappear anonymously — no job, no date, no cause — which means the same job type quietly consumes inventory year after year and nobody can point at it.