Running Two Yards: What Breaks, and What Software Has to Handle

The second yard usually opens for a good reason. A cluster of work forty miles away, a customer whose sites are all in one corridor, a lease that came up cheap. Revenue justifies it easily.

What is less obvious is that the second yard changes the shape of the operation. A set of assumptions that were true and invisible at one location stop being true, and because they were invisible nobody notices they have failed until a crew is standing in a yard that does not have what the load list says.

The assumption that breaks

At one yard, "how many drums do we own" and "how many drums can tomorrow's crew load" are the same question. That equivalence is doing enormous unacknowledged work.

At two yards it dissolves. You can own four hundred drums and be short at the yard that matters. The company-wide number is not merely less useful — it is actively misleading, because it reads as healthy while a crew is blocked.

This is why spreadsheet inventory tends to survive the first yard and fail at the second. A single-sheet count can be maintained by one person who knows what is out there. Two locations require the count to be located, and once every row needs a location, the manual version stops being maintainable in the way that migrating off spreadsheets generally describes.

Transfers are the leak

The mechanic that undoes multi-yard inventory faster than anything else is the unrecorded transfer.

A crew finishing near the north yard drops forty cones there rather than driving them south. Entirely sensible. Nobody records it. Now the south yard's count is forty high and the north yard's is forty low, and neither number is wrong in a way anyone can see — the company total still reconciles perfectly.

That last point is what makes it insidious. The error is invisible to exactly the check most companies run. Totals balance. Only the per-location figures are wrong, and they are wrong in opposite directions, which is the signature of a transfer nobody logged.

Two things fix it, and only two. The transfer has to be recordable at the moment and place it happens, from a phone in a yard, in under thirty seconds — because a process requiring a form in an office will lose to the reality of a crew unloading in the rain. And the receiving end has to be confirmable, so a load that left one yard and never arrived shows up as in-transit rather than silently landing.

What dispatch has to know

Adding a yard adds a variable to every assignment.

Assigning a job stops being crew-and-time and becomes crew-and-time-and-which-yard-has-the-devices. Without per-yard visibility, dispatchers do the only rational thing available: they default to the yard they know best. That is how one location develops a chronic shortage while the other accumulates surplus, and neither shows up in a company-wide count.

It also changes what a job costs. A crew loading from a yard sixty miles the wrong way is carrying an hour of unbilled drive time before the shift starts. That cost is real, it is attributable, and in most companies it is invisible — it disappears into general labour rather than attaching to the job that caused it, which quietly corrupts job costing and, downstream, pricing.

The dispatch question worth being able to answer in ten seconds: for tomorrow's four jobs, which yard should each crew load from, and does that yard have the devices? Our daily dispatch routine covers the wider morning problem; the yard dimension is what a second location adds to it.

What software actually has to do

Concrete requirements, in the order they bite.

Stock by location, not just in total. Every device count must resolve to a yard. A platform that models one warehouse will require you to fake the second one, and faked structure decays.

Transfers as a first-class movement. Not an adjustment out and an adjustment in — a transfer with an origin, a destination, a quantity, a person, and a state. In-transit has to be a real status, otherwise devices vanish between locations.

Mobile capture at the yard. The record has to be made by whoever moved the devices, standing where they moved them. Every alternative depends on someone remembering later.

Allocation against jobs, per yard. Devices committed to Thursday's closure should not appear available for Wednesday's, and that reservation has to be location-aware.

On-rent visibility across locations. Devices sitting on a customer's site for six weeks are neither at a yard nor available, and they still need to be billed. If the platform cannot hold "on rent at customer site, dispatched from north yard," your rental billing and your inventory will disagree permanently.

Per-yard reporting. Utilisation, shrinkage, and turns by location. Without it you cannot tell whether the second yard is earning its lease.

Traffic OS models yards as distinct inventory locations with transfers between them, so devices resolve to a place rather than to a company total, and the features page covers how that connects to jobs and rentals.

Setting a minimum stock level per yard

Once stock is located, the obvious next question is how much each location should hold — and this is answerable with arithmetic rather than instinct.

Start from the largest setup that yard realistically serves. Device counts fall out of the plan: MUTCD Table 6B-4 gives taper length as W times S squared divided by 60 at 40 mph or less and W times S at 45 mph or more, and channelizing device spacing should not exceed a distance in feet equal to 1 times the speed limit in mph for taper channelization and 2 times for tangent channelization. A 12-foot offset at 45 mph is a 540-foot merging taper with devices no more than 45 feet apart — thirteen devices for that taper alone, before the tangent, the downstream taper, or the opposite direction. That is the method in quantity takeoffs, applied to stocking rather than bidding.

Then layer on concurrency and condition. A yard serving three simultaneous jobs of that size needs three setups' worth plus spares, and spares are not optional given that devices are destroyed, buried, and taken. Condition matters as much as count: devices must be maintained to remain visible and effective, which is why sign retroreflectivity and device condition standards belong in the count. Fifty drums of which fifteen are unserviceable is a stock of thirty-five, and only a per-yard record that distinguishes serviceable from present will tell you that.

The wider risk context is worth keeping in view when deciding whether to run a location thin: FHWA reports 891 work zone fatalities in 2022 and 963 in 2021, alongside 94 highway construction worker occupational fatalities in 2022. A crew improvising a setup because the nearer yard was short is not a purely commercial problem.

Where the yard is not really a yard

Worth flagging, because these are the locations that break counts while nobody considers them locations at all.

A long-running project with staged devices. Six weeks of phased work with material left on site is a stocking location whether or not anyone calls it one.

A foreman's trailer with a standing stock of signs. Small, permanent, and entirely undocumented in most companies.

A customer's yard. Common on utility contracts, and a genuine control problem — devices there are on your books and outside your gate.

The test is simple: if a crew can load from it, it is a location. Whether it has a lease is irrelevant to inventory.

Yard discipline is not a software problem

The honest caveat. Software makes multi-yard inventory possible; it does not make it accurate. An unattended yard where any crew can take what they need without recording it will diverge from its count within weeks no matter what platform you buy.

What is required alongside the tooling is narrow: someone accountable for each location, a rule that loading is recorded before leaving, and a physical count on a schedule you actually keep. A monthly per-yard count that gets done beats a quarterly one that does not, and the first honest count after opening a second yard is usually unpleasant — which is the point of doing it.

Reconciling that count against expectation is also how you separate the two problems that look identical from a distance. A shortfall that appears at one yard and a surplus at another is a transfer discipline problem. A shortfall with no matching surplus anywhere is loss — damage, burial, or theft, which is the subject of equipment theft and vandalism and belongs in a different conversation entirely.

Get the locations into the system, make transfers a thirty-second action on a phone, and count on a schedule. The second yard then does what it was opened to do — put devices closer to the work — rather than doubling the number of places your inventory can be wrong.