In short: Yard management fails for a structural reason: the warehouse network stops at the dock door, and the yard beyond it, often 30 or 40 acres holding several hundred trailers, has no coverage of its own. Shunters, gatehouses and trailer trackers fall back to public 4G and paper, and the yard management system reports a yard that does not exist. A network planned for the yard is what makes the software true.
Key Takeaways
- The yard is the blind spot of the supply chain — A 100-door distribution centre can hold 300 trailers outside, and the warehouse system knows the contents of every one and the location of almost none.
- The WiFi stops at the dock door — Access points inside a steel shed do not reach a 40-acre yard, so shunter tablets, gate terminals and trailer tags run on public 4G or on nothing.
- Cover the yard once, then run every system on it — Private 5G radios on the building and the lighting columns give the yard management system, the gatehouse and the trailer trackers one network to believe.
In a nutshell
What is yard management and why does it fail at the dock door?
Yard management is the control of trailers, tractors and drivers between the site gate and the dock doors, and it fails because that space is the one part of a distribution centre nobody built a network for.
A large UK distribution centre is a steel shed of 500,000 square feet or more with 100 or more dock doors, and a yard around it that holds trailers waiting to be unloaded, trailers loaded and waiting for a tractor, empty trailers, and the shunter tugs that move them all. The warehouse management system knows what is on each trailer. The transport management system knows when each was due. The yard management system, from Blue Yonder, Manhattan, C3 Solutions, Peripass or a home-grown spreadsheet, is meant to know where each trailer physically is and which door it is going to next. On most sites it knows that only as well as the last person who typed it in.
The typing happens in the shed, because that is where the network is. The shunter driving the trailer from bay 42 to door 17 is in a cab with no reliable connection, so the move is radioed, remembered and entered later. The gatehouse is a separate building on the far side of the yard with a broadband line of its own. The trailer tracker, if there is one, reports over a public mobile network from wherever it can find signal. The yard management system reconciles all of these into a picture that is minutes to hours behind reality, and the yard runs on the shunter's memory and the supervisor's walk-round.
Where do trailers get lost?
Between the gate transaction and the dock transaction, in the part of the process that has no system of record.
The driver arrives, checks in at the gatehouse and is told to drop the trailer in a bay. The gate records the arrival. The drop happens in a bay that may or may not be the one instructed, because the bay was occupied or the driver misread the sign, and nothing records it. Hours later, the warehouse calls for that trailer. The shunter goes to the bay the system shows, finds a different trailer, and starts driving the rows. On a 300-trailer yard that search is a real cost, measured in the door standing empty and the pick wave that started without its stock.
The knock-on effects land on the road side. Drivers waiting for a door burn their hours under the drivers' hours rules, and the Road Haulage Association has campaigned for years about the unpaid waiting time hauliers absorb at distribution centres. Missed delivery slots incur penalties in most retail supply contracts. Reefer trailers left in a bay the system forgot are a food-safety incident waiting for an audit. And at peak, when the yard fills and the same shunters are asked to move twice as many trailers, the gap between the system and the tarmac widens exactly when the site can least afford it.
Sites in the Golden Triangle, around Daventry, Lutterworth and Hams Hall, run this at a scale where a few percent of trailer moves going wrong is a daily operational problem rather than an occasional one. We have yet to visit a large yard where the yard management system's location data was trusted without a human check.
Why does the warehouse WiFi stop at the dock door?
Because it was designed to, and because extending it across a yard is a poor use of WiFi.
Warehouse WiFi is planned for the inside of the shed, and it is hard enough there, as we set out in our piece on why warehouse WiFi fails. Access points sit on the roof structure and are tuned for scanners in the aisles. The steel cladding of the building is an effective screen, so what leaks out to the yard is a weak signal a few metres beyond the doors. Extending it means outdoor access points on the building, on lighting columns and in the gatehouse, each needing power and a fibre or wireless link back, and WiFi's short range means many of them. Between the access points the shunter's tablet roams, and roaming is where WiFi client devices disconnect. A 40-acre yard with trailers acting as moving steel walls is the same environment that defeats WiFi inside the shed, spread over ten times the area.
So sites fall back to public 4G. For a tablet in a shunter cab this is workable where the coverage is decent, and many distribution parks have decent coverage on the road frontage. It is less reliable at the back of the yard, behind the building, and it is a shared resource: at a large park, the same cells serve every driver waiting in the lorry park, every warehouse colleague on a break and every neighbouring site's devices. The connection is nobody's responsibility, so when a shunter's move fails to post, there is no one to call.
What does trailer tracking need from a network?
A signal in every bay, a low-power way for a tag to report, and a system that trusts the tag more than the typed entry.
Trailer tracking comes in two forms. Fleet trailers get a powered or solar GPS tracker that reports over a public network, which tells the transport office where the trailer is to within a few metres and tells the yard management system nothing it can act on unless the two are integrated. Yard tracking, which is what the site needs, is finer: which bay, which door, since when. That is done with a tag on the trailer, or a reader on the shunter and at each door, or increasingly with cameras reading the trailer's number plate and fleet number at the gate, on the shunter routes and at the doors.
Each approach needs a network that reaches the whole yard. Tags need a gateway in range of every bay. Shunter readers need a live link to post moves as they happen. Cameras need uplink for continuous imagery, which is the demanding case: a dozen cameras covering the yard produce more upstream traffic than the whole warehouse's scanner fleet. Public 4G is the wrong tool for all three, because its uplink is thin, its coverage is uneven at the back of the site and its capacity is shared.
When the network is right, the yard management system changes character. A trailer drop is recorded when it happens, by the reader or the camera, not by the driver's memory. The shunter's next job is dispatched to the cab, with the bay confirmed. The gatehouse sees the yard as it is and can tell an arriving driver that there is no bay yet rather than sending them in to circle. Dock scheduling stops being a plan and becomes a live queue.
How does private 5G cover the yard, the gatehouse and the shunter cab?
By treating the yard as the site's primary outdoor coverage area, with a small number of cellular radios sited to reach every bay, rather than as a spillover from the building.
A private 5G network for a distribution centre puts radios on the corners of the building and, where the yard extends beyond their reach, on lighting columns, using Ofcom's shared access licence in the 3.8 to 4.2 GHz band. A site of 40 acres is typically covered by a handful of radios, planned against the yard layout and the trailer rows, at a power level WiFi cannot match. The shunter's tablet holds one connection as it crosses the yard rather than roaming between access points, the gatehouse joins the same network as the shed, and the cameras on the yard get the uplink they need. The same coverage extends into the dock area and, with indoor radios, into the shed itself, which lets a site retire the worst of its warehouse WiFi in the same project.
The trade-offs are the usual ones. A private network is a capital project with radio planning, a licence, a core to run and devices to migrate, and a small regional depot with 20 doors and a yard the size of a car park will not justify it. For that site, a couple of outdoor access points and a public 4G tablet in the shunter cab may be enough. Our argument is for the sites where the yard is large, the trailer count is in the hundreds, the shunter fleet works around the clock and the yard management system has been bought and not believed. There, the network is the missing component, and it costs less than the software it makes useful.
What should a distribution centre do first?
Measure how long the yard takes to find a trailer, because that figure will carry the business case on its own.
Most sites do not know it. The shunter's search time, the door dwell caused by a late trailer, the driver waiting time at the gate and the number of yard moves per shift are all measurable with the site's existing systems plus a week of observation, and together they describe the cost of a yard that runs on memory. In our experience the number surprises the operations director and is the point at which the yard management system's licence cost stops looking like the problem.
Then survey the yard as the shed was surveyed. Walk the bays with a device and record what the public networks and any existing WiFi deliver at each row, behind the building and at the gatehouse. Map that against where the shunters, readers and cameras will need to work. On most large yards we see the coverage ends where the work begins, and the survey makes that visible to the people who sign the capital request.
Finally, plan the network for the yard first and the shed second, and choose the trailer-tracking method, tags, readers or cameras, on the assumption that coverage will be there.