In short: The modern distribution centre runs on a fleet of autonomous mobile robots, voice-directed pickers and a constant stream of scans, and it has quietly outgrown the warehouse Wi-Fi that has carried it for twenty years. Coverage gaps between racking, roaming handovers that drop a robot mid-aisle, and contention under peak load are no longer minor annoyances — they are the thing limiting throughput. Private 5G gives a large shed a single, deterministic network built for movement, and that is why the UK's biggest logistics parks are starting to specify it.
Key Takeaways
- AMRs are the new floor staff, and they cannot tolerate dead spots — autonomous mobile robots navigating between racking need continuous coverage and clean handover, or they stop in the aisle and wait
- Wi-Fi was built for a static shed, not a moving one — coverage between metal racking, roaming delays and unlicensed-band contention all degrade exactly when peak-season volume is highest
- Private 5G is one network for the whole site — robots, voice picking, scanners, yard management and the trailer park all run on a single licensed network instead of a patchwork of overlapping systems
In a nutshell
The shed has changed faster than its network
Drive past Magna Park at Lutterworth, DIRFT at Daventry, or any of the big distribution parks strung along the M1 and M6, and the buildings look much as they did a decade ago: vast steel sheds, hundreds of thousands of square feet under one roof, lorries queuing at the yard. What has changed is what happens inside. The distribution centre that used to be people with trolleys and the occasional forklift is now a choreographed mix of autonomous mobile robots, conveyor and sortation systems, voice-directed and light-directed picking, and goods-to-person automation that brings the stock to the picker rather than sending the picker to the stock.
That shift has been driven by the same pressures squeezing the whole logistics sector: e-commerce volumes that demand same-day and next-day fulfilment, a persistently tight warehouse labour market, and customers — whether retail chains or online shoppers — who treat fast, accurate delivery as the baseline rather than a premium. Automation is how operators square rising volume and service expectations with the people they can actually recruit.
But almost every one of these technologies has one thing in common: it is wireless, and it is mobile. The robot moves. The picker with a voice headset and a wrist-mounted scanner moves. The forklift with a mounted terminal moves. And the network underneath them was, in most existing sheds, designed for a world where the wireless devices were handheld scanners carried by people walking at human pace and pausing to scan. That network is now being asked to carry a fleet of machines that never stop moving — and the strain is showing.
Why warehouse Wi-Fi struggles with a moving floor
Warehouse Wi-Fi is a mature, well-understood technology, and for two decades it did the job. The reasons it struggles now are specific and worth naming, because they explain why operators are looking past it rather than simply buying more access points.
The first is the physical environment. A high-bay warehouse is one of the most hostile settings for Wi-Fi there is: tall metal racking that blocks and reflects signal, an inventory profile that changes by the hour as stock moves, and aisles that turn into RF canyons. Coverage that was fine when the racking was half empty becomes patchy when it is full, and the dead spots are rarely where the survey predicted. For a person, a momentary dropout is invisible — they keep walking and the scanner catches up. For an AMR closing a navigation and safety loop over the network, a dropout in the wrong place means it stops and waits.
The second is roaming. Wi-Fi handover between access points was designed for devices that move occasionally; in a large shed a robot crosses dozens of access-point boundaries on every run, and each handover is a moment of risk. Roaming delays that a laptop user never notices translate, for a fleet of robots, into stutters, stops and reduced throughput.
The third is contention. Wi-Fi runs in unlicensed spectrum shared with everything else, and under peak-season load — exactly when the operation can least afford it — hundreds of devices competing for the same airtime degrade together. There is no robust way to guarantee that the robot's safety traffic or the picker's next instruction gets through ahead of less critical traffic.
The result is a network that works well enough most of the time and fails precisely when volume is highest and the cost of failure is greatest. For an operation whose entire justification is throughput, "works most of the time" is increasingly not good enough.
What private 5G brings to the warehouse floor
Private 5G addresses the warehouse's specific failures rather than just offering more of the same.
Coverage is the first and most immediate gain. A handful of well-placed 5G radios can cover a large shed more evenly than a dense mesh of Wi-Fi access points, and licensed or shared-licensed spectrum — available in the UK through Ofcom's Shared Access framework — means the operator is not fighting neighbouring networks for airtime. The coverage is planned, predictable, and holds up as the racking fills.
Mobility is the second. Cellular handover is engineered for devices that move continuously; a robot crossing the floor keeps its session as it passes from cell to cell, without the roaming stutter that plagues Wi-Fi. For a fleet of AMRs, that is the difference between a smooth flow and a floor full of machines pausing to reconnect.
Determinism is the third. Private 5G offers genuine quality-of-service and network slicing — the ability to reserve a low-latency, reliable slice for robot control and safety traffic while picking instructions, scans and general traffic share the rest. The traffic that matters most gets the treatment it needs by design, not by chance, and it holds up under peak load rather than collapsing alongside everything else.
There is also a security dimension that matters more as warehouses automate. A private network keeps the operational traffic — the robots, the warehouse management system, the control plane — on infrastructure the operator owns and controls, segmented from the corporate IT network and from the public internet. As the floor becomes more dependent on connected machines, that isolation is a meaningful reduction in the attack surface.
Beyond the picking aisle: one network for the site
The robots make the case, but the value of a private 5G network in a distribution centre comes from how much else it carries on the same infrastructure.
Voice picking and the move off handheld scanners ride it: the headsets and wearables that direct pickers by voice need continuous, reliable coverage to keep a picker's hands free and their pace up. The forklift and reach-truck terminals run on it. The fixed and handheld scanners that track every carton through goods-in, putaway, picking and dispatch run on it. Real-time location systems that show where every asset and, for safety, every person is on the floor run on it.
Crucially, the network does not stop at the shed wall. The yard — the trailer park, the loading docks, the gatehouse — is where a great deal of distribution-centre time is lost, and it is usually the worst-served area for connectivity. A private 5G network planned to cover the whole site brings the yard inside the same network: trailer tracking and yard-management systems that know which trailer is on which dock, automated gate and check-in, dock-door status, and the connected-worker and CCTV systems that keep the outdoor operation safe. Bringing the yard and the shed onto one network closes the gap where the handoff between transport and warehouse so often goes wrong.
The economics: throughput, not gadgetry
The business case for a private network in a distribution centre is not built on novelty; it is built on throughput and reliability, the two numbers the operation already lives by.
The direct return is uptime and pace. Every minute a robot spends stopped waiting for a link, every re-scan caused by a dropout, every picker standing idle for an instruction that did not arrive is throughput lost — and in a peak-season operation running flat out, that lost throughput is the difference between meeting the service-level agreement and missing it. A network that holds up under peak load protects the revenue that the whole automation investment was meant to generate.
The second return is consolidation. The alternative to one private network is the accumulation of several: Wi-Fi for the scanners, a separate system for the robots' vendor, the corporate network for the offices, and whatever covers the yard. Each is a separate contract, a separate security boundary, and a separate thing to maintain and troubleshoot. A single managed network is cheaper to run and far simpler to extend when the next wave of automation arrives.
The third is future-proofing. The trajectory of warehouse automation is clear: more robots, more autonomy, more reliance on real-time data and increasingly on cameras and machine vision for quality, safety and inventory accuracy. Each of those is more demanding of the network than the last, and they are uplink-heavy in a way that warehouse networks were never designed for. Building the connectivity layer for what the shed will be doing in five years, rather than what it was doing five years ago, is the pragmatic choice.
A connectivity decision, not a robot decision
It is tempting to treat warehouse automation as a question of which robots and which warehouse-management software to buy. The operators furthest down the road have learned that the harder and more consequential decision is the one underneath: the network that everything else depends on. The robots are increasingly a commodity; the difference between a fleet that flows and a fleet that stalls is the connectivity carrying them.
Ultimately however, the case for private 5G in the big sheds is not that Wi-Fi has failed — it served the warehouse well for a long time — but that the warehouse has changed into something Wi-Fi was never built for. A floor of continuously moving machines, voice-directed pickers and a yard full of trailers needs a single, deterministic, well-covered network that treats the traffic that matters as the priority it is. For the UK's largest distribution centres, that is increasingly what private 5G means, and it is why the connectivity layer is moving from an afterthought to one of the first decisions in any serious automation programme.