UWB (Ultra-Wideband) Indoor Positioning

Ask a warehouse operations manager where a particular forklift, roll cage or high-value pallet is right now, and the honest answer is usually a shrug and a radio call. Most identification technologies tell you an asset passed a fixed point at a certain time, or that it is somewhere within a large zone, but not exactly where it sits at this moment. Ultra-wideband, UWB, is the technology that closes that gap. It measures the real-time position of a tagged asset to within a handful of centimetres, indoors, where satellite positioning is useless. This guide sits inside the broader warehouse automation complete guide, and it goes deep on one specific question: what UWB positioning is, how it works, and whether the infrastructure it demands earns its place in your building.

The message up front: UWB is not a better barcode and it is not a cheaper GPS. It is a precision indoor positioning system that answers a different question entirely, "exactly where is this asset, continuously", and it does so with an accuracy no other mainstream warehouse technology can match. That precision is real and genuinely useful. It also comes with an infrastructure bill and a design effort that only a minority of use cases actually justify. Knowing which side of that line you sit on is the whole decision.

1. What UWB positioning is

Ultra-wideband is a short-range radio technology defined under the IEEE 802.15.4 family of standards, and specifically the 802.15.4z amendment that added the secure, high-precision ranging capability the warehouse world cares about. The name describes the physical trick that makes it work: instead of transmitting on a narrow slice of the radio spectrum the way Wi-Fi or Bluetooth do, UWB spreads its signal across a very wide band, typically several hundred megahertz, using extremely short pulses measured in nanoseconds or less.

That wide band and those short pulses are not a marketing detail; they are the reason UWB can measure distance so precisely. A short pulse has a sharply defined arrival time, and a sharply defined arrival time can be measured with very little ambiguity. Narrowband technologies, by contrast, rely on measuring signal strength to estimate distance, and signal strength in a warehouse is a noisy, unreliable thing, bouncing off racking, absorbed by inventory, distorted by metal. UWB largely sidesteps that problem by measuring time rather than power.

In practical terms, a UWB system consists of two kinds of device. Fixed anchors are mounted at known positions around the facility, usually high on walls or columns with clear lines of sight into the operating area. Mobile tags are attached to the assets you want to locate, forklifts, pallets, tools, people. The anchors and the tag exchange precisely timed radio messages, the system works out how far the tag is from each anchor, and from those distances it calculates the tag's position in the building. The output is a live coordinate, updated many times per second, accurate to roughly ten to thirty centimetres in a well-designed installation.

2. How UWB works

The core idea is triangulation by distance, or more precisely multilateration. If you know how far a tag is from three or more anchors whose positions are fixed and known, there is only one point in space consistent with all of those distances, and that point is the tag. The clever part is not the geometry, which is centuries old, but how UWB measures those distances so accurately: it times how long the radio signal takes to travel between anchor and tag, and multiplies by the speed of light.

The diagram below shows the arrangement. Anchors on the walls each measure the time-of-flight to the tag, the system converts each time into a distance, and the intersection of those distance rings pins the tag to a single point on the floor.

There are two common ways the timing is done, and the difference matters for how you install the system. In time difference of arrival (TDOA), the tag simply transmits a pulse and the anchors, which are all synchronised to a shared clock, each record when they receive it. The differences in arrival time between anchors define the position. This scales well because the tag only transmits, so battery life is good and you can track thousands of tags, but it demands very tight clock synchronisation across all anchors, which is the hard engineering part.

In two-way ranging (TWR), the tag and each anchor exchange messages back and forth, and each pair measures the round-trip time directly. This removes the need for a shared clock across anchors, which simplifies installation, but it consumes more tag battery and does not scale to huge tag counts as gracefully. Most warehouse deployments choose between these two based on how many tags they must track and how much they want to invest in anchor synchronisation.

3. Time-of-flight and centimetre accuracy

The reason UWB reaches centimetre accuracy while other radio technologies struggle to reach metres comes down to a simple physical fact: radio waves travel at the speed of light, roughly thirty centimetres per nanosecond. To locate an asset to within ten centimetres, you need to measure the signal's travel time to within about a third of a nanosecond. That is an extraordinarily short interval, and it is only measurable because UWB's short, sharp pulses have an unambiguous arrival edge that a receiver can timestamp with that kind of resolution.

Compare that to signal-strength-based approaches, which is how BLE and Wi-Fi positioning estimate distance. Signal strength falls off with distance in theory, but in a real warehouse it is corrupted by everything in the environment: a steel rack, a stack of canned goods, a passing forklift, a person standing between transmitter and receiver. The result is that strength-based distance estimates wander by metres. Time-of-flight does not care about how strong the signal is, only when its leading edge arrives, and that leading edge is far more robust to the clutter of an operating warehouse.

The insight worth keeping: UWB measures time, not power. Every other mainstream indoor radio locator measures power and infers distance from it, which is why they top out at room-level or zone-level accuracy. By measuring the actual flight time of a signal that travels at a known speed, UWB converts distance into a physics problem with one answer instead of a statistics problem with a wide error bar. That is the entire reason it is precise, and it is why no amount of clever software makes a signal-strength system match it.

The honest qualifier is that centimetre accuracy is a best case, achieved with good anchor geometry, clear lines of sight, and enough anchors covering each area. Multipath, where the signal bounces off surfaces and arrives by more than one path, still degrades UWB, though far less than it degrades narrowband methods, because the receiver is designed to pick the first-arriving direct path. Where lines of sight are blocked or anchor coverage is thin, real-world accuracy degrades from a few centimetres toward tens of centimetres, which is still dramatically better than any alternative.

4. UWB versus BLE versus GPS

The three technologies people reach for when they want to locate things are ultra-wideband, Bluetooth Low Energy beacons, and satellite positioning. They are not competitors so much as tools for different jobs, and the table below lays out where each one belongs. The right way to read it is not "which is best" but "which matches the precision my use case actually needs against the cost I am willing to carry."

The pattern the table reveals is that these three technologies barely overlap. GPS owns the outdoors and is worthless the moment an asset moves inside the building, which is exactly why the GPS versus indoor tracking comparison exists as a decision point. BLE is the low-cost workhorse for knowing roughly where something is, presence and proximity at zone granularity, and the BLE asset tracking guide covers where its economics win. UWB is the precision instrument you reach for only when zone-level is not good enough and you genuinely need to know where an asset is to within a hand's width.

5. Use cases: precise asset location, forklift positioning, safety zones

UWB earns its cost in a specific set of scenarios where centimetre precision translates directly into money saved or accidents avoided. Three stand out in warehouse and industrial operations.

Precise real-time asset location. In a large distribution centre, high-value or fast-moving assets, roll cages, returnable containers, specialised tooling, expensive test equipment, have a habit of disappearing into the wrong aisle. Zone-level tracking tells you an asset is somewhere in a two-thousand-square-metre area, which is barely more useful than not knowing. UWB puts a dot on a floor plan accurate enough to walk straight to it. On assets that are searched for frequently, the labour saved from eliminating hunt time adds up quickly, and the reduction in lost or shrinkage-prone equipment is often the larger prize.

Forklift and vehicle positioning. Knowing exactly where every forklift is, at all times, unlocks several things at once: traffic-flow analysis to reduce congestion, automatic task assignment to the nearest available truck, and above all collision avoidance. Because UWB updates position many times per second with low latency, it can drive real-time proximity warnings between vehicles and between vehicles and pedestrians, which zone-level systems are simply too coarse and too slow to do reliably.

Safety zones and geofencing. This is where UWB's precision becomes a safety-critical capability rather than a convenience. You can define virtual boundaries, a robot's operating envelope, the swing radius of a crane, a restricted dock area, and trigger an alert or an automatic slow-down the instant a tagged person crosses into a hazard zone. A safety geofence is only as trustworthy as the accuracy of the positioning behind it; a system that is uncertain by several metres cannot safely define a two-metre exclusion zone, but a system accurate to centimetres can. This is why UWB has become the backbone of many worker-safety and human-robot collaboration deployments.

These use cases sit alongside, not instead of, the identification layer that most warehouses already run. UWB tells you where an asset is; it does not replace the item-level identification and inventory transactions that RFID in warehouse management handles, nor the transactional backbone of the warehouse management system that records what happened. The strongest deployments treat UWB as a precision positioning layer feeding location context into the WMS and safety systems, not as a standalone island.

6. The cost and infrastructure question

This is the section that should come before the purchase order, because UWB's precision is inseparable from its infrastructure demand, and that demand is where most of the total cost lives. Unlike a barcode, which needs only a printed label, or a BLE beacon, which you scatter loosely and calibrate roughly, UWB needs a designed, dense, cabled network of anchors to work at all.

A realistic UWB installation carries several distinct cost lines:

• Anchor hardware and density. You need enough anchors that every point in the tracked area is within range of at least three or four with good geometry. In practice that means a grid, often one anchor every ten to twenty metres, which across a large warehouse is a lot of units. Anchor count, not tag count, usually dominates the hardware bill.
• Cabling and power. Anchors are typically mounted high and need power and often data cabling, frequently Power over Ethernet. Running that cabling across an operating warehouse is a real installation project with real labour, not a plug-and-play afternoon.
• Synchronisation infrastructure. The tight clock synchronisation that TDOA needs is its own layer of equipment and configuration, and getting it right is the difference between centimetre accuracy and disappointing results.
• Tags and their batteries. UWB tags cost meaningfully more than BLE tags or RFID labels, and active tags need batteries that must be managed and replaced across the fleet.
• Software, integration and calibration. The positioning engine, the floor-plan mapping, the site survey and calibration, and the integration into the WMS or safety system are all effort lines that a hardware quote conveniently understates.

The honest limitation: UWB does not scale cheaply across a whole facility the way BLE does. Because accuracy depends on anchor density and clear lines of sight, covering a large, high-racked warehouse to full centimetre precision can require a dense and expensive anchor grid, and the racking itself blocks signals and complicates coverage. Many organisations that specify UWB building-wide discover late that the anchor count, the cabling and the site survey cost several times the tag budget they first estimated. The technology delivers exactly what it promises; the surprise is always the infrastructure, not the accuracy.

7. Where UWB is worth it and where it is overkill

The discipline that separates a successful UWB project from an expensive disappointment is honest matching of precision to need. UWB is worth its cost when three conditions hold together: you genuinely need centimetre or near-centimetre accuracy, the value of that precision is high, and the area requiring it is bounded rather than the entire building.

Where it is clearly worth it: safety-critical geofencing around robots, cranes and automated equipment, where an accurate exclusion zone prevents injury. Forklift collision avoidance in busy mixed-traffic areas. High-value asset tracking where the cost of a lost or endlessly-searched-for asset dwarfs the tag cost. Automated processes, such as guiding automated vehicles or verifying that the right item was placed at the right precise location, where the process itself demands to know position to centimetres. In each of these the precision is not a luxury; the use case does not function without it.

Where it is overkill: if all you need to know is which zone or which room an asset is in, UWB is a hugely expensive way to answer a question BLE answers for a fraction of the cost. If you only need to confirm that an asset passed a checkpoint or entered the building, RFID or a simple barcode scan does the job. If the asset lives outdoors in the yard, GPS is the right tool and UWB would need infrastructure you should not build outside. And if you want to track everything, everywhere, to centimetres, the cost curve almost always says pick the critical zones and accept coarser tracking elsewhere.

The framing I use with clients weighing this is a simple triage. Ask what the worst outcome is if the location is off by two metres. If the answer is "a worker could be injured" or "an automated vehicle could collide" or "a five-figure asset stays lost", UWB earns its place in that zone. If the answer is "someone walks an extra aisle to find it" or "we know it is in the building", the precision is not worth the infrastructure and a lighter technology wins. Run that test zone by zone rather than building-wide, and UWB usually turns out to belong in a handful of high-stakes areas, not everywhere, and the project stays affordable precisely because it is targeted.

8. References

For readers who want to go to the primary sources, two are worth naming. The IEEE 802.15.4 standard, and in particular the 802.15.4z amendment, defines the physical layer and the enhanced, secure ranging capability that gives UWB its precision and its resistance to relay attacks. Anyone specifying UWB hardware should confirm that the devices implement the 802.15.4z ranging features rather than an older, less capable UWB physical layer.

On the interoperability and certification side, the FiRa Consortium is the industry body driving standardised UWB use cases and device interoperability, so that tags and anchors from different vendors can work together rather than locking a buyer into a single supplier's ecosystem. When evaluating a UWB platform, checking for alignment with the FiRa Consortium's specifications is a reasonable proxy for whether the system will interoperate and remain supportable over time. Both of these are the standards worth reading before committing to a vendor; the specifics of any given product should always be verified against its own current documentation.

Final thoughts

Ultra-wideband is the precision instrument of the indoor location toolkit. By measuring the flight time of very short radio pulses between fixed anchors and a mobile tag, it pins an asset to within centimetres in an environment where GPS has no signal and signal-strength methods wander by metres. That accuracy is not marketing; it is a direct consequence of measuring time instead of power, and it is genuinely transformative for safety geofencing, forklift collision avoidance and precise real-time asset location.

The reason UWB disappoints when it disappoints is never the accuracy. It is the mismatch between the precision bought and the precision needed, and the infrastructure cost that a targeted deployment absorbs comfortably but a building-wide one does not. Point UWB at the zones where centimetres genuinely matter, feed its position data into the WMS and safety systems rather than an isolated dashboard, and be ruthless about using cheaper technologies everywhere the precision is not worth the anchor grid. Do that and UWB delivers exactly what it promises. For the wider picture of how identification and positioning fit together across the automated warehouse, return to the warehouse automation complete guide.

Related reading: Warehouse automation: the complete guide, BLE asset tracking, GPS vs indoor tracking, RFID in warehouse management, What is a WMS.