Blog

UWB in 2026: No, Bluetooth 6.0 Didn’t Kill It — Centimeter Accuracy Just Got Real

Posted on by Support

Ask anyone who follows wireless tech what’s going to replace UWB, and half of them will say Bluetooth 6.0’s new Channel Sounding feature. The logic sounds clean: if Bluetooth can now do accurate ranging, why pay for a separate UWB radio? It’s a fair question. It’s also one you should stop asking after about five minutes in an actual factory, warehouse, or hospital — because the answer shows up fast when you watch the two technologies do their jobs side by side.

UWB didn’t disappear in 2026. It got cheaper, got regulatory room to breathe, and quietly became the default answer for the jobs where a few centimeters genuinely matter. Here’s what’s actually happening, and where KKM’s hardware fits in.

Why the “Bluetooth 6.0 kills UWB” story never quite lands

Channel Sounding (CS) is real, and it’s genuinely useful. It lets two Bluetooth devices measure distance to each other — no phone app, no special infrastructure. That’s why it showed up in the spec, and it’s going to do nice things for finding lost keys and unlocking doors as you walk up to them.

But here’s the part that gets skipped in the hype: CS measures a distance between two devices. Industrial tracking doesn’t work that way. A forklift in a warehouse isn’t asking “how far is that one tag?” — it needs to know where it is on a floor plan, continuously, dozens of times a second, in a building full of metal racks, moving stock, and radio noise. That’s a positioning system problem, not a ranging problem.

UWB has spent over a decade being built for exactly that. The FiRa Consortium certification ecosystem, the omlox open standard for locating, chipsets designed for sub-10-centimeter performance with low latency — that infrastructure didn’t exist for CS yesterday and it won’t exist tomorrow. If you’re picking a technology for a safety-critical deployment that has to run for five years, “the spec is new and interesting” is not a procurement strategy.

The 2026 numbers worth citing

Let’s put some actual figures on the table.

The overall UWB market is projected to hit roughly $1.54 billion in 2026, climbing to about $3.81 billion by 2031 — a hair under 20% compound growth (Mordor Intelligence). Not hockey-stick crazy, but steady, and steady is what infrastructure buying looks like.

Two details inside that report tell the real story:

  • Real-time location systems (RTLS) are already the biggest application slice, at around 42% of revenue. Phones still ship the most UWB hardware, but the money is moving to industrial tracking (Mordor Intelligence).
  • The whole RTLS market is growing even faster than UWB itself — from $5.2 billion in 2023 to a projected $16.2 billion by 2028 (FiRa Consortium). UWB is riding that wave because it’s the accuracy layer the other technologies can’t match.

What moved the needle in 2025-2026 specifically? Three things.

Regulators finally got out of the way. The EU’s Decision 2024/1467 raised indoor transmit power limits and opened up outdoor operation in the 6-8.5 GHz band, effectively doubling usable range. China’s MIIT harmonized the 7163-8812 MHz band in mid-2025 and cut certification time dramatically. When the two biggest markets on Earth make UWB easier to deploy on the same week, deployments follow.

Cars dragged UWB into the mainstream. The Car Connectivity Consortium’s Digital Key 3.0 spec mandates UWB secure ranging, and it’s shipping in 2026 model-year vehicles from BMW, Mercedes-Benz, and Hyundai. Relay attacks on keyless entry are the whole reason carmakers went this way — UWB measures time of flight, so a thief with two radios can’t stretch your key fob’s signal into a fake range. Automotive money paid for a lot of UWB engineering that industrial users now inherit.

The silicon got cheap. Module prices fell hard — Qorvo’s DW3000-series modules dropped around 40% between 2023 and 2025, and volume pricing now sits below a few dollars (Mordor Intelligence). That’s the number that turns “nice technology” into “let’s spec it.”

The honest line: don’t buy centimeters you don’t need

Here’s where somebody at KKM has to be straight with you, because it’s in their interest to sell you the right thing, not the expensive thing.

BLE angle-of-arrival (AoA) systems have gotten good — real-world deployments are quoting around 50-centimeter accuracy at a fraction of UWB’s silicon cost. If your problem is “which aisle is this pallet in?” or “is the wheelchair on floor two?”, BLE AoA is often the smarter buy. It’s why KKM makes a full line of Bluetooth beacons and gateways too, and why their Bluetooth Low Energy solution page gets as much traffic as the UWB one.

Buy UWB when the answer needs to be exact, not close:

  • Forklift-pedestrian proximity zones where a false “clear” is a safety incident. Operators tracking vehicles with sub-10-centimeter accuracy have reported forklift-collision insurance claims dropping 30-40% (Mordor Intelligence).
  • Tooling and fixtures that must land in the right rack position, not the right corner of the room — think aircraft assembly or high-mix manufacturing lines.
  • Autonomous mobile robots docking, charging, or handing off loads. A robot that docks 30 centimeters off-center does damage; one that docks to the centimeter doesn’t.
  • Secure access and lone-worker alerts, where the ranging has to be tamper-resistant, not just accurate.

Centimeter accuracy is a tool, not a trophy. The trick is knowing which jobs need it.

KKM’s UWB hardware — and the line they deliberately won’t cross

KKM Smart Solutions — the Shenzhen hardware house with 16+ years in RF, 50+ patents, and an ISO9001-certified factory — keeps its UWB line deliberately small and deliberately open. Three devices, each aimed at a specific job:

  • K4W UWB asset tag — for things. An IP67 industrial tag with a replaceable battery, 10-30 cm typical accuracy via TDoA, OTA firmware updates, and a Qorvo DW-series UWB chipset paired with a Nordic nRF52833 for the BLE side. It also speaks Apple’s Nearby Interaction protocol, which matters more than it sounds — see below.
  • K7W UWB badge — for people. Same positioning core in a wearable ID badge, IP67-rated, with an SOS button, a rechargeable battery built for a full shift, tag-to-tag ranging for worker-interaction zones, and coordinates refreshed every 500 milliseconds. Programmable LEDs and buttons on the badge give your software team physical outputs to hang alerts on.
  • C3W rechargeable UWB badge — a variant with magnetic charging for operations where docking stations are easier than swapping cells.

Two things about this lineup are worth underlining, because they decide whether a deployment thrives or stalls.

First, the chipsets and radios are the industry-standard parts. Qorvo DW-series silicon is the workhorse of industrial UWB, and it means your integrator isn’t debugging against a proprietary radio. The firmware and schematics are documented so your engineers can build on them instead of around them.

Second, and this is the big one: KKM does not sell you a cloud. No forced platform, no proprietary anchor you have to buy from them, no data that has to stop by their servers. The UWB page says it plainly — KKM provides schematics for clients developing their own firmware, and their tags work with your anchors. What you get from KKM is hardware, firmware, API/SDK documentation, and OEM/ODM support. Your positioning engine, your alert rules, your dashboards, your WMS integration — that’s your software, your IP, and it lives in your stack. That division of labor is exactly why integrators and end customers with existing systems keep coming back.

Real world: the sequencing center that counted near-misses

Let’s make it concrete. Picture Südring Automotive Logistics, a sequencing hub that feeds a car assembly plant in southern Germany. They’re the operation that makes sure the right seat, the right dashboard, the right wiring harness shows up at the right assembly line bay, in the right order, every few minutes. Their floor looks like a busy city: tugger trains pulling kitting carts, forklifts feeding line-side racks, and hundreds of people walking between all of it.

Their problem wasn’t finding things. It was the quiet daily grind of near-misses — a forklift turning into an aisle where a picker was walking, a cart left in the wrong bay that cost a line stoppage, and an incident log that relied on people remembering to report what almost happened.

Their setup, all hardware plus software they already ran:

  • K4W tags mounted on the forklifts and tugger trains — the vehicles became precisely trackable points, not “somewhere in zone C.”
  • K7W badges on the floor staff — every picker and driver carries one; the SOS button doubles as a way to call for help instead of shouting over forklift noise.
  • Anchors the client already owned, mounted on pillars and racking across the floor, feeding positioning data to the safety-and-ERP platform the plant’s own software team had built over years.

Here’s the division of labor in practice: KKM supplied the tags, badges, and integration documentation. The client’s team wrote the rules — “forklift within 1.5 meters of a pedestrian badge moving toward it at walking speed equals a proximity alert” — into software they already owned and already trusted.

After the first two quarters, the numbers they shared internally told the story: near-miss reports stopped depending on memory because the system logged proximity events automatically. The safety team finally had a real incident funnel instead of anecdotes. And the line-stoppage from a misplaced kitting cart — the one that used to cost them a shift’s production every couple of weeks — became a thing they could catch in the act and correct in minutes.

Nothing in that story is exotic. That’s exactly the point. It’s a tough tag on a forklift, a badge on a person, and software the client already had. The hardware just had to be accurate enough to trust, rugged enough to survive a factory floor, and open enough that nobody had to call the vendor to make it work.

What to check before you buy UWB in 2026

If you’re spec’ing a UWB deployment this year, run this list:

  1. Who owns the anchor side? If the vendor forces their anchors and their cloud on you, you’re renting your deployment forever. KKM’s model — tags that work with your anchors, schematics for your firmware — is the one that ages well.
  2. What does the accuracy spec assume? “10-30 cm” assumes reasonable conditions. Metal racking, wet stock, and forklift masts eat into that. Ask what the vendor’s hardware does in your environment, and plan anchor density accordingly.
  3. What’s the real update rate and battery math? A badge refreshing every 500 ms draws more than one updating every 5 seconds. Work out your duty cycle before you trust the “all-day battery” line.
  4. Is the silicon ecosystem-standard? Qorvo/NXP-class DW-series or FiRa-conformant silicon means your integrator has a fighting chance. Proprietary radios mean you’re the beta tester.
  5. Where does the data go? Confirm the tags hand coordinates to your positioning engine over documented interfaces — and that nothing in the chain quietly requires a vendor subscription you’ll outgrow.

Bottom line

UWB’s 2026 story isn’t about beating Bluetooth 6.0 in a benchmark shootout. It’s about a market finally maturing: regulators opened the spectrum, automakers paid for the engineering, module prices collapsed, and the use cases sorted themselves into “close enough” (BLE) and “exact” (UWB). The companies winning with it are the ones that picked the technology for the job and kept their software under their own roof.

KKM has been that kind of vendor for 16 years — hardware that’s certified, documented, and handed over with the keys. If your project needs to know where the forklift is to within a hand’s width, every half-second, in a building where radios go to die, that’s the kind of boring reliability you’re actually buying.

Sources

Support