Warehouse RTLS in 2026: Where Channel Sounding Ends and UWB Begins
A tiering guide for warehouse asset tracking: which loads Bluetooth 6.0 channel sounding can carry, where UWB is still mandatory, and how to run both.
Hayat AminPresident of IP, Position Imaging
5 min read
In a warehouse, Bluetooth 6.0 channel sounding ranges to roughly 0.5 to 1 meter in clear aisles and degrades to 1 to 3 meters around steel racking, while UWB holds 10 to 30 centimeters in the same building. The gap is bandwidth: about 80 MHz versus 500 MHz. So assign channel sounding to aisle-level and zone-level tracking of low-cost tags, and reserve UWB for pallet-slot resolution, forklift safety zones, and anything that triggers automation.
Key takeaways
- Channel sounding hops across roughly 80 MHz of the 2.4 GHz band; UWB per 802.15.4z uses 499.2 MHz channels. That 6x bandwidth gap sets the accuracy ceiling.
- Around steel racking, expect 1 to 3 meters from channel sounding and 10 to 30 centimeters from UWB, not the vendor bench numbers.
- A pallet position is about 1 by 1.2 meters. Slot-level answers need error under half a meter, which means UWB.
- Aisle-level questions, which is most of warehouse tracking, sit comfortably inside channel sounding's error budget at BLE tag prices.
- The durable 2026 architecture is tiered: channel sounding for the bulk of tags, UWB islands where precision pays for itself.
- The ranging and fusion math is heavily patented. Licensing granted IP such as US 11,774,249 is faster than reinventing it.
What accuracy do the two radios actually hold on a warehouse floor?
Bluetooth 6.0, published by the Bluetooth SIG in September 2024, added channel sounding: phase-based ranging combined with round-trip time, hopping across up to 72 channels that span roughly 80 MHz of the 2.4 GHz band. On a bench, silicon like Nordic's nRF54 family produces line-of-sight errors in the 10 to 50 centimeter range. On a warehouse floor with 8-meter steel racking, loaded pallets, and moving forklifts, working accuracy settles at 0.5 to 1 meter in open aisles and 1 to 3 meters deep in the racking.
UWB under IEEE 802.15.4z tells a different story. Its 499.2 MHz channels, typically channel 5 at 6.5 GHz or channel 9 at 8 GHz, carry pulses around 2 nanoseconds long. Chipsets like Qorvo's DW3000 series and NXP's Trimension parts hold 10 to 30 centimeters in the same building, including many non-line-of-sight cases.
Neither number is marketing. Both follow directly from bandwidth. Bandwidth is the whole story here.
Why steel racking hurts channel sounding more than UWB
Range resolution for any RF ranging system is approximately the speed of light divided by twice the bandwidth. For channel sounding's 80 MHz aperture that works out to about 1.9 meters. For UWB's 499.2 MHz channel it is about 30 centimeters.
That single number explains warehouse behavior. Steel uprights, wire decking, and shrink-wrapped metal goods are efficient reflectors, so every ranging exchange arrives as a bundle of paths: the direct one plus reflections a few meters longer.
- Channel sounding cannot separate paths closer than roughly 1.9 meters, so the phase estimate blends the direct path with rack reflections. The result is a biased range that drifts as forklifts move and reflection geometry changes.
- UWB resolves paths about 30 centimeters apart, so the receiver can find the leading edge of the direct pulse and discard the reflections behind it.
Both radios see the same multipath. Only one can take it apart. Steel decides which radio you need.
Where channel sounding earns its place: aisle and zone tracking
Most warehouse location questions are not slot questions. Which aisle is the roll cage in? Did the returnable container leave through dock door 4? Is the pallet jack on the mezzanine or the ground floor? A 1 to 3 meter error answers all of these, because aisles are typically 3 meters wide or more and zones are tens of meters across.
Channel sounding wins this tier on economics, not accuracy:
- Single-radio tags. The ranging runs on the same 2.4 GHz radio as ordinary BLE, so a tag is one chip, one antenna, one coin cell lasting years at a few ranging exchanges per minute.
- Anchor reuse. Existing BLE gateway infrastructure upgrades to sounding-capable silicon rather than being ripped out.
- Handheld ecosystem. As sounding reaches phones and scanners, every worker device becomes a mobile anchor.
For the thousands of low-value assets that only need a zone answer, this tier is hard to beat. Which aisle is a channel sounding question.
Where UWB stays mandatory: slot resolution and safety
A standard pallet position is roughly 1 by 1.2 meters, and a typical rack bay holds three of them side by side. Telling adjacent slots apart requires error comfortably under half a meter, so this is UWB territory by arithmetic, not preference.
The same logic covers the warehouse's automation and safety layer:
- Putaway verification. Confirming a pallet landed in the assigned slot, not the one beside it, needs 10 to 30 centimeter confidence.
- Forklift proximity warning. Alert zones a few meters across collapse into false alarms if the position estimate itself wanders by 2 meters.
- Robot handoff. An AMR receiving a pick location tolerates decimeter error, not meter error.
Rule of thumb: if a machine acts on the coordinate, or a person's safety depends on it, the coordinate must come from UWB.
Two-way ranging or TDoA under 802.15.4z both work here; the choice is a density and battery tradeoff, not an accuracy one. Machines act on UWB numbers.
How do you run both radios in one RTLS?
The tiered design that holds up in 2026 puts one location engine over two radio planes.
- Tag classes by asset economics. The bulk of tags, on totes, cages, and returnable packaging, are channel sounding. UWB tags go on the minority of assets where slot resolution or safety justifies the higher tag cost and second radio.
- UWB as islands, not a blanket. Dense UWB anchor coverage at docks, staging lanes, high-value cages, and automated cells; channel sounding coverage everywhere else. Dual-radio anchors keep cabling to one run per point.
- One coordinate frame. The location engine fuses both streams into the same floor map, weighting each fix by its expected error, so the WMS never knows or cares which radio produced a position.
This structure also protects the roadmap. If sounding silicon improves or a workflow later demands precision, you reassign tag classes instead of re-architecting. Tier the radios, unify the map.
What licensing proven positioning IP changes for your build
The hard part of a two-radio RTLS is not the silicon, which Nordic, Qorvo, and NXP will happily sell you. It is the layer above: extracting reliable range from multipath-corrupted measurements, fusing two radio planes of different accuracy into one trustworthy track, and doing it in a field where the core techniques were patented years before Bluetooth 6.0 shipped.
Position Imaging licenses hundreds of granted patents in real-time positioning, radio-frequency ranging, computer vision, and machine learning, a portfolio cited by Apple, Bosch, and other major firms. Granted patents including US 11,774,249, US 12,079,006, US 12,066,561, and US 12,000,947 sit in exactly this tracking-system layer, and they apply whether your tags speak channel sounding, UWB, or both.
Licensing does two things a from-scratch build cannot: it gives you working, field-proven methods you can ship against in months, and it gives you freedom to operate in a portfolio-dense field. License the ranging math, ship the product.
Frequently asked questions
Is Bluetooth 6.0 channel sounding accurate enough for warehouse asset tracking?
For aisle-level and zone-level tracking, yes. Expect 0.5 to 1 meter in open aisles and 1 to 3 meters deep in steel racking, which cleanly answers which aisle, zone, or dock door an asset is in. It is not enough to distinguish adjacent pallet slots, which are about 1 meter apart.
Why is UWB more accurate than BLE 6.0 ranging?
Bandwidth. UWB channels under 802.15.4z are 499.2 MHz wide, giving roughly 30 centimeter path resolution, so the receiver can isolate the direct pulse from rack reflections. Channel sounding spans only about 80 MHz of the 2.4 GHz band, so reflections within roughly 1.9 meters of the direct path blend into the estimate and bias it.
Should I deploy channel sounding or UWB for a new warehouse RTLS in 2026?
Usually both, tiered. Put channel sounding tags on the bulk of assets that need zone answers, because the tags are single-radio and cheap. Put UWB tags and dense anchors where slot-level accuracy or safety matters, such as putaway verification, staging lanes, and forklift proximity zones. Run both through one location engine and one floor map.
Can one anchor network handle both channel sounding and UWB?
Yes, with dual-radio anchors that carry a 2.4 GHz sounding-capable BLE chip and an 802.15.4z UWB chip on one mount and one cable run. UWB anchors need higher density in precision zones, so most designs blanket the building with sounding coverage and add UWB islands where the accuracy pays for itself.
Does choosing a radio protocol protect me from positioning patent risk?
No. The dense patenting in this field sits largely above the radio, in ranging, multipath handling, tracking, and sensor fusion methods that apply to channel sounding and UWB alike. Licensing granted IP, such as Position Imaging's RF ranging and tracking portfolio, addresses that layer directly and is typically faster than designing around it.
Send us your tag classes and accuracy targets, and we will map your warehouse RTLS design against the portfolio in one working session.
Tell us the product. We map the exact scope, what a license covers, and how fast you can ship, all in a 20-minute call.
Book a 20-minute call