Two technologies dominate enterprise RTLS for indoor positioning today: UWB and BLE 5.x with direction-finding. Both work. Both have credible vendors. The decision is almost never about which is "better" — it's about which is appropriate for what you're trying to do, at what budget.
What you get from each
UWB (Ultra-Wideband, IEEE 802.15.4z, FiRa-certified). Uses time-of-flight ranging across multiple anchors. Typical accuracy 10–30 cm in a well-designed deployment. Latency is low (sub-100 ms). Multi-path resilient. Currently the only commodity radio that hits sub-metre indoor accuracy at scale.
BLE 5.x with AoA (Angle-of-Arrival). Uses multi-element antenna arrays on the locator to determine the bearing of a transmitter. Typical accuracy 1–3 m. Standard BLE without AoA is RSSI-only and gives 3–10 m room-level — not the same product.
The price of the extra accuracy
Going from BLE-AoA to UWB roughly triples per-square-metre infrastructure cost. The numbers vary by environment, but as a working rule of thumb across our recent deployments:
- BLE 5.x AoA: one locator per 200–400 m², ~€200–€500 per locator, €20–€30 per tag.
- UWB: one anchor per 50–120 m², €400–€1,000 per anchor, €40–€80 per tag.
Then add cabling, PoE switches, calibration time, and the engineering hours for proper anchor placement. UWB is the more expensive answer everywhere.
When sub-metre accuracy actually matters
The honest answer is: less often than vendors imply. Use cases where you need UWB-grade precision:
- Tool-on-job traceability. Which torque tool is on which part on which line? Workflows with overlapping work envelopes need bay-level discrimination.
- Anti-collision for AGV/AMR. Robot safety zones don't tolerate 2 m uncertainty.
- FOD prevention in aerospace. "Tool last seen near aircraft" is a different problem to "tool last seen in hangar."
- Athlete tracking. Training-load analytics need true biomechanics-grade location.
- High-value asset tracking in shared spaces. Two infusion pumps on adjacent beds need to be distinguishable from each other, not just "in the same ward."
- Process automation interlocks. Triggering an action when an asset enters a specific machine envelope.
When BLE 5.x is the right answer
- Room-level or zone-level workflows. "Pump is in ED bay 7" is enough to drive nurse-call workflow.
- Occupancy and hot-desking. 1–3 m is excellent for hybrid-workplace analytics.
- Wayfinding for staff and visitors. Phone-based RTLS using BLE infrastructure scales beautifully.
- Large estates. If you need to cover 50,000 m² of floor space, BLE economics work and UWB economics often don't.
- Battery-life-critical tags. BLE tags routinely run 3–5 years on a coin cell.
The hybrid pattern
Most of our larger programmes are not all-UWB or all-BLE. They are BLE-AoA across the broad floor area, with UWB infrastructure overlaid in a few high-precision zones — assembly cells, ORs, dispensary, robot lanes. Tags either dual-radio or, more commonly, distinct populations of tags per zone.
This hybrid pattern is cheaper than all-UWB and more useful than all-BLE. It's also where having a vendor-neutral architecture pays back: most large RTLS vendors push you toward whichever radio they make.
One more variable: where the analytics live
Increasingly the decision isn't about the radio at all but about the analytics layer downstream. UWB platforms tend to bundle proprietary engines that are excellent for visualisation but harder to integrate. BLE-AoA platforms have tended toward more open APIs.
Ask the vendor: "If we wanted to feed the position event into our own data pipeline, can we?" If the answer is "yes, via REST or MQTT in a documented schema," that vendor will scale with your programme. If the answer is "you can use our dashboard," it will not.
Quick decision tree
- Do you need to distinguish two assets within 1 m of each other? → UWB.
- Do you need to drive a process / safety interlock? → UWB.
- Is your area > 20,000 m² with a flat budget? → BLE.
- Is your accuracy requirement "which room / which bay"? → BLE.
- Do you need both at different zones? → Hybrid — that's the answer most often.
Want help running this decision against your actual environment? TRACIO technology-selection engagements include site-survey RF modelling and TCO under both architectures.