How Does RTLS Work? Real-Time Location Systems Explained

A real-time location system finds where people and objects are inside a building, continuously, without anyone scanning anything. A tag transmits, fixed receivers hear it, and software turns what the receivers measured into a point or a zone on a floor plan. That last step is where the engineering lives: the same radio, measured differently, produces either 30 centimetres or 5 metres, and that gap is most of what a buyer gets wrong.

This guide covers how RTLS works, which measurement each technology uses, what accuracy each one achieves with a source attached to every figure, which ISO standards govern the field, and why RTLS projects fail.

How does RTLS work?

An RTLS has three parts. Tags transmit a signal, usually called a blink. Receivers at fixed, surveyed positions hear those blinks and record something measurable about them. Software converts those measurements into coordinates or a zone and hands the result to whatever system acts on it.

Tags come in two kinds. An active tag carries a battery and transmits on its own schedule: every second for a forklift, every few minutes for a pallet that rarely moves. A passive tag has no battery, harvests energy from a reader’s field and answers only when interrogated, which is why passive tags are cheap, last indefinitely, and only tell you about the moment something passed a reader.

Receivers go by several names: anchors in ultra-wideband, access points in Wi-Fi, readers in RFID, sensors in infrared. The job is identical. Each sits at a known location and measures the same property of every blink it hears, so the software can compare them.

The comparison is the whole system. Three receivers reporting when a blink arrived can solve for the position that explains all three arrival times. Three reporting only how loud it was can also solve for a position, but far worse, because loudness is not a reliable proxy for distance.

What are the four locating methods?

There are four families of measurement, and the accuracy you get depends more on which family you are in than on which radio you bought.

Time-based methods (time of arrival, time difference of arrival, time of flight, round-trip time) measure how long the signal took to travel. Radio moves at a known speed, so time converts directly to distance. This is the most accurate family, and it is why ultra-wideband leads: a UWB pulse is a couple of nanoseconds wide, so a receiver can timestamp its arrival far more precisely than a narrowband signal that ramps up over microseconds.

Angle-based methods use an antenna array to work out the direction a signal arrived from. Two bearings from two known points intersect at one position. Bluetooth 5.1 added direction finding to the core specification for exactly this purpose.

Signal-strength methods infer distance from attenuation: the weaker the signal, the further away it presumably is. Every radio reports signal strength for free, which makes this the cheapest approach and the least trustworthy. A human body absorbs 2.4 GHz energy. A steel rack reflects it. The same three metres can read as one or as eight.

Proximity does no distance calculation at all. If a reader heard the tag, the tag is near that reader. For “did this cage leave through dock 4”, that is the entire answer, and anything more sophisticated buys precision the question does not use.

How much location accuracy do you actually need?

Most of the money wasted on RTLS is spent buying coordinate-level tracking for a question that only needs zones. Four tiers of coverage, in ascending cost.

Choke point. Readers only at doorways, docks, lifts and corridor junctions. Tells you about transitions, never about position. Cheapest by a wide margin, because the infrastructure count follows the number of openings rather than the floor area.

Zonal. Which room, ward, bay or aisle something is in. Answers most asset-tracking questions ever asked.

Local area. Continuous X-Y coordinates inside a covered space. Needed when position within a room is what matters: a tool on a line, a vehicle in a bay.

Wide area. Coordinates across a whole site, often with an outdoor component and a handoff to satellite positioning at the door.

“Which floor is the infusion pump on?” is a zonal question. Specifying 30-centimetre coordinates to answer it is how an RTLS budget gets spent without the operational problem being solved.

What accuracy does each RTLS technology achieve?

Accuracy figures in this industry are rarely attributed. Checked again in September 2026, the pages ranking highest for general RTLS queries publish accuracy numbers without linking a primary source for any of them. Every figure below carries one, and separates laboratory from deployed performance, because those are different numbers.

Ultra-wideband at 10 cm comes from the Zafari et al. survey of indoor localization systems (IEEE Communications Surveys & Tutorials, 2019). The deployed 50 cm comes from an automotive plant study published in Sensors 2020, running eight anchors per 2,000 to 3,000 m².

Bluetooth 5.1 angle-of-arrival at 0.7 m was measured by Pau et al., Sensors 21(11):3589, 2021, with four anchors in a 25×15 m laboratory. The iBeacon proximity classes (immediate under 1 m, near 1 to 3 m, far beyond 3 m) and the 0.86 to 1.44 m experimental range are from the Zafari survey, which also reviews the Wi-Fi fingerprinting work that puts signal-strength positioning in the several-metre band.

Wi-Fi round-trip time, standardised as fine timing measurement in IEEE 802.11mc, is time-based and beats Wi-Fi signal strength for that reason. No number appears above because the figures in circulation trace to vendor documentation rather than a standards body or a peer-reviewed measurement.

The second group trades coordinates for other properties, and persists because those properties are sometimes the point.

The RFID and ultrasound figures are from the Zafari survey; some UHF passive systems reach about 20 m at a portal. Infrared’s room-level behaviour traces to the paper that founded the field: Want et al., “The Active Badge Location System”, ACM Transactions on Information Systems 10(1), 1992. The GPS figure is from gps.gov, the U.S. National Coordination Office for Space-Based Positioning, Navigation and Timing.

Two things these tables deliberately omit. Cost, because there is no independently verified per-tag or per-anchor pricing to publish, and an invented range would be worse than a visible gap. Battery life, because it varies with transmit interval so widely that a single number misleads: a tag rated for years at a one-minute blink interval may last months at one second. Ask for the figure at your own blink rate.

Why is infrared always room-level?

Because infrared cannot pass through a wall. That is physics, not product maturity, and no future version of the technology changes it.

The 1992 Active Badge paper described badges emitting pulse-width-modulated infrared every 15 seconds with a six-metre range, and observed that the signals are “reflected by partitions and therefore are not directional when used inside a small room” and “will not travel through walls, unlike radio signals that can penetrate the partitions found in office buildings.”

Read that as a limitation if you want coordinates, and as a guarantee if you need certainty that a badge is in this room and not the one next door. Radio rarely offers that: a 2.4 GHz signal passes through plasterboard, so a receiver next door may hear a tag that never left the room. Infrared’s blindness is why it persists in hand-hygiene compliance and clinical contact tracing.

Why is laboratory accuracy not deployed accuracy?

Laboratory figures are measured with clear line of sight and dense anchors. Deployed figures are measured in a building with metal, racking, people and an anchor budget. The honest planning number is always the deployed one.

The most useful thing in the tables above is a disagreement between two credible numbers for the same technology. The Zafari survey records ultra-wideband at around 10 cm. The automotive deployment in Sensors reported 50 cm with eight anchors per 2,000 to 3,000 m². Neither is wrong. They measure different situations, and a factor of five sits between them.

So when a vendor quotes an accuracy figure, three questions decide whether it means anything. What anchor density did it assume, per square metre of your floor area? Was it measured with line of sight, or with the tag on a trolley behind a person? And is it a mean or a 95th percentile? A mean of 40 cm with a long tail behaves very differently in an alerting workflow.

How is RTLS different from GPS?

They solve different problems. GPS is a one-way broadcast from satellites, excellent outdoors and undefined indoors. RTLS is local infrastructure installed in a specific building to answer questions inside it.

The U.S. government’s own gps.gov accuracy page states that GPS-enabled smartphones are “typically accurate to within a 4.9 m (16 ft.) radius under open sky”, that accuracy “worsens near buildings, bridges, and trees”, and lists satellite signal blockage, multipath reflection off buildings and walls, and “indoor or underground use” among its error sources.

Note what is absent: gps.gov publishes no indoor accuracy figure at all, because there is no specified indoor performance. Any page quoting “GPS is accurate to X metres indoors” has invented the number. What an indoor receiver reports is a position with no stated error bound.

What standards govern RTLS?

RTLS is standardised by ISO/IEC JTC 1/SC 31 in the ISO/IEC 24730 series. It matters at procurement, because it separates an interoperable system from one that locks your location data inside a single vendor’s platform.

Parts 21 and 22, both published in 2012, cover DSSS variants using single and multiple spread codes. Part 62 is based on IEEE 802.15.4a, the UWB physical layer most current hardware descends from.

Part 1 “defines a boundary across which application software uses facilities of programming languages to collect information contained in RTLS tag blinks received by the RTLS infrastructure.” It is the API that stops your location history being readable only by the system that produced it. The compliance rule worth taking into a vendor meeting: a conforming system implements Part 1 plus at least one air-interface part. “ISO 24730 compliant” without named parts tells you almost nothing.

What does the evidence show RTLS delivers?

Healthcare carries most of the published RTLS outcome literature, because that is where controlled studies have been funded and run. The numbers below are peer-reviewed and each links to its record.

Radiology scheduling time fell from 12 minutes to 5, a 58.3% reduction, with the idle rate falling from 16% to 12% and a data loss rate under 1%: Tseng WC et al., Radiography 2025;31(4):102999.

A metropolitan hospital tracking beds and equipment reported a 20.9% reduction in delivery time and 86.8% faster equipment search, with 91.2% staff satisfaction: Huang TL et al., Int J Qual Health Care 2025;37(1):mzaf005.

Porter turnaround fell from 28 to 20 minutes in one wing and 26 to 18 in another. Before deployment, 175 of 285 calls, or 62%, went unanswered: Kataria S et al., Cureus 2024;16(9):e69922.

Contact tracing sensitivity reached 60.0% with RTLS against 46.8% for conventional methods, adjusted odds ratio 6.15, 95% CI 1.92 to 28.69: Kim MH et al., Journal of Medical Internet Research 2022;24(10):e41395.

A 2021 systematic review of 42 articles in JAMIA concluded that RTLS “is a useful and effective adjunct methodology in process and quality improvement, workflow analysis, and patient safety”, while noting that “few studies have developed quantitative techniques to effectively analyze RTLS data.” Both halves belong in a business case.

Why do RTLS deployments fail?

Partial coverage, departmental ownership, accuracy oversold at procurement, and no change to the workflow the data was meant to improve. The research on failure is older than the research on success, and more useful.

A three-year qualitative study across 23 US hospitals, with 80 interviews, found “substandard functionality of most real-time location systems in use”, and concluded that the technology’s best use was asset tracking, deployed hospital-wide and managed centrally rather than departmentally: Fisher JA & Monahan T, International Journal of Medical Informatics 2012;81(10):705–12. The hardware has improved since 2012. Its conclusion about deployment shape has not aged.

The four patterns below are our reading of why deployments underperform, informed by that paper rather than quoted from it.

Partial deployment. Coverage in some departments and not others means staff cannot trust the map, so they go back to walking the corridor and looking. Value collapses non-linearly, because every search still begins with “is this area even covered?”

Departmental ownership. Fragmented tag inventories, nobody accountable for the position data being right, and three departments buying three incompatible systems in one building.

Accuracy oversold at procurement. A system specified against a laboratory figure and deployed at half that accuracy fails against the expectation it was sold on, not against physics.

No workflow change. Location data that does not alter a process changes nothing. If porter dispatch works the way it always did, knowing where the bed is saves nobody a minute.

You will also see the claim that nurses waste “$14 billion a year” searching for equipment. It traces to a 2023 review sponsored by an RTLS vendor, and the number is an extrapolation (minutes per shift multiplied by 1.7 million nurses multiplied by an hourly rate) rather than a measurement. Build the case on the peer-reviewed numbers instead.

Where is RTLS used?

In healthcare, manufacturing, logistics and warehousing, with healthcare carrying by far the strongest published evidence base.

Healthcare tracks infusion pumps, wheelchairs and telemetry units, runs staff duress badges, measures patient flow, handles infant security and wander management, and supports hand-hygiene compliance.

Manufacturing uses work-in-progress tracking, tool and fixture location, forklift collision avoidance, and analysis of where material actually accumulates versus where the layout drawing says it should. The Sensors study cited earlier used location clustering to find a production bottleneck and three unplanned storage areas that appeared on no drawing: the first value from an RTLS is often diagnostic rather than operational. Quantified manufacturing outcomes remain much thinner than healthcare’s.

Logistics and warehousing covers yard management, dock scheduling, pallet and cage tracking, and inventory accuracy, mostly at the choke-point and zonal tiers, because the objects are numerous and the questions are about transitions.

How do you choose an RTLS technology?

Start from the question, not the radio. Four decisions, in this order.

Reversing that order is the most common procurement error. A buyer starting at step 4 picks ultra-wideband because it has the best number, then discovers at step 1 that the real question was “which ward”, which a far cheaper tier answers.

One trade-off worth stating plainly: beacons are sometimes the right answer. If the question is zonal, the tag count runs into the thousands and the budget per tag is small, a Bluetooth beacon deployment at 1 to 3 m will answer it and a time-based system will not be approved. The failure is not choosing a cheap technology. It is choosing one and then writing a specification that assumes an expensive one.

Frequently asked questions

What does RTLS stand for?

RTLS stands for real-time location system. It describes any system that automatically determines and reports the position of tagged people or objects, continuously, without manual scanning. RTLS is an outcome rather than a single technology: ultra-wideband, Bluetooth Low Energy, Wi-Fi, RFID, infrared and ultrasound can each deliver it, at accuracies ranging from centimetres to several metres and at very different costs.

How accurate is RTLS?

It depends on the measurement method rather than the brand. Ultra-wideband reaches around 10 cm in laboratory conditions and about 50 cm in a working industrial deployment with eight anchors per 2,000 to 3,000 square metres. Bluetooth 5.1 direction finding measured 0.7 m with four anchors in a 25 by 15 m laboratory. Signal-strength methods on any radio give several metres at best.

Can RTLS work without any new hardware?

Only partially. Wi-Fi signal-strength positioning reuses existing access points and produces accuracy measured in metres, which supports coarse zoning. Wi-Fi round-trip time under IEEE 802.11mc is time-based and does better, but it requires access points and client devices that both implement it. Anything approaching a metre or below needs purpose-built infrastructure: ultra-wideband anchors, Bluetooth direction-finding arrays or infrared sensors, each surveyed into a known position.

What is the difference between RTLS and RFID?

RFID is one of several technologies an RTLS can use, not an alternative to it. Passive RFID reads at 1 to 2 m and tells you that a tag passed a specific reader, which suits portals and high tag counts. Active RFID reaches around 100 m of range but cannot achieve sub-metre accuracy. An RTLS built on RFID answers transition and presence questions; coordinate-level questions need time-based or angle-based methods.

Does ISO 24730 compliance mean two systems will interoperate?

Not on its own. A conforming system implements ISO/IEC 24730-1, the application programming interface, plus at least one air-interface part such as 24730-62 for high-rate pulse repetition ultra-wideband. A vendor stating “ISO 24730 compliant” without naming the parts has made an unverifiable claim. Ask which parts, and confirm that the air interface matches the one the other system uses.

Where do you start on your own site?

Start with the question the location data is supposed to answer, then the coverage tier, then the method, then the technology. The floor plan, the tag count and the decision the data is meant to change settle most of it.

Those three inputs are the ones nobody can supply from a web page. Book a scoping call and we will work through them against your building.