Indoor Positioning Systems: The Costs a Beacon Quote Leaves Out

An indoor positioning system demo is easy to like. The floor plan is clean, the blue dot glides, and whoever is holding the phone walks the route without a stumble. Nobody in the room asks what the thing will cost in month thirty.

The map is the interface. The positioning engine underneath it is the product. Both arrive on one quote, and that is where budgets go wrong. A map is a one-time build. A positioning layer made of battery-powered hardware is a standing commitment to ceiling access, maintenance tickets and re-surveys.

This article is about the second half of that bill, and about the cases where Bluetooth beacons are still the right choice. Per-unit prices are left out on purpose wherever the only published figures come from people selling the hardware. The standards, the legal deadlines and the physics are named and linked.

What is an indoor positioning system?

An indoor positioning system estimates where a person or an asset is inside a building, where satellite signals do not reach, and passes that estimate to an app, a map or a dashboard in near real time.

It has three parts, and they are usually priced as one. A sensing layer produces raw observations: radio signal strength, inertial motion, magnetic field readings, barometric pressure, camera frames. A positioning engine converts those observations into coordinates. A geometry layer holds the floor plans, the walkable space and the routing graph.

Three terms get used interchangeably and should not be. Indoor positioning answers “where am I”. Wayfinding is the routing and instruction layer built on top of it. A real-time location system tracks assets rather than people, often with tags instead of phones. A venue can buy excellent wayfinding software and still have a positioning engine that loses the floor.

Accuracy is also four numbers, not one: horizontal error at a stated percentile, correct-floor rate, latency, and the share of the venue where those hold. A vendor quoting a single metre figure has told you almost nothing.

Why doesn’t GPS work indoors?

Because the accuracy people expect from GPS is an open-sky number. The US government’s own performance page puts a GPS-enabled smartphone at roughly a 4.9 m radius under open sky (gps.gov, accessed September 2026). A roof removes that.

Satellite signals arrive indoors weakened and reflected. The reflections are the worse problem: a receiver that measures a bounced path computes a distance that is too long, and the resulting position can sit in the car park while the user stands at a lift lobby.

Vertical position is harder still, and regulators treat it as its own requirement. The Federal Communications Commission requires nationwide wireless carriers to deliver vertical location within ±3 m of a caller’s handset for 80% of indoor wireless 911 calls, a metric adopted in 2019 (fcc.gov, 911 location accuracy). Getting the floor right is not a refinement. In a hospital or a stadium it is the whole answer.

How does a Bluetooth beacon indoor positioning system work?

Beacons broadcast Bluetooth Low Energy advertising packets at a fixed interval. A phone measures the received signal strength of several beacons at once, and the engine converts those measurements into a position, either by trilateration or by matching them against a survey recorded in advance.

That survey is the part buyers underestimate. Fingerprinting walks the venue, records signal patterns at known points, and stores them as a reference. The positioning quality after go-live depends on how well today’s radio environment matches that recording.

The Bluetooth Core Specification 5.1 added direction finding, including Angle of Arrival, which can tighten accuracy considerably. It also requires hardware that supports it, on both ends, which is why “improve our accuracy” often turns into a hardware refresh rather than a software update.

One constraint sits under all of it. Bluetooth Low Energy shares the 2.4 GHz band with Wi-Fi, microwave ovens, cordless devices and video senders, because that band is unlicensed by design: in the United States under 47 CFR § 15.247 (current edition) and in Europe under ETSI EN 300 328. Interference is not a fault in the deployment. It is the operating condition.

What does a beacon deployment actually cost to run?

The hardware is the smallest line. The costs that repeat are mounting labour, battery replacement, re-surveying after the building changes, and doing all three again each time coverage expands into a new wing, floor or site.

Installation is where a software budget meets a construction budget. Beacons go high up so they are not stolen or knocked off, which means the work involves access equipment, safety sign-off and often night shifts in an operating venue. The trades that do that work are not getting cheaper: the Bureau of Labor Statistics projects employment of electricians to grow faster than the average for all occupations (BLS Occupational Outlook Handbook, accessed September 2026).

Battery life is a trade-off, not a specification. A beacon’s advertised life assumes a slow advertising interval. A blue dot that moves smoothly needs a faster one, and faster advertising drains a coin cell sooner. Read any battery figure next to the interval it assumes; if the datasheet does not state the interval, the figure is marketing.

Then there is the labour multiplier. Replacing a battery that costs almost nothing still requires a ticket, a ladder or lift, a technician, and a window when the space is empty. Multiply that across a large estate and battery maintenance becomes a scheduled operational programme rather than an occasional job.

We have deliberately not printed a per-unit price or a beacons-per-square-metre density here. The figures in circulation come from parties with a position in the argument, and a number you cannot verify is not a number you can budget against. Get both from your own quotes, and insist the quote separate hardware, installation labour and the first five years of maintenance.

Why does beacon accuracy drift after installation?

Because a fingerprint describes the radio environment on the day it was recorded, and buildings do not hold still. The floor plan stays correct. The radio model behind the blue dot quietly stops matching the room.

Shelving gets moved. Partitions go up. A seasonal display lands in the middle of a concourse. Any of those changes the way 2.4 GHz signals reflect and attenuate on the path between a beacon and a handset.

Crowds do it too, and this one surprises people. The 2.4 GHz band is absorbed by water, and a dense crowd is mostly water. A venue surveyed on a quiet Tuesday can behave differently on its busiest afternoon, which is exactly when the wayfinding matters.

None of this breaks a beacon. Every device still reports as healthy on the dashboard while the positions it supports slowly get worse. That is the failure mode to plan for: not outage, but drift that nobody owns.

When are Bluetooth beacons the right answer?

When the thing being located has no sensor suite of its own, when you need dependable micro-proximity at a specific door or shelf, or when a small, stable space needs positioning quickly. Beacons are a real tool, not a mistake.

Asset tracking is the clearest case. A wheelchair, a pallet cage or a tool crib carries no accelerometer and runs no app, so a tag and a reading infrastructure are the honest way to find it. Retail triggers are another: knowing a phone is within a couple of metres of one fixture is a proximity question, not a navigation question.

Angle of Arrival deployments earn their cost where sub-metre precision has a hard requirement behind it and cabling already exists, such as a manufacturing cell or a secure area. A venue that already runs access points with Bluetooth radios has part of the infrastructure paid for.

The trade-off cuts the other way too. Beacons need Bluetooth switched on. Some visitors turn it off for battery, some for privacy, and some never see the prompt. Whatever share that is in your venue, it is the ceiling on how many people ever experience the system you paid for.

What are the alternatives to a beacon network?

Four families compete for the same job: Wi-Fi ranging, ultra-wideband, sensor fusion on the handset, and camera-based positioning. Each moves cost somewhere different, and none of them is free of a survey.

Wi-Fi round-trip time measures time of flight instead of signal strength, which makes it less sensitive to the reflection problems that trouble fingerprinting. It depends on access points that support Fine Timing Measurement, accurate records of where those access points physically are, and handset support that varies by model.

Ultra-wideband is the accuracy leader and the capital-cost leader with it. Anchors need power and cabling, and every tracked object needs a tag. For a hospital tracking infusion pumps that maths can work. For guiding a hundred thousand visitors a year through a mall, it does not.

Sensor fusion shifts the burden into software by combining inertial sensors, magnetic field readings and barometric pressure from the handset. It removes the battery fleet from the ceiling, and it introduces a different dependency: the quality of the mapping and modelling work, plus whatever corrects inertial drift over a long walk. Ask about drift over a 400 m route, not over a 20 m demo.

Camera-based positioning is accurate and carries a governance cost. If it processes images of identifiable people in Europe or the UK, it engages Regulation (EU) 2016/679, the General Data Protection Regulation (official text, OJ L 119, 4 May 2016), which means a lawful basis, a retention position, and a conversation with your privacy office before the pilot, not after.

Does an indoor positioning system have to meet accessibility law?

If a public entity in the United States operates it, yes. The Department of Justice’s ADA Title II rule sets WCAG 2.1 Level AA as the technical standard for web content and mobile apps, published in the Federal Register on 24 April 2024 (ada.gov fact sheet, 2024).

That reaches further into indoor navigation than most buyers expect. Public universities, municipal transit, city-run venues and public hospital systems are Title II entities, and the wayfinding app is web content or a mobile app. WCAG 2.1 Level AA is a published W3C Recommendation dated 5 June 2018 (W3C), so the criteria are knowable in advance rather than a matter of opinion.

Federal agencies have a parallel obligation under Section 508 (section508.gov). Either way, put the standard in the procurement document by name. A retrofit of turn-by-turn instructions for screen reader users costs several times what specifying it up front would have.

How do you evaluate an indoor positioning system before you buy?

Test the system where it will actually be used, and make every claim arrive as a number with a percentile attached. A demo in a clean lobby tells you about the lobby.

Run the pilot in the worst part of the building. Pick the atrium, the loading dock corridor, the basement with the metal racking, and the concourse at peak. Anything that works there will work in the lobby.

Ask what happens in the failure cases, because they are the ones your front-of-house team will field. What does the app show when Bluetooth is off, when the visitor declines location permission, when a device is three generations old, when a floor has just been re-laid? Ask for the accuracy figure again after your first renovation, in writing.

One last question is worth more than most of the technical ones. How does coverage get added when you open a new building, and what does that cost per square metre? A system that scales through configuration and a system that scales through a construction project look identical in a demo and diverge sharply by year three.

Frequently asked questions

Short answers to the questions buyers ask most about indoor positioning systems.

What is the difference between an indoor positioning system and indoor wayfinding?

An indoor positioning system works out where a person or asset is. Wayfinding is the layer above it: the map, the routing graph and the turn-by-turn instructions that use that position. Wayfinding with a poor positioning engine produces a beautiful map that sends people the wrong way, which is worse than a static directory board, because users stop trusting their own judgement once a blue dot is on screen.

Do Bluetooth beacons work if a visitor has Bluetooth switched off?

No. A beacon-based system needs the handset’s Bluetooth radio on to receive advertising packets. Visitors switch it off to save battery or for privacy reasons, and some never respond to an in-app prompt. Whatever proportion of your visitors that represents is a hard ceiling on adoption, so ask any vendor what their app shows to a user who declines, and treat that screen as part of the product.

How accurate does indoor positioning need to be for wayfinding?

Accurate enough to place a person on the correct floor and the correct side of a wall. For corridor-level guidance in a mall or a campus, a few metres of horizontal error is usually workable; a floor error is not, because it sends the user to a different building level entirely. Regulators treat vertical accuracy separately too: the FCC’s ±3 m z-axis rule for wireless 911 calls is a floor-level requirement.

What is the largest recurring cost in a beacon deployment?

Labour, in two forms. Battery replacement and re-calibration are both cheap in parts and expensive in access: lifts or ladders, safety sign-off, a technician, and often an after-hours window in a venue that cannot close. Neither cost appears on a hardware quote, and both continue for the life of the system, which is why a five-year view of total cost looks very different from the first invoice.

Can one indoor positioning system cover a whole multi-site estate?

Technically yes, but the cost shape differs by technology. Infrastructure-based approaches scale close to linearly with floor area, because each new site repeats the purchase, installation and survey cycle. Software-led approaches scale more like a mapping and configuration exercise. Ask for the per-square-metre cost of site number twelve, not site number one, and ask who performs the survey each time.

What should you bring to a scoping conversation?

Three things, and they change the answer more than any product question does: current floor plans, an honest list of your worst areas, and what you plan to build next.

Bring those to a scoping call and we will go through your site’s constraints, the accuracy your use case actually needs, and which of these approaches fits the building you have rather than a reference venue. Book a scoping call.