Ten Questions to Ask Before Choosing an Indoor Positioning (RTLS) Vendor

Three proposals lie side by side on your desk. Turn the first page and all three quote similar accuracy figures; the reference pages carry similar factory photos. The only thing that differs is the price on the last page.

Deciding on price alone at this point is risky. Indoor positioning is not a matter of buying equipment; it is a matter of engineering around your site conditions, so two projects that start from the same spec sheet can end very differently. And that difference rarely shows itself before the contract is signed.

This article collects ten questions that surface the differences a proposal alone will not show. Once you have heard the answers, the three documents on your desk will no longer look the same.

I. Catalog Accuracy Cannot Tell Them Apart

Most proposals list an accuracy figure. But that number comes with conditions: an open space, anchors with clear line of sight, tags mounted in ideal positions.

Your site is not that place. Metal equipment stands everywhere, forklifts cross the line of sight, and materials are stacked above head height. How the same hardware behaves under these conditions depends on design and operating know-how. We covered why demo accuracy and field accuracy diverge in Why Positioning Wavers in the Field.

A common objection comes up here: if everyone says roughly the same thing, why not just pick the cheapest? If the three vendors were truly selling the same thing, that would be sound math. But the very fact that identical requirements produced widely different quotes means the three are assuming different jobs. One sized the anchors generously, one sized them to the minimum, and one has not yet put your site into the calculation at all. A price gap is not an answer; it is a question.

So what you should ask about is not the number, but the conditions behind the number, and what happens when those conditions get worse.

II. The Ten Questions

1. Have you done a site with the same conditions as ours?

Ask about conditions, not industries. Even within manufacturing, an assembly line and a metal materials warehouse are completely different radio environments. The vendor should be able to name a site that resembles yours in ceiling height, metal density, outdoor sections, and the kind of objects that move. Why metal environments deserve a question of their own is covered in Tracking Molds Where Metal Is Everywhere.

2. Under what conditions was that accuracy measured, and what is the worst case?

An average alone is half an answer. A system that averages 30cm but occasionally jumps 3m and a system that averages 50cm but never jumps serve different purposes. For finding assets, the first may do; for detecting entry into a danger zone, every jumping coordinate is a false alarm. The vendor that volunteers the worst case is the one that knows its own limits, and only a vendor that knows its limits can design around them.

3. What happens when the signal drops, and how fast does it recover?

On a real site, signals drop. Forklifts block, people stand, materials pile up. What to ask for is not a promise that it will never drop, but the behavior when it does: does the last position hold, does the object vanish from the screen, how many seconds until recovery? A vendor that answers on the assumption of dropouts is a vendor that knows the field.

4. How many anchors do we need, and how was that number calculated?

If it was divided out of floor area, ask again. Anchor count comes not from area but from the shape of the space and the placement of obstacles. A count produced without a drawing is not a design; it is a number made to fit a quote.

5. How many years does the tag battery last, and under what conditions?

Battery life trades against transmission interval. "Three years" by itself tells you nothing; you need to hear at what transmission rate it is three years. Ask for the figure converted to the interval your use case needs, and the number often changes.

6. Where is the data stored, and does it run without an external network?

Location data carries worker movements and production information. This answer decides two things. One is the security review: an architecture that sends data out means more departments to convince and a longer review. The other is availability: when the outside line goes down, does monitoring stop with it, or does the site keep running? The advantages of a configuration that completes on-site are covered in Edge and On-Premise Monitoring.

7. How does it integrate with our existing systems?

"We provide an API" is not enough. Confirm the format, the interval, and whether it is one-way or two-way. Feeding positions into an MES or WMS is one level of difficulty; receiving work orders back from them is another.

8. When we scale up, does the current configuration simply grow?

Ask whether the anchors and servers used in the pilot extend as-is to full deployment, or whether equipment must be replaced along the way. If replacement is required, a large share of the pilot investment becomes sunk cost.

9. Do you build the hardware and the software in the same place?

Tags and anchors from company A, the monitoring screen from company B, integration by company C: this arrangement is more common than you would think. While everything works, there is no problem. The problem comes with ambiguous failures, like an asset that occasionally disappears in one zone. Is it the tag, the anchor layout, the positioning engine, or the screen? When every layer belongs to a different company, even comparing logs becomes a negotiation. A single-source build settles that question internally. Half of incident response is the time spent narrowing down the cause.

10. Can our own staff change settings during operation?

Adjusting zone boundaries, changing alert rules, registering new tags: once operation starts, these come up every week. Confirm whether your staff can do this directly in the admin screen, or whether every change goes back to the vendor. If it goes back to the vendor, every small change carries a cost and a wait, and while the site changes, the system stays with the old site. More often than not, this is the item that decides satisfaction in year two rather than year one.

III. Where the Answers Diverge

The same question draws answers of very different character.

Question	An answer you can trust	An answer to ask again
Accuracy	Conditions and worst case together	A single average figure
Anchor count	A layout based on your drawing	A number divided from floor area
Signal loss	Recovery time and behavior explained	A claim that it never drops
Scaling	Configuration changes by stage	Only a claim of unlimited scaling

The right-hand column does not imply bad intent. But answers like those signal that the vendor has not yet looked at your site, and a quote given in that state is likely to change after the project starts.

IV. What Questions Cannot Settle, a Pilot Can

Even after all ten questions, something remains: the radio environment of your site can only be measured. That is why it is safer to put a small verification zone before the contract.

If you pick the easiest area for verification, it becomes a demonstration, not a verification. Include the hardest conditions, and the pilot will speak for the rest of the site. What to measure and what to set as pass criteria is covered in Designing an RTLS Proof of Concept.

V. Points to Consider Before Adopting

1. Prepare a drawing first: half of the ten questions only get proper answers when a floor plan is on the table.
2. Ask everyone under the same conditions: if the three vendors receive different information, their quotes cannot be compared.
3. Put operations into the contract: battery replacement cycles, unit prices at scale-up, and settings authority are best fixed before signing.
4. Require documents that outlive the people: anchor layouts, configuration values, and integration specs are what make a handover possible.

Closing

The three proposals looked alike because your site was not in them yet. The questions above are tools for getting answers, and at the same time a device that reveals how closely a vendor has looked at your site. Hear the answers, look back at the desk, and the three documents are no longer the same proposal.

ORBRO builds its own tags, anchors, edge servers, and monitoring software, designs the layout around site conditions, and proceeds by verifying in a pilot zone before scaling. If you want the ten answers for your site, send us a drawing and your site conditions. We will answer them item by item.