Accuracy in Real-Time Location Systems

Real-Time Location Systems (RTLS) do exactly what the name says, they let you see where tracked things and people are right now. They have any number of obvious business applications in any sector you care to name and form the basis of a long list of business-critical operations. 

Table of contents:

• What is the accuracy in Real-Time Location Systems?
• Start with Why
  • What level of accuracy will get the job done?
• What’s wrong with being too accurate?
• Focus on ROI

What is the accuracy in Real-Time Location Systems?

Equipment, personnel, industrial assets, elements in a shipping process—all that and more can be efficiently tracked and inventoried with RTLS far more efficiently than, for example, old school processes involving logs and registers that are prone to human error and require constant manual updates. As an integral component of Industry 4.0, such indoor positioning tools drive many of the key benefits of IoT applications.  Still, there is no standard fit for all RTLS use cases and the required precision for asset tracking varies greatly from one case to another. We’re talking about the degree of accuracy here, not the overall reliability of the results. Tracking by zone- or room-level location is perfectly sufficient for the overwhelming majority of use cases. There are some contexts where one or two-meter accuracy might be desirable, like in a warehouse where all pallets or boxes look the same and being able to locate a particular one is important, but those are the minority of use cases.  Where your needs fall on the spectrum of required accuracy will determine what kind of RTLS deployment you need.

Find out how to drive successful RTLS projects.

Download a free whitepaper here.

Start with Why

Do you need to track things that are in constant motion, like people or certain manufacturing assets, or do you need to track something that it mostly stationary after it’s put into place, like the contents of a warehouse? In a recent post about choosing the right hardware for your beacon deployment, we advised anyone in the planning stage to start by answering the basic question of what job you want to accomplish. When it comes to determining how much accuracy you need your Real-Time Location System to deliver, we suggest you start with a similar question:

What level of accuracy will get the job done?

The answer you give will have a strong influence on the design, performance and cost of your Real-Time Location System deployment. This happens to be an especially easy question to answer since there are basically only two options—”enough” and “very high”.  In 99% of cases, it’s enough to know that, for example, a particular person or machine is in, say, Room 5. Unless Room 5 is incredibly huge, finding the person or machine shouldn’t be a problem. Do you really need to know that the precise location is in the northwest corner of Room 5, forty-eight centimeters from the wall on the west side and seventy-three centimeters from the wall on the north side? If you want that kind of precision, you can get it (and pay for it), but is it really necessary?  If your tracking needs change after your initial deployment, you can always match the granularity that you need by expanding the physical network of sensors you use or by making it more dense. There’s no point in starting out with an RTLS that delivers super-accurate results unless you’re sure that’s what you need right from the beginning.

What’s wrong with being too accurate?

“Too much accuracy” sounds like a good problem to have, but in fact it’s just a way to generate unnecessary costs and headaches. Those costs mostly come from the extra sensors and hardware needed to broadcast a signal from more places, allowing tracked people or things to respond more often, thus creating more precise positioning. Again, if you don’t need it, what’s the point?  But the costs go beyond excess hardware. You can also create additional, and completely avoidable, problems by using the wrong settings on the hardware that you do need. Beacons have something called an advertising interval. It’s the frequency with which they broadcast their “I’m here!” message to the world. The more often they send the signal out, the more feedback they get, which translates into greater accuracy in tracking applications.  The problem is that sending that signal out more frequently takes more battery power. In use cases involving locating things that stay put for extended periods, you don’t need to ping their location every two seconds. Adjusting the settings of your RTLS hardware to fit your needs means getting longer battery life and reducing the costs in time and money of replacing the batteries earlier than you need to.  Remember that we’re not trying to discourage anyone from using a Real-Time Location System for a very high level of accuracy, just to make sure that the requirements of your use case justify it. Otherwise, you’re using (and paying for) much more tech than you need.  The costs of the infrastructure and computing power required to supply a very high degree of accuracy need to be offset by business benefits delivered by that precision. This could be time saved in searching, faster selection from warehouse shelves or any number of other advantages. If a highly-precise RTLS pays for itself in terms of the benefits it brings, great, but be sure you’re only spending as much as you need to. 

Focus on ROI

It’s important to remember that the degree of accuracy needed for a particular Real-Time Location System deployment can vary and, beyond a certain point, more accuracy doesn’t always make technical or financial sense. There’s no need to pay for the hardware necessary to deliver pinpoint accuracy when room- or area-level location is enough. When the efficiency of a location system is measured in terms of ROI, like any other capital investment, spending beyond requirements simply distorts the math and obscures the real value gained from the installation.  Excess capacity might sometimes be useful in other contexts but when you have surplus RTLS hardware for an application that doesn’t need it, it remains just that—surplus.