How accurate are indoor positioning systems?

January 16th, 2019

The last thing you want happening when using or navigating with GPS is faulty position. What good is a positioning system when it can’t show where you are? Naturally, this concern exist for Indoor Positioning Systems (IPS) as well, and perhaps even more so since it’s a younger technology.

First of all, it bears to mention that when it comes to evaluating IPS there are other factors  that can be more important than accuracy – more on that later. At any rate, in this blog post we’ll look specifically at accuracy.

Although it was more common before, we still occasionally get questions about accuracy. When the technology was young, there may have been issues with inconsistent accuracy (among many other things). However, it’s safe to say that those days are gone. Over the last couple of years algorithm improvements have pushed the reliability and overall accuracy to unprecedented levels.

To get a sense of IPS accuracy, we can compare it to two other positioning technologies; GPS and Wi-Fi positioning.

positioning-systems-comparison

Smaller scale, higher requirements

Since indoor environments (eg. inside an office) are on a much smaller scale – compared to the scale of the outdoor environs – we humans tend to have a higher standard of accuracy when indoors. And there’s a good reason for that. The GPS being 5-10 meters off is not a big deal when driving on a road – as long as we’re on the right road and moving in the right direction there are few arguments for a higher accuracy. But inside buildings the same error in meters would result in showing the blue dot in a completely different room. Fortunately, indoor positioning systems provides accuracy tailored for indoor environments. With a accuracy of 1-2 m and smart processing of the calculated position, IPS is definitely within the threshold of acceptable accuracy.

outdoors-vs-indoors-scale

Maturing industry, converging performance

In 2016, an independent study of 6 indoor positioning providers found that most of the tested IPS deliver a high accuracy, as is shown in the graph below. In the years since the test, positioning accuracy have only increased and become more reliable.

indoor-positioning-providers-comparison

As you can see in the chart above, almost all indoor positioning providers displayed high accuracies. The convergence of accuracy is a sign that the IPS industry as a whole is reaching and many times exceeding the requirements of most use cases. A strong indication of the level of maturity of the technologies and the progress the industry has made over the last few years.

What is the perceived accuracy?

While the raw position matters, equally important is the perceived accuracy. So what do we mean by that?

The raw position refers to the unfiltered output of a positioning system. Perceived accuracy means how accurate the user perceives it to be. To create a better wayfinding experience, a couple of tricks can be deployed to cater to the user.

For example, when using a GPS while driving, theoretically you could get position outputs off the road. Now, since the system is assuming you’re driving, it can make an educated guess that you’re probably driving on the road, and snap the blue dot to the road.

gps-while-driving

Although the raw position could be off the road (left), the system assumes you’re on the road, and snaps to the road when displaying the blue dot (right).

A common solution to improve the perceived accuracy is by adding some sort of interpolating of the positions. The simplest of the algorithms for this is the “moving average”. In short, it works by re-calculating the position as the average of the last couple of position updates. While this might seem like a good and robust solution, it often comes with some irritating flaws. When applying an interpolation on the position updates, it comes with the cost of responsiveness. This solution often gives a smooth result but to the user, it feels like the position lags behind.

raw-position-vs-interpolation-filter-15

We advise against using a moving average filter because it introduces unnecessary lag to the wayfinding experience. If you’re more interested in our reasoning behind that, and the various techniques that can be used to improve the blue dot from a user experience point-of-view, I highly recommend you check out our three part series on the topic. (A Blue Dot Experience Part I, II and III)

Additional criteria for performance

While important as a hygiene factor accuracy is far from the one-and-only metric for performance. As mentioned above, there are a number of other questions that are equally or more important for a great IPS experience. For example:

  • Does the IPS work in background mode?
  • What is the refresh rate? Lag time?
  • Does the quality of positioning remain over time?
  • Does it work on most or all devices?
  • Can it handle a large number of concurrent users?
  • What beacon density is required?

Read more about beacon density, what it it and why it matters ➞

In conclusion, the accuracy of today’s IPS is well within the acceptable range for most, if not all indoor use cases. The evidence is clear from independent testing, but also from the number of large-scale commercial roll-outs.


Want to learn more about how to improve the employee experience? Make sure to download our Workplace Survey Report.