In Part 1 we explored the requirements for a technology that could be used to build the indoor Google Maps. By analysing what made other navigation solutions (GPS and search engines) successful, we were able to identify eight dimensions ranging from accuracy, to cost and scalability.
This blog post will evaluate the technological contenders against each other on the eight dimensions we have previously determined. To keep it simple we’ll use a table to compare the different technologies that you can subsequently use to evaluate different indoor positioning technologies.
Currently there are several different approaches to solve the indoor navigation problem. These range from basic WiFi networks up to futuristic VR/AR mapping at Facebook’s Reality Labs.
However, when was the last time you used an indoor navigation app? For most people the answer will be “never”. Therefore, it is safe to say that no one technology has been able to meet all of the necessary requirements. Or there are far greater challenges to indoor positioning than to GPS and search engines which we did not capture. Nevertheless, as you will see from our analysis below some technologies are closer than others in meeting our eight requirements.
Some upfront concepts
Before we dive into the comparison, we need to clarify some important concepts. These will help you to understand some of the reasoning we provide to support our evaluation:
- The same technology but different positioning algorithms can yield very different accuracies. In our evaluation of Bluetooth Low Energy we refer to the classic positioning algorithm using trilateration and Relative Signal Strength Indication (here an excellent article that explains the concepts in more detail). While there are alternatives leading to higher accuracies (e.g. Angle-of-Arrival and Time-of-Flight) these come with substantially higher installation and ownership cost, as well as lower scalability. This is due to the fact that they require a high density of cabled beacons.
- Reliability and consistency of indoor positioning solutions is often dependent on calibration. Calibration or fingerprinting is the concept of mapping out venues for signal strength (e.g. WiFi strength at any given point in the building). Calibration is important for some technologies since the position of a person will be assessed based on the signal strength between the receiving device (e.g. person’s phone) and emitters (e.g. WiFi access point). Positioning technologies that require calibration are unreliable. The constant change of furniture and human presence in indoor environments alters the signal strengths.
- Many positioning technologies use physical gateways. The role of gateways is to configure transmitter devices, upgrade them and report their status. While gateways decrease scalability due to their physical nature, without gateways it is impossible to support remotely and maintain large indoor positioning systems.
In the table below we evaluate the most commonly deployed technologies against each other. We assign scores to each technology across the eight dimensions on a 5 point Likert scale ranging from ‘Very Low’ to ‘ Very High’. The dimensions are stated in their ideal state (e.g. ‘Low Installation Cost’), hence, ‘Very High’ is the optimal score a technology can get across all the dimensions. While this is our subjective assessment, we provide objective reasons for each score.
Now if we sum up the above analysis into a digestible visual, we end up with the following:
Now where does this leave us? We can see that most technologies need to balance accuracy against cost and scalability (remember our upfront comments on different Bluetooth Low Energy positioning algorithms).
Technologies combining Bluetooth Low Energy and ultrasound seem to deliver the best tradeoff between accuracy, cost and scalability. This can be confirmed by looking at Microsoft’s last indoor positioning competition in 2018. 33% of the participating companies used ultrasound based technologies. While not perfect, a combination of Bluetooth Low Energy and ultrasound seems to be the best contender to be used in the next Google Maps of the Indoors. At least for now.
There are futuristic technologies such as Facebook Reality Labs’ Live Maps project which tackles the problem from a completely different angle. These technologies will be facing other pros (e.g. mapping) and cons (e.g. privacy), different to outdoor and internet navigation. These additional requirements unique to the indoor positioning industry, we will cover in a future article.
We’d love to hear your thoughts on the topic and this article – discuss with us in the comments section on Medium!