Where is the Google Maps for Indoors? Part 1 – The Requirements

We have managed to figure out how to build airplanes and how to fly. We have invented the internet and can communicate with people on the other side of the world. We have even launched satellites and rockets into space and are thinking about colonising Mars. But despite spending almost 90% of our time indoors (not to mention present times with Covid-19) we still do not have an app that allows us to navigate any building we enter. We are left with the tools of caveman when it comes to navigating the indoors: entering a building and then searching and asking our way to our point of interest. Is indoor navigation really rocket science?

Whenever we search for something (and I do not mean in a spiritual way), it is either (1) indoors, (2) outdoors or (3) on the internet. Outdoor navigation we solved with GPS in the 1960/70s and internet navigation we solved with search engines in the 1990s.

It is always difficult to envision the attributes a novel technology needs to have in order to take off. But looking at common success factors of adjacent technologies, such as GPS and search engines, we can at least get a rough estimate of the success factors. So let’s take a look at some of the most important commonalities between GPS and search engines in search for some clues on the technological requirements for the future Google Maps of indoors: 

1. Sufficient/High Accuracy: GPS has an approximate accuracy of 5 meters which is sufficiently accurate to find larger objects outdoors. And Google’s search engine accuracy is certainly one of the most important reasons why it has a 92% search engine market share.

2. Reliable and Consistent: While GPS uses a mix of sensors and data (e.g. telecom, inertial and pressure sensor data) to enable the GPS experience as we know it, the key requirement is line of sight between the receiver device and the satellites. This is also one of the main reasons it does not work for indoor navigation. But for the outdoors it is fairly reliable across all weather conditions anywhere in the world. Similarly, as long as you have an internet connection you will get reliable and consistent search results from search engines.

3. High update speed: Both GPS and search engines offer results with low latency. In fact, one of Google’s original OKR’s was to “make the web as fast as flipping through a magazine”

4. Low installation cost: GPS and search engines services  do not come with any installation costs. Furthermore, both technologies operate on a freemium model.

5. Scalable: GPS technology initially required huge receivers for usage, now requires only a small chip. Search engines are a software product and do not require any special hardware from users. The small chip for GPS and software solution for search engines makes both technologies very scalable, especially in combination with the next commonality.

6. Smartphone compatible: The beauty and advantage of both solutions is that they do not require any additional device apart from our smartphone which we already use and recharge every single night.

7. Low infrastructure requirement: Apart from the initial 31 satellites in orbit, GPS does not require any more hardware to scale. However, it does require software. In particular, without large lookup table and maps, latitude and longitude data are not very valuable. Search engines share a similar lookup problem. For example, Google uses a combination of custom database technology (Bigtable), an estimated 2.5 million servers (in 2016) and several other software tools to manage its 3.5 billion searches per day. While there are significant infrastructure required from Google on the server side, it can aggregate those in a few specialised office facilities.

8. Low total cost of ownership (compared to value/profit): The initial GPS system has been launched and is maintained by the US government. Developing and launching the GPS system was expensive, but today the benefits inside and outside the US government by far outweigh the costs and therefore the system is considered a public good nowadays. Total cost of ownership for a search engine, especially for Google, are high. It requires continuous work and investments to ensure accuracy, reliability and scalability. However, the costs are worth it. Google’s dominance over the search engine market results in large advertising revenues. These left the company with ad revenues of $98.1 billion in 2019 and a profit margin that fluctuates around 20%.

Now let’s translate these attributes into technological requirements for an indoor navigation solution:

1. Accuracy: Indoor environments are very complex, with different floors, areas, rooms and even sub-rooms. This means that the next Google Maps for Indoors will need to have a resolution down to an arm’s length, being able not only to distinguish if but also where you are in a particular room.

2. Reliability and consistency: The next Google Maps-like technology will need to produce consistent and reliable performance over time and in varying environments. For indoor environments this is means accurate results despite changing density of people, moving objects, different building construction types and interior re-designs. Furthermore, if phones are used, the technology will need to support them being in the pocket or the apps being in the background (e.g. room occupancy use cases). This problem recently turned out to be an insurmountable challenge for many governmental contact tracing apps (e.g. UK). For privacy reasons, Apple’s operating system limits the access of Apps to Bluetooth functionality if they are not actively used.

3. Update speed: Since in an indoor environment we can enter or leave a room within less than a second, we require low latency on the movement updates. This translates into several updates per second for the future Google Maps of indoors.

4. Installation cost: The cost function for existing indoor positioning technology is largely driven by beacons, gateways, power outlets and cabling, power consumption and setup time. To drive down the cost hardware needs to be replaced with software whenever possible. In addition, mapping of and building navigation models for the indoors is a difficult and costly task due to the complexity of buildings and the lack of good floor plans. The future Google Maps of indoors will need to find cost efficient ways to map the indoor world.

5. Scalability: Scalability of the technology will be less of a problem in terms of number of people it can support, since you can only pack so many people into an indoor space. For the indoors it rather means the number of supported buildings (different sizes, different construction material, etc.) and devices (tags, smartphones, smartwatches, etc.). Furthermore, we also need to keep in mind that any technology using infrastructure needs to have a scalable status-check and re-calibration process.

6. Smartphone Compatibility: An indoor positioning technology that will scale to the success of Google Maps will hardly use tags or a special device. It will have to be supported for mobile devices, across all models (current and future ones) and major operating systems such as iOS and Android.

7. Infrastructure requirement: Ideally the future solution should be infrastructure-less, similar to GPS or search engines. This will help in lowering costs and increasing scalability.

8. Total cost of ownership: The total cost of ownership is the sum of infrastructure, configuration, calibration and maintenance. Technologies which are relying on fingerprinting (a way to improve accuracy by mapping out the signal strength of Wifi or Bluetooth transmitters) will require frequent re-calibration and will not be feasible for the next Google Maps of the indoors.

As you can see, indoor navigation is really not an easy problem to solve. It requires higher accuracy and latency than GPS and similar latency to search engines. Since indoor environments are more subject to change than outdoor environments, indoor positioning solutions need to be more reliable and consistent than GPS, but less so than search engines (with thousands of new webpages popping up every day). Furthermore, they need to be as infrastructure-free as GPS or search engines to really take off and scale. Otherwise the cost drivers will erode any revenues. And there is no way around being smartphone compatible. 

In summary, the Google Maps for Indoors has a difficult set of requirements to meet. In our next article we will compare the most prominent indoor positioning technologies based on the requirements outlined above. This way we will be able to see which technology is the most promising for indoor positioning and navigation.

We’d love to hear your thoughts on the topic and this article – don’t hesitate to discuss with us in the comments section on Medium!  

Blog:

1 Minute Elevator Pitch  –  Forkbeard’s Indoor Positioning Solution

02.12.2020
Blog:

Hello, Forkbeard: Meet Eirik!

25.11.2020
Blog:

Where is the Google Maps for Indoors? Part 2: The Technologies

18.11.2020