How does Google Indoor Navigation work?

Question

For a project at my university I have to collect information about Google Indoor Maps. I spent several hours searching information about how Google Indoor Navigation really works. I've found several links on how to create Google Floor Plans like this:

http://support.google.com/gmm/bin/answer.py?hl=en&answer=1685896

And how you can improve location accuracy with the Google Maps Floor Plan Marker app:

http://support.google.com/gmm/bin/answer.py?hl=en&answer=2667756&topic=1685871&ctx=topic

They say the app collects public broadcast data. But nowhere information about which information is collected and how. Several discussions here and a few articles I found say that google uses wifi-access points and the cellular antennas for indoor positioning. But I couldn't find any official information from Google, which I can use in my project.

Can you suggest me where I can find answers to the following questions:

1. How does the Google Maps Floor Plan Marker really work? Which information is collected and how? A official technical Google paper would be nice, so I can use that for my work.
2. How can I increase the accuracy of the Google indoor navigation, by adding Wifi Access Points, Bluetooth stations or other technology?
3. How can I add additional information like Emergency exits and toilets to a floor plan, so the user can navigate with his smartphone to this special points?
4. Is it possible to add information about rooms, so the user can tap with the finger on them to get additional information? For Example which department is situated in this room and so on?

I found a lot stuff about indoor positioning and indoor navigation. But nothing about how google uses them.

Thanks in advance.

Photoconduction

Great! But Google recomends asking questions at stackoverflow using the google-maps tag. Is there a way to ask Google directly, like you suggested? – Photoconduction

have you succeed to integrate google indoor api on android application? – Aquatic

Answer 1

This project just provides maps for indoor places. It doesn't do anything special to augment navigation in these places beyond the standard (GPS, WiFi, etc).

When they talk about (augmenting) positioning with Wifi and Cellular data, they are talking about the same stuff they use on the regular Google maps. This data is collected from cell tower info, and the Wifi stuff is sniffed from the Google street-view cars and then augmented by user-data as people with devices come into contact with these devices and send info back to Google.

It is not very accurate - or even accurate as GPS. It is kind of a "ballpark" estimate. For example, if a Google street-view car detected your WiFi signal when it drove by your house - if you had no GPS, and Google Maps saw your WiFi box, it could estimate that you were somewhere within a few hundred feet of where the Google Street View car detected the signal.

Not as accurate as GPS - but not too bad if you conciser within a few hundred feet better than "no idea". In short - these sources to not give you better or more refined positioning of indoor places. They would be good - for example - to be able to tell you that you are in that particular building if you were inside, and out of GPS coverage.

Manufacture

If we consider your given example you could be right. But I read that it is possible to increase navigation accuracy indoor by using the triangulation of wifi signals, bluetooth stations and the cellular antennas: <engadget.com/2012/01/08/…> You have to track your map and messure the signal inetnisty of these stations at your current position. So you can determine in which room you are. I think this can be also done with the goole map. But could not find whicht technlogies google uses and how you can increase the accuracy. – Photoconduction

Here the correct link. Don't know what happened: engadget.com/2012/01/08/… – Photoconduction

It definitely does do that - but it's not that accurate. Less accurate than GPS, althought GPS doesn't always work (or work well) indoors. You also have to remember, Google doesn't necessarily have a great idea of where these WiFi hubs are, or how things like walls, floors, etc will effect signal strength. Also, Wifi routers are constantly being added, moved and removed, and Google is trying its best to guess at where they all are, but without a solid basis of information. – Manufacture

Yes you are right. Google doesnt know and this is less accurate than GPS. But as you can see in the link, it is possible. Because YOU know where your wifis are. And with the Google Map Marker App you can walk and track your position, so that Google knows too. I read that you have to position at least 3 access points that are reacheable from everywhere in the building. And the App will track where you are. I need something from google which describes which data will be collected by the app and how the data will be used for navigating. How can we increase accuracy by providing the right data. – Photoconduction

Answer 2

Indoor positioning uses WLAN interface. To estimate the real-time location of a user, location systems have to perform a number of steps and various calculations. The calculation of more than one distance from several Access Points (APs) could be used to estimate the exact location. The main principle states that signal strength at the receiver is inversely proportional to the square of the distance that the signal travels. RSS fingerprinting method is used which is based on recording and sampling of patterns of radio signals in specific environment called as pattern recognition or fingerprinting.

Carlin

Answer 3

Since this post is a little old I thought I would still add my thoughts. First of all you can make edits for your own custom maps using MyMaps. As the previous answers stated Google Maps uses cell data, wifi data, gps, etc for tracking. It can also use phone sensor data such as your accellerometer. With the current level of wifi infrastructure it can actually be quite accurate using a method of wifi fingerprinting (The more access points the more accurate it will be given they are effectively positioned). There are several scholarly articles demonstrating its viability down to a couple meters of accuracy. I would suggest reading an article on RSSI fingerprinting. Also check the google maps developer documentation for more detail on editing map information.

Also I was reading that they will be using light detection for additional indoor accuracy (not sure if implemented yet).

Enolaenormity