ESP8266 Community Forum

- Mon Jan 05, 2015 11:11 am #6357 Hello everyone.

Has anyone experimented with location tracking using these modules? Since they can act as both a server and client very easily, I feel as though they would be a perfect candidate for this application.

Imagine a large, square room with 4 stationary modules at each corner, and a single mobile module that moves freely about the room.

Using the RSSI value of the mobile module, relative to all four of the stationary modules, would it not be possible to estimate the mobile modules location within the room?

I have just started working with these modules and the Lua Nodemcu firmware. I am also quite interested in indoor tracking, so I plan to investigate this area of application using these modules. However, I wanted to see if anyone has already started experimented in this area or is familiar with what some of the common pitfalls might be.

Thanks.

- Mon Jan 05, 2015 12:15 pm #6361 I tried something like that on the freertos but Wifi scan speed not good for this application.

ESP8266.com Moderator
Lets get the party started and +1

[url=https://www.banggood.com/custlink/KvGGhGF4wG[/url]

I like ESP8266 and ESP8285 and ESP32

- Mon Jan 05, 2015 4:50 pm #6397 In theory that should be possible, some companies such as ekahau have been doing this for a long time, using the existing Wifi APs/HotSpots and their RSSI.

In practice, you need to establish a 'heatmap' of RSSI values of all the Wifi hotspots in sight by doing a site survey. Then the mobile to be located will compare its RSSI values to the heatmap and guess the position.
I think that Cisco also has built-in a protocol and support in all of its Wifi hardware, the feature can be purchased to enable the capability.

Beware that Wifi signal RSSI fluctuates a lot, is subject to reflections, hygrometry, humans bodies moving in the room with laptops or other more or less conductive items (such as a bottle of water), and many factors, which IMHO is why the technology has not been so widespread since its inception about a decade ago.
With ekahau and a thorough survey, we've had precision of about 1 to 2 meters, but with sometimes random erratic values. It's not bad, but often not enough to discriminate exactly in which room the user is, so you have to apply 'physics' corrections that take into account the fact that you can't cross a wall without going through a door before, for example...

You also have to consider the cost of keeping the beacon devices powered, which may or may not be that easy (batteries you have to change, and usually electric sockets are at ground level, whereas you would want your beacons close to the ceiling to minimize interferences, ok for an office but less for a private place with no access to space above ceiling.

Same thing has also be experimented with Bluetooth 'beacons', which used to be a preferred choice because the cost per beacon was lower than Wifi, and power consumption could be lower too (now with BTLE could run on a coin cell for weeks probably)

GhW

- Tue Jan 06, 2015 9:30 am #6458 Thank you for the valuable replies!