Using Docker when subscribing to the mqtt topics on the Find3 mosquitto server using mosquitto_sub on the same server no messages are received.

I can can confirm that the mosquitto server is running I got a "Connected not authorized" when connecting without user and password.

When i log in with the family name, it starts a download of a file with extension .gz which i am not able to open and not the page with the dashboard.

Hi,

First of all, thanks for a great piece of software!

I'm hoping to combine FIND with some very low power BLE scanners, currently submitting their output to mqtt. Would be great if this continuous stream of mqtt data could land in FIND's passive scanner, s.t. I can follow BT only devices (e.g. xiaomi miband3 can be tracked almost live this way). I think that requires 2 changes:

• MQTT listener for passive/reverse scanning (current form is active only right?)
• Accept sensordata without wifi fingerprints (if BT or other data is available)

Unless I misunderstood, the current MQTT listener posts data as if they were active scan results correct? I could make sure MQTT data is submitted conform your sensordata model, or in the current hexrssi format.

What do you think? Cheers

I have created a very experimental, early stage client for the ESP32 / ESP8266 microcontroller, which is available from $1. This should allow people to create very cheap beacons.

Update: see https://github.com/DatanoiseTV/esp-find3-client

Please note this is a very early stage and experimental thing. Any input welcome!

Can I run this on a raspberry pi?

I did try but I got this error:

pi@officepi:~/find3/find3 $ sudo docker build -t find3 .
Sending build context to Docker daemon   14.9MB
Step 1/32 : FROM ubuntu:18.04
 ---> 7295fb90f21f
Step 2/32 : RUN apt-get update
 ---> Running in 351b1a998416
The command '/bin/sh -c apt-get update' returned a non-zero code: 139

I'm wondering if this is related to the pi's armv6l architecture or possibly a memory limitation?

Any ideas? Thanks!

The dashboard needs a little work. I have no idea how it is possible to have 0% accuracy on 56 readings.

Also, the number of data readings stored is always directly below the first character. This look great until you name a room something large. I.e. Beer pong room. Then the number of data readings directly overlap the accuracy readings.

The original Find app appears to be significantly more accurate currently during real world test. I am wondering if it takes the a.i. a few days to figure out what to prioritize?

Is there any chance we can expect any neural networking code anytime soon? I am debating on if this is considered a.i. or just computer learning.

Would it be possible to program into equation if a few modems disappear it doesn't effect the outcome too much. I noticed if I train with 12 WiFi signals detected, and then take away the 6 farthest modems, the code can't figure out the location.

Realistically, over two years many of the BLE and WiFi signals will change. Neighbors get new WiFi modems or add extenders, BLE devices are replaced or eliminated. WiFi hotspots never stay in the same place for long. etc.

As long as 6 of 12 signals are detected with the same readings that they should have for that room, I don't see why it wouldn't be possible to code in some kind of solution that ensures this scenario doesn't result in unknown location.

Unfortunately I haven't found a way to reproduce it and this is the only time I was able to capture the log messages.

...
2018-04-17 17:03:04 [INFO] [PID-5268] server.go func1:978 192.168.1.140:50470 POST /passive 19.691983ms
2018-04-17 17:03:10 [WARN] [PID-5268] analysis.go AnalyzeSensorData:117 [qg] nb1 classify: problem preparing SQL: database is locked
2018-04-17 17:03:12 [WARN] [PID-5268] analysis.go AnalyzeSensorData:117 [qg] nb1 classify: problem getting key: database is locked
2018-04-17 17:03:12 [INFO] [PID-5268] fingerprints.go updateCounter:87 have 9 new fingerprints for 'qg', re-calibrating since last calibration was 2562047h47m16.854775807s
2018-04-17 17:03:12 [INFO] [PID-5268] server.go func1:978 192.168.0.1:59900 POST /passive 7.742035674s
2018-04-17 17:03:13 [INFO] [PID-5268] calibration.go dumpSensorsToCSV:399 [qg] dumping 317 fingerprints to /home/find/go/src/github.com/schollz/find3/server/main/data/m0gbo9ga.csv
2018-04-17 17:03:14 [INFO] [PID-5268] server.go func1:978 192.168.0.107:53874 POST /passive 22.710843ms
2018-04-17 17:03:20 [INFO] [PID-5268] server.go func1:978 192.168.1.140:59420 GET /ws?device=all&family=qg 95.342µs
...

Server version: c787ee806f46016ee17b007d5cc690914496df6f Go version: go1.10.1 linux/amd64

I'm researching tools and technologies for a little home-automation project I'm working on. I have a bunch of Eddystone beacons that I'm looking to track around my site. The documentation talks mostly about passive scanning for wifi. Will find3-cli-scanner running on Rasberry Pi with the -passive flag also pick up bluetooth or is it wifi-only?

Thanks much! Looking forward to working with and contributing to the project.

Does the 10 second timeout happen only during the pull room request or does it also apply during the training phase? I'm ok with training taking longer if it results in more accurate results. I do not want the pull room request to take long though.

Bumps github.com/gin-gonic/gin from 1.3.0 to 1.7.0.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Bumps certifi from 2018.8.24 to 2022.12.7.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Hello, congratulations is an excellent solution.

I would like to know if I can do a Wi-Fi heat map as an example.

The idea would be to know the most visited points of a store, with passive scanning and train the limits of the store, and according to those limits everything that is within those limits, the idea is to calculate the coordinates within the limits to paint a heat map.

It would be very interesting to implement it, has anyone implemented it or do you know how it could be done?

I thank you very much.

Thanks.

I want to know that in FIND3 application, if I entered family name, device and enabled the gps option and start the scanning process without the learning process then it should show me the device on maps and provide gps co-ordinates right? Does it works this way?

Following Tracking your Phone document

In FIND3 application I entered the details family name, Device, location. I enabled the tracking mode which went in to learning mode and then start scan as per mentioned in the document. I also enabled the GPS option for precise location. After 15 min of scanning I found the device and the location it shown somewhere near Africa and actually it should show somewhere near my Pune's home address. I used the https://cloud.internalpositioning.com server.

I am following the document of Tracking your Phone.

When I open the app I enter the details family name , device then I select the GPS option and press start scan but after 18 seconds it gets crashed. Having crashing issues with the app. If I am not providing any location and I want to track the device with GPS option enabled it should give me the co-ordinates of my device where it is lying right?