Real World Use?

[neostim]

Hi there,

I'm curious if this project has been used in any real world use cases yet? I'm curious what kind of accuracy you were seeing.

I currently track assets using BLE tags outdoors (with several solar scanners + triangulation) but I'm now looking to also track indoors and found this repository interesting.

[j-tobias]

Hi,

thank you for your interest. I'm sorry the Project hasn't been used in real world yet. I'm solely using it for tracking myself at home and controlling lights for entering and leaving a room.

The accuracy is quite good even though the latency is a bigger issue for me. I unfortunately don't have any metrics supporting these findings. Additionally, the accuracy is heavily dependent on the amount of sensors you use.

But there are still many ways to Improve the locationing capabilities

[neostim]

Thanks a lot for the reply, it's appreciated. I want to give your code a try, but do you have any documentation on how you created the trained model (the pkl file)?

I can do a site survey to generate the x,y, reader and RSSI values but I'm not sure how to go from there to a trained model. Any help or links is appreciated :)

[j-tobias]

I've uploaded my complete Pipeline in utils/model_pipeline.ipynb. There you can see how my Dataset looked like and how I configured the RandomForrest.

I unfortunately don't have any documentation 😬

[neostim]

Thank you very much! I'll go through this today.