With ChatGPT, I have created C++ code which is uploaded to four T-Beam SUPREME units and causes each unit to send every 10 seconds its coordinates. Each unit records it’s transmission and it records on its SD card incoming transmissions from the other units. In addition, I have a base station “sniffer” Lora (Semtech 1303 chip) on a Raspberry Pi 4B server which captures all of the field units and publishes what it captures over UDP. I have a logger on the server which then records in CSV the information. I haven’t built the PostgreSQL database component, yet; that will digest all of the units’ transmission and receiving logs and the sniffer’s logs. What I make with all that high precision data has yet to be seen.
I then built a web platform which subscribes to the UDP stream and then publishes live on a map current positions as published by each unit. Each unit’s position fades with passage of 5 seconds.
While tracking and logging tracks is not the main goal for my planned Reticulum field test, the ability to track is a critical component. I have created a Forgejo code repository where I have exercises I that build upon one another. Currently I have SD Card confirmation and high precision GPS time discipline, e.g. to the 1/1,000,000th of a second, so that each unit can work in parallel and not transmit at the same time. Every unit is working in complete synchronization… so I hope and have designed.
Here’s a screen capture video a a test run where I move one unit to another window in the house, and then I take unit C, “Cy”, named after a popular and an active GIS advocate, for a quick walk out into the street. The video is sped up 500% and lasts 49 seconds.
The code is at: https://salemdata.net/repo/jlpoole/microReticulumTbeam/src/branch/feature/fieldtest-beacon-sd-provision and this demonstrates Exercise 12.
Leave a Reply