Thank you!

[jorge0136]

I have been developing an idea along these lines and am very excited to find someone else is already on their way. Thanks for open sourcing your work!

I am curious to know, has the information provided by your sensors correlated to actual trail conditions?

[eddietheengineer]

Glad you can find it useful! The last thing I did was doing longer term testing in my back yard trying to make sure the code was low power enough that it would stay charged from the solar panel and battery. I found out that I needed to power the sensors off of an output pin instead of direct connecting them to 3.3V/GND--then they could be shut off when the device was in deep sleep between readings.

The other critical addition was code to check battery level on wake--if the battery level was too low, prevent the device from starting the cellular service. I had the device fully discharge the battery, and there was a bug in the Particle boot loader or firmware that caused it to corrupt, I had to recover the device using a ST-Link I believe it was.

Finally--as far as correlating sensors to actual trail conditions, that is the most difficult part. Unfortunately the top piece I had that attached the solar panel and sealed the container came loose during a storm and filled everything with water--so the device is no longer running and I have not purchased replacement parts. I was hoping to be able to get long term data with the resistive and capacitive moisture sensors and hopefully be able to correlate, but early data showed that it was difficult to do. I think based on the soil type, how the sensors are installed, etc--the readings can vary a lot and seem to be unpredictable. I'm sure others have found a way to have consistent measurements, but I have not yet. If you try it out, please let me know how it goes!