Long-Term Autonomous Solar Meshtastic Node
I am currently planning a fully autonomous Meshtastic node for long-term outdoor deployment on a street light pole or mast and I am looking for advice from people who may already have built something similar. The main goal is maximum autonomy with as little maintenance as possible. The environmental conditions are quite harsh, with summer temperatures reaching up to 45°C and winter temperatures dropping to around -20°C, combined with snow, humidity, condensation and potentially very little sunlight for extended periods during winter.
The node should run entirely on solar power and be capable of surviving several days or even weeks of poor weather conditions with only limited solar input. Right now I am leaning towards using a low power nRF52-based platform such as the RAK4631 together with a LiFePO4 battery pack and an MPPT charge controller. To reduce power consumption as much as possible, I do not want WiFi, Bluetooth or GPS to remain permanently active.
One idea I had was inspired by the power management philosophy of the Voyager probes. Under normal conditions the node would remain in a very low power state and only once every 24 hours it would wake up, enable WiFi, connect to a nearby WLAN, contact a remote configuration or C&C server, check for new configurations or firmware updates, synchronize time if necessary and then completely disable WiFi again. The goal is to keep the power budget extremely low while still maintaining remote manageability, because after installation the node would be physically difficult to access.
At the moment I am mainly worried about long-term reliability problems such as condensation inside sealed enclosures, charging batteries below 0°C, brownout boot loops during low battery situations, failed OTA updates on inaccessible nodes, flash wear caused by excessive logging or configuration writes, thermal issues caused by direct sunlight during summer and long-term UV degradation of enclosures and cable materials. I am also wondering how much a metal street light pole may negatively affect LoRa performance and antenna radiation patterns.
I would really appreciate hearing from anyone who has already attempted similar off-grid Meshtastic deployments or other autonomous LoRa infrastructure projects. I am especially interested in existing open source projects, recommended hardware combinations, watchdog or supervisor implementations outside of Meshtastic itself, real-world winter deployment experiences and lessons learned from failures. Photos, GitHub repositories, BOMs or deployment write-ups would also be extremely helpful.