2026-07-06 –, 154
This sensor network, commissioned in 2024, measures flood levels at the Snowy River passage, a mile into the cave. From the online data we determined that the water level sensor stopped working in March of 2025. The remaining network continued to work until May 2 when it quit forwarding messages. In the cave we found that the water sensor microcontroller had a dead battery as did the first Longwire terminal. Our team did the planned annual battery changeout in the cave.
Afterwards we found that the network was broken just a few nodes into the cave. After several months we discovered five dead batteries in some of the first nodes. We discovered a dip switch setting on two of them were not back in the low power mode. After another month we discovered all 26 batteries had subsequently become depleted! Entrance to the cave ceased due to a government shutdown, followed by an annual bat hibernation closure. Unable to access the cave, Ted Lappin tested a trial network at home and discovered that a binary bit in the sleep mode configuration settings allowed radios to wake up earlier when they lost synchronization with the coordinator node. The radios would be operating continuously without sleep intervals in a few weeks as their clocks were adjusting. This caused the costly 3 -month battery lives from the dip switches being wrong on two nodes. It snowballed to kill all of the batteries as all 26 nodes stayed awake once their sleep cycle modified itself.
John T. M. Lyles