You can learn a lot about the health of a hive by listening to the sounds that bees make, as a beekeeper would do. The developers of the LongHive project, featured in Make Magazine Vol 75, tell us about their efforts to use machine learning to detect the presence of the queen bee by recording the sounds of the bees in the hive.The developers are Evan Diewald, a PhD student in Advanced Manufacturing at Carnegie Mellon; Antonio Scala, a computer science and math major at Villanova; and Nathan Prihala, a finance major at university of Pittsburgh. Nathan knew beekeepers and saw the problems they were having with their hives.
The LongHive project uses Raspberry Pi running a version of TensorFlowLite, a LoRaWan microcontroller, which is designed for long-range low data transfer, and communicates via the Helium Network. Here’s a diagram of the LongHive system.
The LongHive Project was entered into developed as an entry into the Hackster “IoT for Good” competition and it won the Grand Prize. Get the code, build files and even more details at hackster.io.