Development of an energy optimization Algorithm for honeybess monitoring systems.

Abstract

Several state-of-the-art honeybee monitoring systems utilize the Raspberry Pi IoT platform. However, their biggest challenge is high device power consumption, which scales with the number of sensors connected. Previous approaches for power optimization such putting the sensors to deep sleep, turning off the GPIO pins whenever data is not needed, collecting data at specific time intervals, and turning off the power and activity LEDs are difficult to optimize. There is therefore a need to develop efficient energy optimization methods for the Raspberry Pi. We set out to develop an energy optimization algorithm for the Raspberry Pi 4 B Model, targeting the insect monitoring work under the under AdEMNEA project at Makerere University. We developed an algorithm which could manage the CPU, GPU and RAM clock, turn on and off USB ports, HDMI ports Bluetooth, WIFI, ethernet port and onboard LEDs. On testing when the USB, HDMI, Ethernet Port, Bluetooth, and Onboard LEDs were off; WIFI on; and varying the clock speed of the CPU, GPU, and RAM; we were able to save power consumed by the Pi by 200 mW to 650 mW, for the clock range of 100MHz to 1000MHz. Our result demonstrated an efficient approach to energy optimization for the Pi, which is a feasible option since the power consumed by the Pi itself is way above what the sensors draw from it.