Hypnos: A Hardware and Software Toolkit for Energy-Aware Sensing in Low-Cost IoT Nodes

Abstract

Through the Internet of Things (IoT), autonomous sensing devices can be deployed to regularly capture environmental and other sensor measurements for a variety of usage scenarios. However, for the market segment of stand-alone, self-sustaining small IoT nodes, long-term deployment remains problematic due to the energy-constrained nature of these devices, requiring frequent maintenance. This article introduces Hypnos, an open hardware and software toolkit that aims to balance energy intake and usage through the adaptive sensing rate for low-cost Internet-connected IoT nodes. We describe the hardware architecture of the IoT node, an open hardware board based on the Arduino Uno form-factor packing the energy measurement circuitry, and the associated opensource software library, which interfaces with the sensing node’s microcontroller and provides access to the low-level energy measurements. Hypnos comes equipped with a built-in, configurable, modified sigmoid function to regulate duty-cycle frequency based on energy intake and usage, yet developers may also plug in their custom duty/sleep balancing function. An experiment was set up, whereby two identical boards ran for two months: one with the Hypnos software framework and built-in energy-balancing function to regulate sensing rate and the other with fixed sensing rate. The experiment showed that Hypnos is able to successfully balance energy usage and sensing frequency within configurable energy ranges. Hereby, it increases the reliability by avoiding the complete shutdown, while, at the same time, optimizing performance in terms of the average amount of sensor measurements.