As the need to monitor agriculture parameters intensifies, the development of new sensor nodes for data collection is crucial. These sensor types naturally require power for operation, but conventional battery-based power solutions have certain limitations. This study investigates the potential of harnessing the natural temperature gradient between soil and air to power wireless sensor nodes deployed in environments such as agricultural areas or remote off-grid locations where the use of batteries as a power source is impractical. We evaluated existing devices that exploit similar energy sources and applied the results to develop a state-of-the-art device for extensive testing over a 12-month period. Our main objective was to precisely measure the temperature on a thermoelectric generator (TEG) (a Peltier cell, in particular) and assess the device’s energy yield. The device harvested 7852.2 J of electrical energy during the testing period. The experiment highlights the viability of using environmental temperature differences to power wireless sensor nodes in off-grid and battery-constrained applications. The results indicate significant potential for the device as a sustainable energy solution in agricultural monitoring scenarios.

The rapid advancement of Industrial Internet of Things (IIoT) has heightened the need for efficient data processing and transmission, particularly in energy-constrained environments. This study introduces a novel wavelet-based edge computing methodology designed specifically for low-power IIoT sensors using energy harvesting. Unlike existing implementations that rely on computationally complex instructions, this approach optimizes the wavelet transform (WT) for resource-limited microcontrollers (MCUs) without sacrificing data quality. By leveraging a lightweight assembly-level WT implementation, the proposed solution significantly reduces computational costs and energy consumption. A comprehensive analysis performed on ARM Cortex-M7 MCU on an industrial vibration dataset demonstrates energy savings of assembly language (ASM) up to 87% with discrete wavelet transforms (DWT) and 32.1% with fast wavelet transforms (FWT), compared to C-based implementations. This work is distinct in its ability to dynamically adjust data transmission levels based on available energy, ensuring robust operation in batteryless IIoT environments. Moreover, the method offers flexibility in signal reconstruction, supporting scalable compression ratios and facilitating long-term predictive maintenance applications, making it a pioneering step in sustainable industrial monitoring.

Managing energy in batteryless Internet of Things (IoT) nodes, especially in solar-powered mesh networks, presents significant challenges. This paper introduces an advanced solar irradiance model that simulates detailed daily energy profiles. The model considers various azimuth and elevation angles of solar panels, as well as cloud cover. Accurate simulation of daily energy production is critical for optimising the behaviour of solar-powered IoT nodes. The results highlight the utility of the model as a robust tool for research and simulations involving batteryless IoT devices, emphasising its enhanced capabilities for precise energy management and optimisation. This study offers a reliable framework for predicting and managing energy production in solar-powered IoT networks, thus supporting the development of more efficient and sustainable IoT systems.

Internet of Things (IoT) sensors designed for environmental and agricultural purposes can offer significant contributions to creating a sustainable and green environment. However, powering these sensors remains a challenge, and exploiting the temperature difference between air and soil appears to be a promising solution. For energy-harvesting technologies, accurate soil temperature profile data are needed. This study uses meteorological and soil temperature profile data collected in the Czech Republic to train machine learning models based on Polynomial Regression (PR), Support Vector Regression (SVR), and Long Short-Term Memory (LSTM) to predict the soil temperature profile. The results of the study indicate an error of 0.79 °C, which is approximately 10.9% lower than the temperature error reported in state-of-the-art studies. Beyond achieving a lower temperature prediction error, the proposed solution simplifies the input parameters of the model to only ambient temperature and solar irradiance. This improvement significantly reduces the computational costs associated with the regression model, offering a more efficient approach to predicting soil temperature for the purpose of optimizing energy harvesting in IoT sensors.