*Corresponding Author: kumar.saurabh00@gmail.com 133 DOI: https://doi.org/10.52756/ijerr.2025.v47.011 Efficient Utilization of Energy in IoT Devices Using Machine Learning Algorithms Kumar Saurabh 1 *, Manish Madhava Tripathi 1 and Satyasundara Mahapatra 2 1 Department of Computer Science & Engineering, Integral University, Lucknow, UP, India; 2 Department of Computer Science & Engineering, Pranveer Singh Institute of Technology Kanpur, UP, India E-mail/Orcid Id: KS, kumar.saurabh00@gmail.com, https://orcid.org/0009-0009-2088-576X; MMT, mmt@iul.ac.in, https://orcid.org/0000-0003-3441-5733; SM, satyasundara123@gmail.com, https://orcid.org/0000-0002-9587-6659 Introduction The term the Internet of Things (IoT) depicts the connection of different objects or devices on the internet, for example, sensors, actuators, smart devices etc that collect and exchange data, playing a crucial role in industries like health care, machinery, agriculture, and smart cities (Al-Fuqaha et al., 2015; Jain et al., 2023; Saurabh et al., 2023). They have been extremely successful. Nevertheless, issues abound with their use, one of them being that of energy consumption. Many IoT devices, especially those deployed and designed to be placed in non-manned regions or difficult-to-reach areas, hence the need to use battery-powered devices, which may also not be easy to change (Cheng et al., 2022; Mao et al., 2021). Thus, it is critical to keep these devices running for as long as possible by making smart use of energy or energy-harvesting techniques to improve the efficiency of their energy consumption model (Hisham et al., 2020). Traditional methods for improving the efficiency of energy usage in Internet of Things devices typically rely on predefined rules that are applied to unique scenarios (Sadhukhan et al., 2021; Schmidtke, 2020; Dawn et al., 2023; Gudumian et al., 2024). Traditional approaches, on the other hand, usually have difficulty adapting to the rapidly changing and diverse environments in which IoT devices operate. An intriguing approach to addressing issues with energy management is the crossover of ML integration into IoT networks (Zantalis et al., 2019). Some of the capabilities that are improved with the application of machine learning include the ability for the device to learn usage patterns, provide precise energy forecasts, optimize device operations and adapt to the environment or usage (Cui et al., 2018). This paper examines the efficiency benefits of employing ML in IoT devices, focusing on their best efficiency-focused ML models and their implementation, as well as contrasted with traditional strategies for managing energy. Article History: Received: 24 th Nov., 2025 Accepted: 26 th Mar., 2025 Published: 30 th Apr., 2025 Abstract: The expansion of the Internet of Things (IoT) has revolutionized various industries, allowing for the automation of processes, monitoring in real time, and smart decision-making. One of the most significant difficulties confronting IoT devices is energy efficiency, given that many operate on constrained power sources. This study describes the application of ML algorithms in the energy optimization of IoT devices and also analyzes currently available tools for energy efficiency improvement, including predictive modeling, adaptive resource distribution, and energy-aware algorithms. The proposed ML-based adaptive GPS scheduling algorithms show improvement in efficacy in terms of energy consumption and at the same time maintain positional accuracy. This study compares and shows that using the flexible scheduling option is more energy-efficient, especially for users who have diverse patterns of mobility. Furthermore, this study also looks at the potential of ML approaches such as reinforcement learning, supervised learning, and unsupervised learning for predicting device usage, improving energy efficiency, and extending the battery life, which leads to reducing energy consumption while maintaining QoS. Keywords: Adaptive resource management, Energy efficiency, Energy-aware algorithms, Machine learning, Predictive modelling How to cite this Article: Kumar Saurabh, Manish Madhava Tripathi and Satyasundara Mahapatra (2025). Efficient Utilization of Energy in IoT Devices Using Machine Learning Algorithms. International Journal of Experimental Research and Review, 47, 133-145. DOI: https://doi.org/10.52756/ ijerr.2025.v47.011 Int. J. Exp. Res. Rev., Vol. 47: 133-145 (2025)