Azam Muzakhim Imammuddien, Septriandi Wirayoga, Syamsul Hadi, Sulistyono, Isac Ilham Akbar Habibi, Miftakhul Huda
The rapid development of electric vehicle (EV) technology has driven the adoption of electric bicycles (e-bikes) as a practical and energy-efficient transportation solution. However, challenges remain in ensuring energy efficiency and extending battery lifetime, particularly in geared e-bike systems. This research presents the design and implementation of an Internet of Things (IoT)-based monitoring and prediction system for e-bike batteries using a microcontroller, voltage divider circuit, and Genetic Algorithm (GA). The system employs four deep cycle batteries (12 V per cell) as the main power source, with voltage sensors and a PZEM-017 module to measure electrical parameters such as voltage, current, power, and energy. A DHT22 sensor is also integrated to monitor environmental conditions that influence battery performance. Data are processed by the microcontroller and transmitted to an Android application via the internet for real-time visualization. Calibration tests of the voltage sensors against a multimeter demonstrated high accuracy with an average error of 0.04%, ensuring reliable monitoring. Experimental results from 10 trials revealed that battery lifetime is highly dependent on temperature and inter-cell voltage differences. Stable conditions \left(26-28^{\circ} \mathrm{C}, \Delta \mathrm{V} \leqslant 0.05 \mathrm{~V}\right) predicted lifetimes of 3.8 -4 years, whereas high temperatures \left(\geqslant 32^{\circ} \mathrm{C}, \Delta \mathrm{V}\right. up to 0.30 \mathrm{~V}) reduced lifetimes to 2-2.2 years. This system effectively enables predictive maintenance, energy efficiency, and sustainable e-bike operation. © 2025 IEEE.
State Polytechnic of Malang, Mechanical Engineering Departement, Malang, Indonesia; State Polytechnic of Malang, Electrical Engineering Departement, Malang, Indonesia