Open Access

ARTICLE

A Digital Twin Driven IoT Architecture for Enhanced xEV Performance Monitoring

J. S. V. Siva Kumar¹, Mahmad Mustafa², Sk. M. Unnisha Begum³, Badugu Suresh⁴, Rajanand Patnaik Narasipuram^5,*

1 Department of Electrical and Electronics Engineering, GMRIT, Rajam, 532127, India
2 Department of Electrical and Electronics Engineering, Methodist College of Engineering and Technology, Hyderabad, 500001, India
3 Department of Electronics and Communication Engineering, Sasi Institute of Technology, Tadepalligudem, 534101, India
4 Department of Electronics and Communication Engineering, Koneru Lakshmaiah Education Foundation, Guntur, 522302, India
5 Energy Group, Cyient Ltd., Pune, 411036, India

* Corresponding Authors: Rajanand Patnaik Narasipuram. Email: ,

(This article belongs to the Special Issue: AI in Green Energy Technologies and Their Applications)

Energy Engineering 2025, 122(10), 3891-3904. https://doi.org/10.32604/ee.2025.070052

Received 07 July 2025; Accepted 18 August 2025; Issue published 30 September 2025

Abstract

Electric vehicle (EV) monitoring systems commonly depend on IoT-based sensor measurements to track key performance parameters such as vehicle speed, state of charge (SoC), battery temperature, power consumption, motor RPM, and regenerative braking. While these systems enable real-time data acquisition, they are often hindered by sensor noise, communication delays, and measurement uncertainties, which compromise their reliability for critical decision-making. To overcome these limitations, this study introduces a comparative framework that integrates reference signals, a digital twin model emulating ideal system behavior, and real-time IoT measurements. The digital twin provides a predictive and noise-resilient representation of EV dynamics, enabling enhanced monitoring accuracy. Six critical parameters are evaluated using root mean square error (RMSE), mean absolute error (MAE), maximum deviation, and correlation coefficient (R²). Results show that the digital twin significantly improves estimation fidelity, with RMSE for speed reduced from 2.5 km/h (IoT) to 1.2 km/h and R² values generally exceeding 0.99, except for regenerative braking which achieved 0.982. These findings demonstrate the framework’s effectiveness in improving operational safety, energy management, and system reliability, offering a robust foundation for future advancements in adaptive calibration, predictive analytics, and fault detection in EV systems.

---

Keywords

---