Research Projects

Electric vehicles (EVs) are gaining prominence globally, with India aiming for 80 million EVs by 2030. Despite this rapid growth, challenges in battery safety, reliability, and performance persist, with recent EV battery explosions raising concerns. Addressing these challenges is crucial for India's EV sector, ensuring public trust, economic sustainability, and alignment with the 'Make in India' initiative. This project aims to develop AI-based techniques for monitoring and diagnostics of EV batteries, focusing on safety, reliability, and efficiency. The scientific objectives include: 1. Developing advanced AI models to identify potential battery failures, enhancing safety, and preventing incidents like thermal runaway. 2. Optimizing hyper-parameters and machine learning models for battery diagnostics to improve precision and computational efficiency. 3. Designing lightweight AI models with a low memory footprint for integration into embedded systems for real-time monitoring. 4. Investigating key battery performance metrics like capacity fade, state of charge (SoC), and state of health (SoH) to ensure reliable operation under diverse conditions. The hypothesis is that AI-driven predictive analytics, optimization, and lightweight integration can improve EV battery safety, reliability, and performance. The project will use practical battery emulators, real-time simulation & testing, and real-world EV data to simulate different operating conditions and validate the models. Key experiments include, but not limited to, : • Developing advanced AI frameworks for condition monitoring of EV batteries. • Investigating the effects of appropriate selection of hyper-parameters on diagnostic accuracy and efficiency. • Evaluating the suitability of low-memory AI models for deployment in resource-constrained environments. • Deploy and testing in significantly limited capability embedded device compared to high-end PCs. On successful execution, the project will provide crucial insights into battery behavior and failure mechanisms, improve safety protocols, reduce operational costs, and extend battery lifespans. Lightweight AI models will enable real-time monitoring and predictive maintenance, making embedded systems more scalable and efficient. These advancements will significantly accelerate EV adoption by addressing critical safety concerns. Outcomes will be disseminated through high-impact journals, patents, and technology transfers, ensuring broad application and industry uptake. Conferences and workshops will further promote knowledge exchange, collaboration, and engagement with stakeholders. This project will meet India's growing EV needs, foster sustainable transportation, and enhance India's competitiveness in the global EV ecosystem, aligning with the nation's vision for a greener future and the 'Make in India' initiative.