Research Projects

The global battery industry was valued at approximately $112 billion in 2021 and is projected to grow to $424 billion by 2030. Lithium-ion batteries (LIBs) are expected to remain the dominant segment within this market. Research in this area has become increasingly crucial, particularly with the rise of electric vehicles, where batteries serve as a fundamental component. However, managing end-of-life LIBs are still plagued by economic inefficiencies and pose environmental and safety risks. Although significant progress has been made in LIB recycling technologies in recent years, major challenges remain, particularly in the recovery of metals such as aluminium, copper, nickel, cobalt, manganese, and lithium from spent batteries. Advancing these solutions is essential to building a circular economy within the LIB industry and supporting the sustainable growth of electrification technologies. The proposal aims at utilizing advanced AI tools such as Knowledge Graph Construction and Large Language Models (LLMs) to construct the entire superstructure representing the life cycle of different batteries using the existing literature. The superstructure representing different routes are then analysed for recyclability of key components and cost effectiveness to ensure the circularity and the sustainability by design. The large language models will be used effectively to obtain the recyclability of different routes present in the super-structure. Further, the cost effectiveness will be obtained using the temperature, pressure and the operating conditions used in each route. By utilizing knowledge graph representations, we aim to systematically analyze and evaluate diverse recycling parameters to identify the most effective combination of processes for sustainable battery recycling. Further, the existing online LLM models can provide only the qualitative responses to end-of-life cycle queries on recycling of batteries which is of only limited use. Therefore, we propose to develop a knowledge graph based chatbot which involves re-training the current LLM models by including the rules based on chemistry and thermodynamics for providing more useful quantitative responses on task specific Recycling process and its cost minimization that can be implemented in the real applications. Further, advanced methodologies such as Graph RAG will be used to obtain the important information about the ease and accessibility of key technologies by analyzing the superstructure.