The evolution of research at the intersection of industrial ecology and artificial intelligence

Abstract

The intersection of artificial intelligence (AI) and industrial ecology (IE) is gaining significant attention due to AI's potential to enhance the sustainability of production and consumption systems. Understanding the current state of research in this field can highlight covered topics, identify trends, and reveal understudied topics warranting future research. However, few studies have systematically reviewed this intersection. In this study, we analyze 1068 publications within the IE–AI domain using trend factor analysis, word2vec modeling, and top2vec modeling. These methods uncover patterns of topic interconnections and evolutionary trends. Our results identify 71 trending terms within the selected publications, 69 of which, such as “deep learning,” have emerged in the past 8 years. The word2vec analysis shows that the application of various AI techniques is increasingly integrated into life cycle assessment and the circular economy. The top2vec analysis suggests that employing AI to predict and optimize indicators related to products, waste, processes, and their environmental impacts is an emerging trend. Lastly, we propose that fine-tuning large language models to better understand and process data specific to IE, along with deploying real-time data collection technologies such as sensors, computer vision, and robotics, could effectively address the challenges of data-driven decision-making in this domain.