Data-driven insights into ESG dynamics and transportation systems

Abstract

In an era where Environmental, Social, and Governance (ESG) considerations are increasingly paramount and the digitization of industries accelerates, this thesis pi- oneers an exploration of the confluence between ESG metrics and Big Railway Data analytics. By weaving together these ostensibly disparate strands, the research uncov- ers a rich tapestry of data-driven insights and predictive models that serve to bridge the gap between corporate sustainability efforts and operational efficiency in the railway sector. The first topic delves into the world of ESG scores, an area that has historically re- ceived limited attention due to disagreements among ESG rating agencies and varying perceptions of ESG performance among investors. This research is pivotal, as it fills a significant gap in our understanding of how ESG factors influence corporate practices and financial outcomes. Through a combination of theoretical predictions and em- pirical research, it uncovers the hidden relationships between ESG performance and corporate outcomes. High ESG scores are revealed to be inversely correlated with the likelihood of ESG-related scandals and stock returns during such events. The results also illuminate the heterogeneity of ESG scores and their dual-edged nature, high- lighting the potential for both positive and negative consequences associated with a strong ESG reputation. Moreover, a novel model emerges, illustrating how firms grap-ple with an optimization problem when determining their optimal level of ESG invest- ment. This model posits the existence of two equilibria shaped by the delicate balance between ESG scandal loss and ESG adjustment cost, providing insight into why some firms make divergent ESG decisions. The second topic of the thesis transitions from the theoretical exploration of ESG scores to a practical application in the railway industry. It presents a detailed case study of the Mass Transit Railway (MTR) Corporation’s use of artificial intelligence (AI) to enhance its ESG profile. By implementing a sophisticated AI model known as a-LSTM, which combines Long Short-Term Memory networks with attention mecha- nisms, MTR has significantly improved its journey time predictions. This improvement has led to a reduction in carbon emissions and passenger waiting times, showcasing the practical benefits of integrating Big Data and AI technologies in enhancing operational efficiency and achieving ESG goals. Through this comprehensive study, the thesis illustrates the importance of a data- driven approach in addressing ESG challenges and operational efficiencies. It provides valuable insights into how the analysis of ESG scores can inform strategic corporate decisions and demonstrates the potential of AI and Big Data in improving sustain- ability outcomes in the transportation sector. By bridging the gap between these two seemingly disparate domains, this research contributes to a deeper understanding of the role of data analysis in advancing both corporate sustainability and operational excellence.