Computational Approaches for Estimating Missing data in Life Cycle Assessment
Life Cycle Assessment (LCA) is a widely used analytical tool that examines the environmental impacts of a product along its whole life cycle, including raw materials extraction, manufacturing, transport, use, and disposal. An LCA study is data intensive. Traditional ways to collect data for LCA involves on-site investigation of manufacturing processes and laboratory tests, which are time-consuming and expensive. In my research, I propose a computational framework to predict the missing data in LCA based on existing data. Specifically, I use techniques in data science, such as link prediction, neural networks, and random forests to develop models to quickly predict missing data in LCA. I use these computational approaches to explore the underlying patterns of the LCA data and reveal the interrelationship between manufacturing processes and the environment, between properties of contaminants and their hazard impacts. Correctly extracting the patterns behind the LCA data helps estimate the missing data without relying on the time-consuming, expensive empirical data collection process. The computational approach will significantly reduce the cost of and save time for LCA studies, therefore help broaden the application of LCA for sustainability decision making.
Ping Hou is a Ph.D. candidate in the joint Ph.D. program in School for Environment and Sustainability (SEAS) and Michigan Institute for Computational Discovery & Engineering (MICDE). Her research interests focus on developing and applying computational and data-driven methods to advance the science and engineering of sustainability assessment. Her current research centers on reducing data collection efforts and increase the feasibility of Life Cycle Assessment.