Machine learning in life cycle assessment

Abstract Machine learning (ML) has been used in life cycle assessment (LCA) to estimate the values of environmental impact characterization factors and to conduct sensitivity analyses. ML has even been used to develop surrogate LCAs, which have enabled prediction of future products’ full life cycle environmental impacts based on design-phase product characteristics. Outside of LCA, applications of ML algorithms have included data cleaning, predicting system output flows or performance, ecosystem informatics, and system optimization. Considering these uses and capabilities of ML, opportunities exist for using ML in cleaning data for life cycle inventories (LCI), estimating flow data for unit processes, improving the quality and quantity of data used to determine impact characterization factors, and generating inventory data for scenario analyses. In this chapter, we introduce LCA and ML, look at how ML has been used in LCA and in development of surrogate LCAs, and discuss other applications that could inform future ML-based tools for LCA.