We all know the problem with Life Cycle Assessment (LCA): It’s too hard.
To calculate the carbon footprint of a single product, analysts spend months digging through PDFs, manually entering data into spreadsheets, and struggling with missing values.
In my recent paper published in PLOS Climate, I explored how we can modernize this outdated workflow using Data Science.
The New Framework
Instead of treating Machine Learning (ML) as a “black box,” we mapped specific ML techniques to the four standard ISO phases of LCA.
Figure: A roadmap for integrating AI into environmental assessment.
How It Works
Goal & Scope (The Setup): Normally, defining the system boundary is manual. We used NLP (Natural Language Processing) to automatically scan thousands of literature abstracts to suggest boundaries.
Inventory (The Data): Missing data is the biggest headache. We proposed Probabilistic Imputation—instead of guessing a number, we use statistical distributions to fill gaps with confidence intervals.
Impact (The Calculation): Complex chemical models take forever to run. Neural Networks can act as “Surrogate Models” to predict impact scores in milliseconds.
Conclusion & Resources
The future of sustainability isn’t just about better chemistry; it’s about better data.
If you are interested in the technical details or the bibliometric code used in this review, check out the links below: