Predicting CO2 Emissions in Circular Economy Transitions: A Bayesian Network Approach under Data Uncertainty

For the transition to a circular economy, an accurate assessment of CO2 emissions from production processes is essential to make informed end-of-life decisions. Product passports and management shell models can fill these gaps but are currently at the very beginning of their introduction in industrial processes and are too comprehensive and data-intensive for widespread use across many components. To address this problem, we propose a Bayesian network model that predicts the CO2 emissions of the production of a component and incomplete data. It combines information from metadata with concrete knowledge of process flows to provide statistically supported quantitative statements.