Data-Driven Decision-Support Tool for Environmental Performance Evaluation: Integrating Automation

Life Cycle Assessment (LCA) is widely recognized as a methodology for supporting sustainable decision-making, particularly through the eco-design of product systems. However, its application remains constrained by persistent challenges: data collection is labor- and time-intensive, inventory mapping relies heavily on expert judgment, and existing solutions are fragmented, often addressing isolated tasks without offering a unified approach. In particular, practitioners lack integrated solutions that combine structured product lifecycle modeling, automated inventory mapping, and scenario-based simulation within a single decision-support framework. This work develops and presents a modular Python-based LCA decision-support tool that addresses these limitations through three core capabilities: (i) structured lifecycle modeling across all phases, adapted from prior literature and enriched with insights from OEM manufacturing use cases at GE HealthCare, (ii) automated mapping of Bill of Materials entries to Ecoinvent processes using large language models with human validation, a feature integrated from recent advances in the literature, and (iii) scenario-based simulation to compare multiple configurations and identify high-impact elements, visualize environmental hotspots, and support eco-design and circularity strategies. A case study at GE HealthCare used as a demonstration for the tool by comparing two spare-part supply chains: (1) a repair scenario, where parts are repaired and reused, and (2) a prime scenario, where parts are replaced by purchasing new ones. Results show that the repair configuration reduces climate change impacts by around 70% and resource depletion by 90%. Additionally, the study illustrates how the tool facilitates LCA applications and its contribution to supporting sustainability-oriented decision-making in industrial product systems.