Brightcon 2026, hackathon and courses

Highlights
1. Most AI-assisted LCA research focuses on data gap filling, automated foreground system construction remains underexplored.
2. A user-driven pipeline is proposed: users input a rough system description, and LLMs automatically construct the foreground system with system boundaries and key inventory parameters.
3. The framework will be implemented in Python and integrated with Brightway for LCA calculation.
4. The AI-constructed foreground system will be evaluated against traditional manual LCA approaches in terms of efficiency and consistency.
5. The pipeline aims to lower the barrier for LCA practitioners without deep modeling expertise.

Context
AI applications in Life Cycle Assessment (LCA) automation have grown rapidly in recent years. Existing studies have explored two main directions: using machine learning or large language models to fill data gaps in the life cycle inventory (LCI) stage and developing AI-driven pipelines to automate the overall LCA workflow. However, the automated construction and calculation of foreground systems specifically remain underexplored.

Approach
This study proposes an AI-driven pipeline in which users provide a rough description of their system, and the tool automatically constructs the foreground system. Users can then review and adjust the outputs before proceeding to calculation. This approach significantly reduces the manual effort required in conventional LCA practice.

The framework will be implemented in Python with Brightway as the core LCA engine. The AI-constructed foreground system will be compared against traditionally built models to evaluate consistency and time efficiency, and its ability to meet user-defined requirements will be assessed.

Brightcon Spirit
The proposed pipeline will be developed as an open-source Python tool built on Brightway, ensuring full transparency and reproducibility. Although the codebase is currently under development, the session will present the framework design, preliminary results, and invite community feedback on both the methodology and implementation.