ongoing
Operator Learning
Learning the mapping of partial differential equations using neural operators
Lead: 黄凯期
Projects
Research Projects
Students enter real projects from day one
AI for EDA
ongoing
EDA Yield Database Development and Open-Source Release
Organizing and Open-Sourcing EDA Yield Algorithms, with Future Incorporation of Yield Algorithm Development.
Lead: 刘博豪
ongoing
Development and Open-Sourcing of an EDA Optimization Algorithm Engine
Organizing and Open-Sourcing EDA Optimization Algorithms, with Future Incorporation of Optimization Algorithm Development
Lead: 梅沁心
ongoing
Yield Optimization
Yield-Aware Sizing Optimization
Lead: 区宝文
AI for LCA
ongoing
EPD Intelligent Retrieval
Building an LLM-based framework to automatically extract key environmental indicators (e.g., GWP) from unstructured Environmental Product Declarations (EPDs) and construct a structured knowledge base.
Lead: 黄凯期
ongoing
LCI Data Synthesis Framework
Predicting numerical values in ecoinvent tables using a large language model combined with fine-tuned small models; confirmed usable predictions can then be extended to additional unknown datasets.
Lead: 商靖海
ongoing
LCI Data Synthesis Framework
Developing an automated assessment framework based on a large language model agent (LLM Agent). Leveraging the agent’s planning and tool invocation capabilities, it enables an end-to-end workflow from parsing the raw bill of materials (BOM), automatically retrieving and matching environmental databases (e.g., Ecoinvent), to calculating carbon emissions and generating LCA reports, aiming to significantly reduce manual intervention and enhance the timeliness of accounting.
Lead: 黄凯期、刘佳展
ongoing
LCA Optimization
Integrating large language model agents with operations research optimization techniques into life cycle assessment (LCA) to enable end-to-end automated decision-making, from bill of materials analysis to green material substitution.
Lead: 刘佳展
ongoing
LCI Data Synthesis Framework
Using AI to assist in matching components in LCI inventories with corresponding names and activities in environmental databases (e.g., Ecoinvent).
Lead: 区宝文
Project Process
1
Baseline Reproduction
First month: reproduce Lead-verified baseline
2
Core Experiments
Core experiments and analysis
3
Paper Writing
Participate in draft writing
4
Open Source
Organize code and open-source release
Want to work on these projects
Join and enter real projects immediately, no sandbox period
Apply Now