Program 4

Analytics and machine learning for carbon performance

Program 4 focuses on leveraging analytics and machine learning to enhance carbon performance across the infrastructure lifecycle. This program aims to address the challenges of processing and interpreting complex, uncertain, and noisy emission data through innovative projects. Key initiatives include developing federated learning systems for accurate carbon performance modelling without data sharing, creating risk-carbon-cost optimized maintenance strategies using AI for sustainable operation, and devising explainable visualization tools for real-time infrastructure performance monitoring.

These projects are designed to provide a comprehensive toolkit for managing carbon emissions effectively, ensuring long-term carbon neutrality and supporting the infrastructure sector’s transition towards more sustainable practices. Through advanced analytics and machine learning, Program 4 is set to revolutionize how we understand and improve carbon performance in infrastructure development.

Co-Lead:

Professor Lihai Zhang

Project 1

Project lead: Prof Flora Salim (The University of New South Wales)

Integrated data analytics and machine learning for carbon performance modelling, forecasting, and option explorations

This project aims to develop AI and machine-learning based techniques to develop capability in:

1. Developing a novel AI method and model for carbon performance and lifecycle modelling with strong generalization capabilities, capable of predicting carbon performance across the infrastructure’s lifecycle, even with limited training data.
2. Integrated and federated learning of carbon performance from knowledge bases and historical sensor / IoT data.
3. Generating novel synthetic datasets and benchmark for machine learning of carbon performance.
4. Visualisation for real-time performance monitoring and management, generating recommendations and suggestions for carbon emission reduction, towards carbon neutral infrastructures with AI-aided solutions, with the varying considerations and tradeoffs.

Project 2

Project lead : Prof Lihai Zhang (The University of Melbourne)

Developing risk-carbon-cost optimised maintenance strategy

The overall aim of this project is to develop a highly efficient, generic multi-criteria, multi- objective optimised maintenance strategy. This strategy aims to optimise structural risks, financial costs, and carbon emissions throughout the service lifecycle of key infrastructure, including buildings, bridges, and pipelines.