Cross-Regional Insights on Sustainable Water Supply Chains using Machine Learning

The Research Problem
Water security is an escalating global challenge driven by climate variability, rapid urbanisation, aging infrastructure, and governance pressures. Countries across different regions are experiencing water scarcity, declining water quality, and increased exposure to extreme events such as droughts and flooding. While significant investments have been made in water infrastructure and monitoring, existing approaches often treat water management as a technical or environmental issue rather than as a complex, end-to-end supply chain involving extraction, treatment, distribution, consumption, and reuse. At the same time, advances in machine learning (ML) offer powerful opportunities to improve prediction, risk identification, and decision-making in water systems. However, these tools are rarely integrated with systemic perspectives on water supply chains or applied comparatively across regions with different institutional and ecological conditions. As a result, policymakers and utilities lack robust, transferable insights into how predictive analytics can enhance the sustainability and resilience of water supply chains. This research addresses this gap by examining sustainable water supply chains under water quality stress across Australia, Mexico, South Africa and the United Kingdom, generating cross-regional insights that combine data-driven analytics with supply chain thinking.

Research Design
The project is guided by three core research questions:

1. How do water quality risks, climate pressures, and governance factors differ across water supply chains in Australia, Mexico, and the UK?
2. How can machine learning models improve the prediction of droughts, contamination risks, and flood vulnerability across these contexts?
3. Which socio-environmental and governance variables most strongly influence water supply chain resilience?

The study adopts a comparative, data-driven research design. National and regional datasets on water quality, climate variables, and socio-economic indicators will be harmonised and analysed using ML techniques such as time-series forecasting, classification, and clustering. The project is delivered through a WUN research partnership involving universities in Australia, Mexico, South Africa and the UK, combining expertise in machine learning, water governance, and supply chain management.

Project Objectives
The project aims to:

1. Develop harmonised, cross-country datasets linking water quality, climate, and governance indicators.
2. Build and validate ML models to forecast water risks and support proactive decision-making.
3. Produce a comparative framework for sustainable and resilient water supply chains.

WUN supports this research by enabling access to diverse datasets, interdisciplinary expertise, and cross-regional collaboration that would not be possible within a single institution. The project aligns with WUN’s focus on sustainability and responsible use of emerging technologies, while contributing to global efforts under the United Nations Sustainable Development Goal 6 (Clean Water and Sanitation).