Project Note
Predictive Water Main Analytics
Translated machine learning research into an operational risk model for prioritizing aging water infrastructure.
Problem
Water infrastructure failures were difficult to anticipate, leading to costly emergency repairs.
Role
Defined the operational problem, provided municipal data, and aligned outputs with field crews.
What shipped
- Built an ML pipeline with asset, maintenance, and geospatial features.
- Developed a risk score to prioritize inspections.
- Published findings in academic and public data channels.
Impact
- Gave operations teams a defensible prioritization model.
- Strengthened the case for proactive infrastructure investment.
Working themes
- Applied analytics: translated research into a practical risk-scoring workflow.
- Data foundations: combined asset, maintenance, and geospatial data into usable operational signals.
- Public infrastructure: supported proactive decisions for infrastructure prioritization.