Skip to main content

Water pipe failure prediction

Posted by: data61

April 29, 2015

Data61 is using data-driven techniques to improve prediction of pipe failures for water utilities. Intelligent predictions reduce maintenance costs, prioritise capital spend and minimise disruption to water supplies and the community.


The Challenge and Data61’s Approach

Australian water utilities currently spend $1.4 billion per year on reactive repairs and maintenance, including the consequence cost of social and economic impact. Focusing the asset maintenance efforts on preventative repairs has the potential to save the water industry $700 million on reactive repairs and maintenance.

Condition assessment of water pipes is an expensive and disruptive process with water utilities typically inspecting only 1% of their network assets per annum. Data61 has developed an advanced failure prediction tool using non-parametric machine learning techniques; with this we expect to double the precision of prediction within that 1% of the network that is inspected.

Industry Engagement

Data analysis completed for:

  • 23 global utilities
  • 700,000 failure records
  • 9 million pipe assets
  • 525,000 kms of pipes.



  • P. Lin, B. Zhang, T. Guo, Y. Wang, F. Chen, “Interaction Point Processes via Infinite Branching Model”, The 13th AAAI Conference on Artificial Intelligence (AAAI), 2016.
  • Y. Wang, B. Li, Y. Wang, and F. Chen, “Metadata Dependent Mondrian Processes”, The 33rd International Conference on Machine Learning (ICML), 2015.
  • P. Lin, B. Zhang, Y. Wang, Z. Li, B. Li, Y. Wang and F. Chen “Data Driven Water Pipe Failure Prediction: A Bayesian Nonparametric Approach”, The 24th ACM International Conference on Information and Knowledge Management (CIKM), 2015.
  • B. Li, B. Zhang, Z. Li, Y. Wang, F. Chen, D. Zhang, and D. Vitanage,  “Multi-level Data Analytics for Risk Water Pipe Selection”, IWA Leading-Edge Strategic Asset Management (LESAM), 2015.
  • Li, B., Zhang, B., Li, Z., Wang, Y., Chen, F., Vitanage D. “Prioritising water pipes for condition assessment with data analytics”, Australia’s International Water Conference & Exhibition (OzWater), 2015.
  • Z. Li, B. Zhang, Y. Wang, F. Chen, R. Taib, V. Whiffin, and Y. Wang, “Water pipe condition assessment: A hierarchical beta process approach for sparse incident data”, Machine Learning (ML), 2013.


  • B. Li, Y. Wang, F. Chen, and Y. Wang, “Group Infrastructure Components”, application filed (N14 014-PROVAU), 2015.
  • Z. Li, Y. Wang, and F. Chen, “Bayesian nonparametric method for infrastructure failure prediction”, WO 2014/085849 A1, 2014.
  • B. Zhang, Y. Wang, and F. Chen, “Extended Hawkes process for infrastructure failure prediction”, application filed (N14 012-PROVAU), 2014.
  • B. Zhang, Z. Li, Y. Wang, and F. Chen, “Determining a health condition of a bridge”, application filed (N12 023-PCT), 2012.


SW logo              UKWIR_logo_RGB


Download the Water Pipe Failure Prediction Brochure.