Case Study

Assessing Wastewater Pipes with AI in Southeast Australia

Greater Western Water & VAPAR Case Study

AUTHOR

Related Topics

Condition Monitoring

Artificial Intelligence

CCTV

Western Water have completed CCTV condition inspections of significant parts of their wastewater network since 2014. While the worst condition pipes had been identified and repaired, there remained almost 40 km of inspected pipes where the risk was not easily quantifiable due to the available data and the way it was structured.

THE CHALLENGE

VAPAR presented a proposal to the Western Water Reliability Team to use our AI algorithm to quickly process this large dataset and present the inspection results along with consequence of failure data. This methodology provided a risk-based framework to convert a large pipe dataset into distinct risk cohorts to provide outputs that will allow more informed decision making for budget allocation and developing rehabilitation programs.

Most of the inspection footage uploaded had been categorised with a structural grade of 4 and varying levels of service grading. There was not a high level of confidence that these grades alone provided the optimum method of planning future works. The time, cost, and resources required to review years’ of videos scored by a wide range of people and companies was the primary challenge. 

THE SOLUTION

VAPAR used a combination of automated and advanced auditing to provide consistent scoring and defect analysis options through the VAPAR.Solutions™ platform. All pipes went through VAPAR’s renewal recommendation module that uses smart logic to provide consistent recommendations based on specific pipe defect categories. Recommendations can include relining, patching, root treatment, jetting, or a combination of these measures.

THE RESULTS

A collaborative workshop between Wester Water and VAPAR defined the consequence of failure and aligned the results with the Wester Water’s corporate risk matrix. These outcomes were then used to define risk cohorts in the dataset. The result allowed for more detailed inputs to be considered and improved budget estimates for targeted future capital works directly informed by risk. Approximately $2M was saved from the previous capital works program estimate that was based on the available summary data prior to analysis using VAPAR. 

Project Snapshot:

  • Approximately 38,000m of wastewater pipe footage assessed.
  • More than 15,000 pipe features and defects identified and categorised using AI.
  • Collaboration between VAPAR & Western Water to quantify likelihood and consequence of failure.
  • Risk-based prioritisation resulted in a $2M refinement of the initial priority program estimate.
Featured SWAN Members: