The City of Guelph uses Waterlix Condition Assessment Map

October 21, 2019

The city of Guelph, Ontario uses Waterlix condition assessment to identify vulnerable locations in their water network and prioritize repairs. Using Waterlix artificial intelligence solution, the city can accurately measure the impact of replacing any water main.

 

The city of Guelph has a progressive asset management program and Waterlix contributed to the program by increasing the accuracy of the water main condition map. In this assessment the geographical data layers and satellite images were used to include the ground shift information in the analysis. 

 

A feature of the Guelph analysis was the relatively small amount of break history data on hand. AI models get stronger with increasing numbers of data points in order to identify a pattern. The problem was addressed by merging the historical break data from previous projects including data from the cities of Kitchener and London. We also added to the historical data for each pipe type (each material and size) which enhanced the accuracy of the Guelph model.

 

Using the newly enhanced model our condition assessment solution is capable of working in cities with any size and historical data. The models now have data sets for many pipe types and historical data from each city that better handle unusual or special cases. This added data gives additional weight when assessing pipes vulnerability.

 

After enhancing the Guelph model we identified specific sites that the risk of breaks (especially circumferential breaks) was 5 times more than other locations. This proved the importance of considering the geography in condition assessment models. Such Information will reduce the total cost of ownership for cities' assets and is a valuable source of knowledge for future developments. Our current model uses 76 additional factors and precision of our model in Guelph was 91%.

 

Through our other analysis efforts we found that having the break type in the data set increases the precision. The same model had an accuracy of 97% for the city of London because of the availability of break types.

 

It was a pleasure to work with a professional team in Guelph. Ms. Jessica Angers and Mr. Daryush Esmaili, who were instrumental supporting this project. 

Please reload

Our Recent Posts

Waterlix in The fifth National Water and Wastewater Conference

October 28, 2019

Waterlix Inc. in Communitech News

October 28, 2019

The City of Guelph uses Waterlix Condition Assessment Map

October 21, 2019

1/1
Please reload

Tags

Please reload