Identification of Mains Most at Risk of Failure

Region of Peel, Ontario


The client – the Regional Municipality of Peel, in Ontario, Canada, has implemented a continuous improvement strategy to proactively manage its water infrastructure in response to population increase and the resulting ever-growing demands on the water supply system.

As part of this proactive approach, the Region has collaborated with CANN Forecast to leverage innovative methods such as Artificial Intelligence and Machine Learning to identify the water mains cohorts that are most at risk of failure, allowing decision-makers to predict more accurately the remaining life of the assets and to prioritize renewal programs. In addition, this predictive modeling information will also be used to plan water main inspection and replacement programs, allowing staff to maximize the service life of their linear assets and optimize the available replacement budget.


InteliPipes is a decision-support system based on Machine Learning that identifies the most at-risk pipe cohorts within a municipality’s water network. Once these vulnerable groups are identified, it allows utility managers to :

(a) have a better understanding of the network’s degradation over time;

(b) tailor inspection plans and replacement programs;

(c) optimize water main investments in order to provide an improved level of service.

«  We came across CANN Forecast Intelipipes’ algorithm and right away realized that this system will be able to meet our needs. The data science team at CANN Forecast have delivered on this big time and help us accurately forecast watermain cohorts that we should be proactively targeting for inspection, maintenance and replacement thus help us improve the level of service to our customers. »

Imran Motala, P.Eng
Manager, Water and Wastewater Asset Management, Region of Peel



In total, InteliPipes’ cohorts-based algorithm was able to identify 18 km of linear water main assets – representing 0.39% of the Region’s water network – that have experienced an average break rate greater than 82 breaks/100km/year and a yearly likelihood of failure between 21% and 41% for the 2016-2020 period. Given the excellent overall risk index of the Region of Peel – a global break rate of 2.6 breaks/100km/year and a yearly likelihood of failure of 0.26% – prioritizing this 18 km of assets will allow for the most efficient usage of available funds and resources.