001-40 WC 143 JULY-AUG25 PT - Flipbook - Page 24
WASTEWATER
grease build-up, offset joints, and structural deformation.
The system uses computer vision and machine learning models
trained on a massive dataset of more than 600,000 previously
coded inspections. As the AI “learns” from expert-coded examples,
it becomes increasingly accurate at flagging both structural and
operational defects according to PACP condition grades.
“We leveraged over 600,000 inspections and trained the AI
model to see patterns,” says Ben Gready, AI specialist with the
City of Edmonton. “By focusing on the lowest-risk factors, our
experts could concentrate on the most important ones.”
Once the footage is processed, the AI outputs a preliminary
condition report, complete with time-stamped defect locations
and severity scores. Human reviewers then validate only higherrisk segments, significantly reducing the time required for full
inspection analysis.
Since implementation, EPCOR reports a 37 per cent reduction
in time spent on low-risk reviews, freeing up capacity for more
critical assessments and accelerating capital planning and
maintenance decisions.
water, listening as it goes. As it moves through the system,
the device uses sensitive acoustic sensors to detect the highfrequency sounds associated with escaping water or trapped air.
With the help of tracking sensors along the pipeline, its location
is simultaneously recorded, allowing engineers to map the
precise site of each potential defect once the device is retrieved
and the data analyzed.
During deployments in 2018 and 2019, SmartBall helped
Calgary identify several previously undetected issues—including
a rare circle break in a large-diameter concrete watermain—
before they escalated into major failures.
These targeted findings allowed the city to prioritize repairs
with precision, strengthening infrastructure resilience while
avoiding costly excavations, unplanned service interruptions, or
emergency response measures.
HAMILTON
Advancing CSO monitoring and sewer inspections
CALGARY
Inline acoustic inspection with SmartBall technology
Technology providers: Xylem Smart Infrastructure
Project status: Ongoing
Application: Leak and gas pocket detection in pressurized watermains
Technology providers: City of Hamilton
Project status: CSO real-time monitoring and inspection program in
early stages of expansion
Application: Overflow reporting, targeted inspections, infrastructure
planning
WHILE EDMONTON’S AI scans video footage for trouble, Calgary
is listening for it—literally. The City’s water utility has adopted
inline leak detection technologies from several reputable vendors,
one of which is Xylem’s SmartBall—an acoustic sensor device that
travels through active pressurized watermains to detect leaks and
gas pockets without interrupting service.
About the size of a baseball, the SmartBall slips into a live
pipeline through an existing access point and travels with the
24
WATER C AN ADA • JULY/AUGUS T 2025
for water quality in Hamilton—particularly during storms,
when wastewater and rainwater overwhelm the system and
discharge directly into local waterways. To manage this issue, the
City has implemented a robust real-time CSO reporting system,
with ongoing efforts to strengthen and expand its capabilities.
While Hamilton has begun installing new monitoring infrastructure at CSO outfall locations, the system is still evolving.
Real-time data is available, but its operational integration—such
as informing inspections or predictive responses—is still largely
WAT E R C A N A D A . N E T
Getty Images
COMBINED SEWER OVERFLOWS (CSOS) have long posed a challenge
