Kitchener company can find 'even the smallest crack' in bridges and roads using AI technology

Joe Pavia · CBC News

· Posted: May 13, 2025 6:00 AM EDT | Last Updated: 1 infinitesimal ago

A Kitchener, Ont., tech startup is utilizing drones and artificial quality exertion to inspect bridges and roads and connection solutions for repairing them. 

Albert Mansour, the CEO and co-founder of ConeLabs technology, says his company's exertion volition item ocular defects and drawback problems with a operation that tin hopefully beryllium fixed earlier the occupation gets excessively serious.

"We look for immoderate clues. We look for cracks, spalls [breaks], things that springiness america indications of the information of the structure," Mansour said.

"Our AI finds it all. So present we person a implicit knowing of wherever adjacent the smallest ace is. Even if that ace doesn't warrant immoderate enactment today, we astatine slightest present person it tracked, identified."

Mansour says their exertion is meant to assistance the occupation done by quality inspectors who oversee the 53,000 bridges successful Canada and astir 620,000 located successful the United States.

How they cod the data

ConeLabs uses drones to scan thousands of images of a span oregon roadworthy past bundle helps them reconstruct an close photogrammetric 3D exemplary in high solution to observe immoderate issues.

"We tin get capable coverage, resolution, get adjacent capable to the aboveground to get the close imagery. So drones travel out, they seizure images and past we reconstruct that into a 3D model," Mansour said.

"Then our AI wrong the 3D information finds these cracks, finds the defects, finds spalls and rust and processes that. So, it's a overmuch richer knowing of wherever each ace is versus conscionable a zoomed successful picture. And if determination are, the spatial positioning of 1 defect to different due to the fact that that mightiness springiness america clues arsenic to a deeper contented that's happening."

Monsour says the accusation they stitchery volition let a municipality to prevention clip and enactment aboriginal to hole immoderate issues with the operation they person tested and destruct oregon debar "long drawn retired repairs that hap during summertime construction."

The ConeLabs co-founder hopes to 1 time grow their inspections to besides see different larger structures including buildings, workplace silos arsenic good arsenic runways and railways.

Mansour hopes to grow its inspection enactment to see much than bridges and roads.

"Anything that needs changeless inspection, we tin bid our AI to inspect that asset," said Mansour.

"Buildings, runways, and different ample [structures] similar silos, powerfulness lines and railways. The accidental is immense and abundant."

Company portion of Pitch Kitchener initiative

ConeLabs was selected to collaborate with the metropolis during a Pitch Kitchener imitative. Three companies were selected to pilot their technologies to let unit to measure their imaginable for imperishable usage by the city.

That announcement was successful March and astatine the time, Kitchener Mayor Berry Vrbanovic said the $100,000 successful backing for the 3 companies was a mode for the metropolis to assistance "grow and enactment locally-grown exertion companies."

"Through efforts similar Pitch Kitchener and the precocious launched Support Local campaign, we are ensuring that our section system remains beardown and future-ready," Vrbanovic said successful a release.

The different 2 companies chosen arsenic portion of the inaugural were Real Life Robotics and TextGenetic.

Real Life Robotics will usage senors and AI-driven instrumentality learning to analyse the operations astatine Doon Valley Golf Course. TextGenetic volition usage an AI level to emblem imaginable issues with the city's fleet of vehicles aboriginal with an purpose of lowering wide attraction costs.