Announced at Google Cloud Next 2025, the new capabilities in Bentley’s Blyncsy product offering—which applies AI to crowdsourced imagery for automated roadway asset detection and inspection—will help infrastructure stewards improve roadway maintenance and support disaster recovery According to a Bentley press release, firm software will be able to provide highly detailed analysis of assets along with visual references, while Google’s Vertex AI adds the ability to build and maintain models to alert agencies of changes to infrastructure assets before they become safety hazards.

[….]

Bittman notes that hiring companies to inspect highways with specialized equipment could cost an average of $300 per mile. Blyncsy's new application could reduce that cost by at least half, he says. "Agencies were already collecting [Street View] images for another purpose," he says. That data, he adds, can be repurposed for decision-making on maintenance, upgrades and potential disaster recovery. Compared to the time it would take to collect data on an agency's road assets—often thousands of miles—the repurposing of existing Google data plus Bentley's data analytics capabilities could also save some 98% in time, he says.

"We’ve been able to take that same information to market at half the cost and deliver it in 98% less time, because it’s already been collected. And now, with Google, we have access to exponentially more infrastructure information," he says.

[…]

Looks like Google continues its world dominance. I wonder if a FOSS option exists for this type work. But it is cool to see an AI application that is somewhat useful in the construction industry.