Blog

Axceta Featured in La Presse: Physical AI for Remote Construction Site Monitoring

Top photo: Christophe St-Pierre, partner at Excavation St-Pierre & Tremblay, and Sherif Zaroubi, CEO of Axceta, on an excavation site in La Prairie. Photo: Martin Chamberland, La Presse

We’re proud to be featured today in La Presse’s PME Inspiration section. Journalist Lila Dussault tells the story of how three Canadian SMEs turned a simple idea into a working Physical AI system on construction sites. Read the full article in La Presse (in French).

It started with a conversation

Excavation St-Pierre & Tremblay, founded in 1994, now employs more than 125 people and runs 15 to 20 job sites a week between Montréal and Sherbrooke. Its project managers spend a large share of their time driving from site to site.

In the 2026 off-season, partner Christophe St-Pierre looked for a way to monitor his excavator fleet remotely. He found rugged, self-cleaning cameras made in Canada by ExcelSense Technologies in British Columbia. But they couldn’t stream live video. ExcelSense connected him with Axceta.

What we built

Our team connected the ExcelSense cameras to a compact edge computer powered directly by the excavator. We then developed the software that streams live video and uses AI to flag objects and events that matter to site managers. Within a few months, prototypes were running on three excavators.

AI-enabled camera mounted on the railing of an excavator, with the operator in the cab
Photo: Martin Chamberland, La Presse

This is Physical AI as we practise it: sensors, embedded electronics, connectivity and AI working together on existing equipment, in mud, dust and Québec’s November weather.

Early results

  • Time and travel savings. Christophe St-Pierre estimates remote monitoring saves the company hundreds of thousands of dollars a year in project manager time and travel. He alone avoids two or three round trips of about 200 km each week.
  • Lower carbon footprint. Fewer trips mean fewer emissions. And the AI itself is lean: it doesn’t need large-scale model training.
  • Better visibility. Instead of a snapshot during a site visit, managers can follow progress on several sites continuously.
  • Respect for crews. The cameras record no sound and give a general 120° view of the work. They track overall progress, not individual workers.

What’s next

Excavation St-Pierre & Tremblay plans to install about ten more cameras on its fleet this year. On our side, we’ll keep improving the model as it learns from more sites. And we’re glad to show what three 100% Canadian SMEs can build together.

Have equipment that needs to see what’s happening in the field? Talk to us about your Physical AI project.

Related content

AgTech

Grow Yield and Increase
Productivity

Energy

Take Energy Management
to the Next Level

Mining

Optimize Mining
Operations