Predictive maintenance for a city fleet: the Kitchener pilot

Chosen as one of three companies from ten finalists at Communitech's Fast Track Cities showcase, TextGenetic AI won a $100,000 Pitch Kitchener pilot to apply AI-driven, real-time predictive analytics to the City of Kitchener's vehicle fleet.

The challenge

A municipal fleet is hundreds of vehicles and assets with different ages, duty cycles and maintenance histories. Breakdowns are expensive twice over: the repair itself, and the service that doesn't get delivered while a vehicle is off the road. Maintenance schedules based on fixed intervals either over-service healthy vehicles or miss the ones about to fail.

What we built

Our platform ingests historical work orders and telematics data and forecasts the likelihood of component failures, so crews act before a breakdown rather than after. It is predictive maintenance in the practical sense: a prioritized list of what to look at this week.

Flags potential maintenance issues before they become costly breakdowns
Reduces vehicle downtime and extends the life of city assets
Lowers operating costs and supports environmental goals

Why it matters beyond fleets

The pilot is where our forecasting engine earned its credibility in live operations. The Government of Canada then selected the technology for its Innovative Solutions Canada Testing Stream, a program that tests late-stage innovations (TRL 7–9) in real federal environments. The same engine — and the same discipline of measured, reviewable prediction — now powers our neuroanalytics work in brain health.

Productized as Axion Fleet

The pilot has been productized as Axion Fleet: asset lifecycle, automated work orders and predictive maintenance for municipal and commercial fleets.