News · August 21, 2026

UNB's Machine Learning Health Vault Wins CFI Infrastructure Funding

A $97,026 Canada Foundation for Innovation award will help Scott Bateman's team build privacy-preserving synthetic health data — part of $397,026 across three UNB projects announced August 20.

By NB Tech News Staff · 3 min read

Abstract vault door beside a synthetic data lattice for privacy-preserving health research

The University of New Brunswick landed infrastructure money from the Canada Foundation for Innovation for three campus projects — and the most tech-facing of them is a Machine Learning Health Vault meant to produce realistic health data that is not anyone's real chart.

UNB's newsroom posted the campus breakdown August 20, the same day CFI announced $118,188,220 for 357 projects at 68 postsecondary institutions nationally. All three UNB awards come through the John R. Evans Leaders Fund, the CFI stream for equipping individual researchers — 334 of the 357 projects, and $97.5 million of the total — rather than the College Fund that covers the rest. UNB's share is $397,026 from CFI, plus institutional and provincial matching on top.

Synthetic health data without the patient

Dr. Scott Bateman (Computer Science; Data Science and AI Research Institute) leads the Health Vault with co-applicants Dr. Stijn De Baerdemacker (Chemistry; same institute) and Dr. Ted McDonald (Political Science; DataNB). CFI's share is $97,026, with another $48,513 from UNB.

The project combines advanced AI with administrative-data research to generate synthetic health data — records built from detailed patterns in real health information without corresponding to any real person's file. The goal, per UNB, is data-informed policy and patient-care decisions made without risking exposure of New Brunswickers' personal health information.

That is the tradeoff administrative-data shops have been chasing for years. It matters here because the province's health data already sits in production systems and patient-facing apps — the same records surfaced through tools like MyHealthNB — and researchers who want to model on them run into privacy limits long before they run into compute limits. A synthetic layer is the standard workaround; the CFI award funds the machinery to build one against New Brunswick's own patterns.

McDonald's involvement is the tell on scope: DataNB is UNB's administrative data research centre, the shop that already works with linked provincial records. Pairing it with the Data Science and AI Research Institute puts the modelling and the data governance in the same project rather than in two solitudes.

Microplastics and aluminum dross, same envelope

The other two awards are engineering and chemistry infrastructure with clear provincial stakes:

  • Dr. Zunaira Asif and co-applicant Dr. Ikrema Hassan (Engineering) receive $150,000 (plus $14,616 from UNB) for "From Source to Sink," tracing microplastics across air, water, and sediment with drones, Raman microscopy, and advanced computing — aimed at the standardized datasets that environmental monitoring still lacks for the smallest particles.
  • Dr. Barry Blight (Chemistry) and Dr. Kyle Rogers (Chemical Engineering) receive $150,000 from CFI to build a lab that scales their aluminum dross digestion technique, which cuts refinery waste while producing specialty materials and zero-carbon fuel. That project also draws $133,349 from UNB and $150,000 from ResearchNB.

Vice-president research Dr. David MaGee tied the three together in the UNB release: the CFI money supports researchers "in helping us collectively address the growing impact of microplastics, make better healthcare decisions while protecting privacy, and improve the environmental sustainability of manufacturing and energy systems."

For the provincial tech beat, the Health Vault is the piece that ties campus AI capacity to DataNB's administrative-data world — with a dated dollar figure attached.

Sources

Related coverage

Tags: unb, cfi, health-data, privacy, ai, fredericton, datanb, research