# Introduction I am Gregor Kos, a Senior Lecturer at # Concordia University in # Montreal ,
| Source: Mastodon | Original article
Gregor Kos, a senior lecturer in Chemistry and Biochemistry at Concordia University, announced the launch of a hyper‑local urban air‑quality research platform that blends machine‑learning techniques with chemical analysis. The initiative, unveiled at a university press briefing on Monday, will deploy a dense network of low‑cost sensors across Montreal’s downtown core and feed real‑time measurements of pollutants into predictive models that Kos and his graduate team have been refining in his “Machine Learning for Chemists” course.
The project matters because it moves air‑quality monitoring from city‑wide averages to street‑level granularity, exposing exposure hotspots that conventional stations miss. By coupling isotope‑ratio data and n‑alkane profiling—areas where Kos has published over 40 papers—with probabilistic ML algorithms, the platform aims to forecast short‑term pollution spikes and identify their chemical sources. Such fine‑scaled insight could inform municipal traffic‑management policies, guide public‑health advisories, and provide a testbed for AI‑driven environmental stewardship in other Nordic and North‑American cities.
Kos will release the first batch of sensor data and model code on an open‑source GitHub repository by the end of the quarter, inviting collaboration from the broader AI‑for‑science community. The university has secured a three‑year, CAD$2 million grant from the Canada Foundation for Innovation, and a partnership with the City of Montreal’s environmental department is already in place. Watch for a pilot study report slated for September, which will compare the platform’s forecasts against traditional monitoring stations, and for potential scaling talks with Stockholm’s urban sustainability office later this year.
Sources
Back to AIPULSEN