ľ¹Ï¸£ÀûÓ°ÊÓ

Summary of dissertation

Air pollution varies dramatically from street to street - what you breathe can change significantly within just a few street blocks. Traditional monitoring stations are too widely spaced to capture these hyperlocal differences. My PhD research applied mobile monitoring, by mounted sensors on Google Street View cars, to reveal invisible air pollution distribution patterns at an unprecedented 50-meter street-level resolution across Amsterdam, Copenhagen, and Rotterdam.

By combining AI with street-view imagery, my models have uncovered surprising urban pollution hotspots. For instance, neighborhoods with more chimneys and intersections with traffic lights consistently show higher pollution levels. The produced high-resolution pollution maps, available for different times of day, empower millions of residents to make informed decisions about their daily routines, from choosing less polluted commuting routes to scheduling outdoor activities when air quality is better.

Beyond air quality, I've expanded this mobile sensing approach to map urban noise and heat, demonstrating how a single monitoring system can track multiple environmental risks. This work lays the foundation for what I call an "Exposure Fundamental Model" - a future AI-powered system that could comprehensively monitor various urban environmental hazards simultaneously.

My research provides city planners, policymakers, and public health experts with powerful new tools to understand and mitigate urban environmental risks. These detailed exposure maps enable targeted interventions to protect vulnerable populations and create healthier cities. By making invisible environmental threats visible, we can work toward urban spaces that are not just more livable, but actively health-promoting.

PLEASE NOTE: If a candidate gives a layman's talk, the live stream will start 15 minutes earlier.