MCCHE Precision Convergence Webinar Series with Flora Salim
Foundation Models for Time-Series and Spatio-Temporal Data
By Flora Salim
Professor University of New South Wales
Date: Friday, February 13, 2026
Time: 8:00 a.m. to 9:30 a.m.
Location: Online
Abstract
This talk explores the foundations of AI for time-series and multimodal sensor data, emphasizing the pressing challenges and frontier solutions for real-world spatio-temporal learning. Time-series data from sensors in domains such as transport, energy, and urban systems are often riddled with missing values, heterogeneity, irregular sampling, high noise, and label scarcity. These issues are compounded by modality differences across sensors, domain shifts, and dynamic environments. We present a comprehensive overview of recent advances, grounded in a series of foundational works. We also introduce a massive traffic forecasting, building IoT time-series, and human mobility datasets and benchmarks. and pretrained models for generalizable spatio-temporal inference across diverse urban contexts. We ground this discussion in broader trends outlined in a recent comprehensive survey on foundation models for spatio-temporal data science, which articulates how pretraining, cross-domain transfer, and unified architectures are reshaping the field.
In summary, this talk offers a unified vision of foundational AI for time-series and multimodal sensors, combining robust temporal modeling, cross-modal alignment, and scalable representation learning to unlock new capabilities in dynamic, real-world environments.