The project entitled ‘Spatiotemporal Models for Predicting Delays in Transportation Networks during Extreme Weather Events’ is a joint collaboration project between Dr Simon Hu at Zhejiang University and Dr Huy Tran at Aerspace Engineering at University Illinois at Urbana Champaign.
Millions of people depend on transportation networks on a daily basis. However, increasingly frequent and uncertain extreme weather events continue to disrupt their service. For example, Hurricane Sandy affected between 13,500 and 17,500 flights within 16 major US airports daily. Increasingly available transportation data may help mitigate disruption impacts through data-driven modelling and prediction capabilities. Challenges to fully realizing these capabilities include the need to handle large spatial scales, complex temporal dynamics, and emergent behaviors stemming from network connectivity. The project aims to address these challenges by developing machine learning algorithms that cohesively in-corporate weather, temporal, and network-level features to predict impacts of extreme weather events on transportation networks. We will focus on predicting flight delays in air transportation networks (ATNs), due to projected growth in air travel; e.g. air travel in China is expected to grow 6.4% annually over the next decade. We will also consider application to rail networks, along with explora-tion of multi-modal networks utilizing both air and rail links.