AI-Powered Smart Grids

Smart Grids are the most single important part of smart cities and factories. One of the main concern of of AITOWN is to to to design reliable, secure and resilient multi-networked   solutions to empower Smart Grids with 

Electric Grid Event Detection Using Heterogeneous and Spatiotemporal Data

Traditionally electrical networks have been treated mainly as physical entities that connect electricity suppliers to consumers. However, a modern grid is empowered by the internet of things, distributed generation, and networked computational subsystems to support the incorporation of renewable energy resources, electric vehicles, and energy markets. The induced dynamic and stochastic due to the new paradigm in smart grids require high-resolution measurement and agile decision support techniques for system diagnosis and control. Where muli-stream measurement data come from various sensors in various locations of the network, classical machine learning methods are not the ultimate solution. We are developing multi-task and scalable learning machines for event detection in societal scale systems such electric grids. 

Data-Driven Resilience in Smart Cities

Characterizing Interoperability of Multi Networks for Resilience

Adversery impacts of large-scale natural and man-made disasters are primarily the result of the inability of city’s components such as trasnposration and electricity networks to efficiently cope with random and dynamic changes, which translated into resilience deficiencies while exposing gaps in data availability and data analytics. Therefore, providing optimal and fast infrastructure restoration solutions based on lessons learned and data gathered are critical for governments and emergency responders. However, lack of data due to the shortcoming of monitoring systems in cities and the scarce nature of disasters remains a big challenge for data-driven approaches. We are developing inference and estimation algorithms that can fill the gap in data sparsity and provide integrated resilience for urban infrastructure.