Team Led by ECE’s Siami Awarded $1.1M by NSF for Multi-System Modeling Framework
ECE/EAI Assistant Professor Milad Siami is leading a multi-university $1.1M grant from the National Science Foundation, in collaboration with ECE Dennis Picard Trustee Professor Mario Sznaier and Assistant Professor S. Farokh Atashzar from the NYU Tandon School of Engineering, for “Modeling and Control of Non-Passive Networks with Distributed Time-Delays: Application in Epidemic Control.” Motivated by the ongoing COVID-19 pandemic, this project seeks to develop an end-to-end framework for synthesizing robust, data-driven control laws for complex interconnected systems subject to uncertain time delays.
Milad Siami, assistant professor, electrical and computer engineering (ECE), is leading a multi-university team in the creation of a comprehensive framework that could help model and predict complex interconnected systems—such as the continued spread of the ongoing COVID-19 pandemic.
The team was awarded a $1.1 million grant from the National Science Foundation (NSF) to develop an end-to-end framework for synthesizing robust, data-driven control laws for complex interconnected systems subject to uncertain time delays.
“With the spread of COVID, we’re dealing with a large-scale network and many different time delays, including the time it takes to detect, the time to make other people sick, and the time to recover,” says Siami. “These effects lead to ‘reflective’ spreading waves, resulting in multiple phases of infection spread that are both hard to predict and control.”
The time element is what makes this team’s work different from other modeling and prediction methods. The project, titled “Modeling and Control of Non-Passive Networks with Distributed Time-Delays: Application in Epidemic Control,” seeks to provide accurate spatio-temporal predictions of the successive echoing waves of the spread between various population clusters.
While Siami’s expertise is in modeling and control of networks, the team also includes collaborators Mario Sznaier, Dennis Picard Trustee Professor, ECE, who works with systems identification and semi algebraic optimization; Peter Boynton, CEO of the George J. Kostas Research Institute for Homeland Security at Northeastern, with expertise in emergency management and bridging research and practice; and S. Farokh Atashzar from the NYU Tandon School of Engineering.
The end goal of the framework is to help individuals and stakeholders prepare for new strains and waves of COVID using the resources available to them, such as masks or vaccines, while also finding the right timing to balance the tradeoff in terms of cost and frustration. If predictions are presented accurately and at the right time, the team’s model could mitigate the spread of the disease for years to come.
“Because there are so many types of data and parameters that all contribute to the spread of COVID, we can use less data,” says Siami. “Our framework is intended to use a scalable algorithm to be able to make predictions wisely and using less information, while still offering robust and trustworthy results.”
Once developed, the framework could also be used for many other multi-system, dynamically connected problems, from the spread of other airborne illnesses to the use of new technology or the distribution of less-than-credible news stories.
Abstract Source: NSF