Assistant Professor and BBISS Fellow, Bistra Dilkina, is the Georgia Tech lead investigator for a $10 million, 5 year National Science Foundation “Expeditions in Computing” grant.  The funds will establish the Computational Sustainability Network, or CompSusNet, with Cornell University as lead institution.  CompSusNet will be comprised of 12 academic institutions, as well as domestic and international, private and non-profit organizations.  Dilkina will also sit on the executive council for the Expedition. 

Computational sustainability is a young field of study.  Highly trans-disciplinary research teams focus on mathematical and computational models to aid in management and decision making for the major challenges related to sustainability, such as poverty mitigation, renewable energy, and biodiversity conservation.  Computational sustainability problems tend to be dynamic and complex, often utilizing combinatorial decision making algorithms, big data, citizen science, dynamical models, machine learning, and optimization.

CompSusNet is one of three Expeditions in Computing grants awarded by the NSF.  These grants represent some of the NSF’s largest investments in computer science research.  The two other 2015 Expeditions in Computing grants went to “The Science of Deep Specification” at Princeton University, which aims to eliminate software “bugs” that can lead to security vulnerabilities, and “The Evolvable Living Computer Project” at Boston University, which will study which computing principles can be applied repeatedly and reliably to synthetic biology.

Dilkina’s expertise and research interests, as well as the resources available at Georgia Tech (such as the Institute for Data and High Performance Computing), dovetail well with goals of the Computational Sustainability Network grant.  Dilkina’s project focuses on developing decision support models to optimize biodiversity conservation by balancing habitat connectivity with other urban planning considerations.  As human-centered landscapes become more developed, wildlife habitat becomes fragmented and isolated.  Maintaining landscape connectivity has been shown to be an effective strategy to preserve biodiversity, however it is extremely complex to implement. Together with ecologists and computer scientists at Cornell University, Dilkina will develop a habitat connectivity-focused conservation plan for the Andean bear populations in Ecuador.  However, the models and techniques that will be developed will not necessarily be specific to species, ecosystem, or spatial scale.

Read more BBISS Big Ideas Stories.