Institute for Computational Sustainability Hosts Inaugural Conference

The inaugural conference in the new field of Computational Sustainability brought a wide variety of researchers to Ithaca, including biologists, conservation workers, economists, computer sciences and operations researchers.   The conference provided an opportunity for researchers in sustainability to learn about modeling and computational techniques, and for computer scientists and operations researchers to learn about computational aspects of challenges in sustainability.   Turnout for the conference was much larger than expected, according to Carla Gomes, Professor of Computer Science and Director of the institute.  ORIE's Professor David Shmoys, Associate Director of the Institute, played a major role in organizing the conference.

Over the course of four days, formal presentations, poster sessions, panel discussions and working group sessions covered a range of topics from interactive decision support tools to challenges in environmental monitoring.   According to Gomes, the major application areas for the institute and the conference are conservation and biodiversity, balancing socio-economic demands and environment, and renewable energy.  A variety of computational themes, based on a "transformative synthesis" of areas from the science of computation to dynamics, optimization and statistical modeling to agent integration can be brought to bear on these application areas, she said.

Among the speakers on operations research topics, Princeton Professor Warren Powell (whose former Ph.D. students Huseyin Topaloglu and Peter Frazier are now on the ORIE faculty) spoke on the use of "approximate dynamic programming for a stochastic, multiscale energy policy model," and ORIE Professor David Shmoys discussed "mathematical programming-based heuristics: telecommunication network design meets species distribution planning."    Shmoys discussed the established operations research problem of designing survivable telecommunications networks and showed that models for this problem can be adapted to problems in the survival of species. As an example, he discussed an optimization model for planning long-term acquisition of woodlands as a means of managing red-cockaded woodpeckers, a threatened species.   In one of the poster sessions at the conference, recent ORIE B.S. graduate Adam Elmachtoub, Computer Science Ph.D. students Dan Sheldon and Bistra Dilkina and Ryan Finseth, an M.S. student in Applied Economics and Management, presented a poster on their collaborative work with Shmoys and others on this model.

In their poster presentation, Finseth and Elmachtoub pointed out that the red-cockaded woodpecker is a so-called "keystone" species that helps create nesting sites for at least 26 other vertebrate species, but has been in decline due to the degradation of the longleaf pine ecosystem in which the live.  The Conservation Fund and its partners are buying land tracts adjacent to current surviving populations of the woodpecker in North Carolina  in order to aid in the recovery of the woodpecker species.  To help the Fund, the team (which included Fund representatives) developed a large scale network flow model, based on the spread of influence in social networks, to represent the spread over time of a species through adjacent territories.   For any specific set of territories they were able to simulate the spread and so determine the properties of the system under different odds-driven scenarios.  They then used an optimization model to compute how most effectively to use a limited budget in maximizing the number of territories that have active woodpecker colonies after a specified number of years have elapsed.

In an introductory overview of the conference, Gomes said "I believe that Computational Sustainability is a fundamentally new intellectual territory with great potential to advance the different disciplines involved and with unique societal benefits."

According to an article in the Communications of the ACM, conference attendee Michael Runge, a research ecologist at the U.S. Geological Survey's Patuxent Wildlife Research Center, said he and his colleagues had believed there were no solutions to many of the complex ways they wanted to formulate ecological decision problems. "I've realized that we were over-constraining how we were thinking about problems," he said. "I've had my eyes opened to the number of tools available from the mathematics and computational side. The question is: How do we connect these amazing tools and the huge demand for their application to ecological problems?  We need people to bridge communication between all these fields, people who can see that a disease dynamics or water supply contamination problem looks a lot like a telecommunications network problem," said Runge.