Funding Opportunities

Cyber-Innovation for Sustainability Science and Engineering (CyberSEES)

Due Date: April 8, 2014


This Cyber-Innovation for Sustainability Science and Engineering (CyberSEES) solicitation aims to advance computing and information sciences research and infrastructure in tandem with other disciplines to develop analyses, methods, prototypes, and systems that lead to an increased understanding of sustainability and to solutions to sustainability challenges. Proposals are expected to forge interdisciplinary collaborations among the computer and information sciences, social and natural sciences, geosciences, mathematical sciences, engineering, and associated cyberinfrastructure to address challenging sustainability problems.

Computational challenges are woven into many areas of sustainability research and their solutions are opening up possibilities for radical new approaches and unforeseen opportunities. Close intellectual partnerships among computing and other disciplines are essential to inform new approaches to these integrative challenges, respecting the specific characteristics of complex phenomena, data, and models; understanding the tradeoffs and consequences inherent to design and decision processes; and effectively dealing with limitations and constraints of computational techniques. Some prominent underlying challenges are outlined below. These descriptions are meant only as examples of integrative challenges, not intended to span the scope of program interests.

  • Large-scale Data Analysis and Management. Data sets used in many sustainability-related topics are challenging because of their complexity (such as control variables in smart infrastructures), heterogeneity (such as social or behavior data), or variability and uncertainty (such as environmental sensor data). Additionally, sustainability data often have distinct short-term and long-term end uses, and data sets must be appropriately curated with these uses in mind. The sustainability context requires attention to issues such as the use, management, and preservation of data over a large range of time scales; data modeling and ontologies; metadata extraction, management, and data provenance; and cognitive prediction/optimization applied to data sets of differing ontologies and semantics, aiming for causal inference to inform effective policies, decisions, management and control.
  • Robust Observation, Sensing, and Inference. Monitoring of human, built and natural systems, such as hydrologic, waterway, or atmospheric phenomena; species migrations; water, gas, or electricity distribution systems; requires scalable observation infrastructures. Effective infrastructures must be nondisruptive, easily deployed and managed, and be robust under sensor failures, and/or satisfy applicable security or privacy requirements. These monitoring systems must be able to operate unattended for long periods of time yet be remotely configurable, controllable, and testable for correct operation. Interdisciplinary research is needed on integrated sensing, monitoring and modeling for decision making in relation to air, water and soil environments, as well as the optimization of complex sensor systems.
  • Modeling of Complex Systems. Behavior in systems facing sustainability challenges, from caribou migration to power generation, is the result of complex interactions between human, built, and natural systems. Modeling, simulating, analyzing, and optimizing these systems may require combinations of physics- or economics-based models, model-based reasoning, and statistical models built from data, or interactions involving entities or individuals with different and often conflicting interests. New interdisciplinary research is needed for accurate and efficient methods of tackling complexity, scaling, uncertainty quantification, reliability, coupling between systems, massively parallel architectures and constrained optimization in order to achieve needed performance.
  • Dynamic and Intelligent Decision Making. Solving sustainability problems, particularly those relating to systems near tipping points (such as extreme events), often requires contending with dynamically evolving or unstable situations. These scenarios present challenges to modeling and decision making due to inadequate or changing data; model, data, and system response uncertainty; resource constraints; fast response needs; and risk, safety and security concerns. A key characteristic of these environments is the need to continuously assimilate new data as it becomes available and to effectively handle new or altered constraints. Techniques are also needed to proactively discover such changing constraints within these environments. Additionally, both dynamic and static sustainability-related data sets can leverage intelligence and machine learning methods exploiting massive parallelism, scaling relations in universal nonlinear function approximations, and ladders of uninformative priors to enable open-minded inference in the presence of spatial complexity, leading to data-enabled discovery.
  • Control and Management of Infrastructure.Smart and sustainable management of built systems such as future electric grids, involving insertion of renewable sources such as wind and solar power, and other energy infrastructures; transportation systems; manufacturing systems; and smart homes and buildings, requires comprehensive and flexible control systems that satisfy performance, policy, reliability, and other constraints. The design and study of these systems will require bold new computing research and knowledge in areas such as resource management algorithms and architectures, systems analysis, real-time coordination and communications. For smart grids in particular, challenges include but are not limited to efficient and secure electrical power management at multiple scales from grid-level to personal systems, and integration of renewable energy resources and home energy systems into an aware and enabled electric grid. The human and organizational contexts of control systems must be well understood and reflected in system design, in order to make them usable and sustainable.
  • Human-centered Systems. Large and long-lived impacts on sustainability will depend on human behavior, human factors, and collaboration across large communities. Sociotechnical systems built on sound understanding of people, organizations, and how to inform and assist them, will play a key role in sustainability efforts. Computational approaches will lead to tools and systems that support engagement and decision making by the public; collecting, modeling, and presenting information about resource usage via usable interfaces, appropriate visualizations, and persuasive technology; preference elicitation and decision support/automated decision making for effective and efficient use of resources; and models, methods and tools for dissemination and increasing awareness of sustainable practices.
The interdisciplinary challenges considered above are invariably shaped by human, societal, and economic factors, requiring consideration of the human interface, security and privacy, socio-cultural norms, non-compliance, herding behavior, economic incentives, and amenability to deployment in the real world. Sustainable systems must be designed for transparency, legitimation, and participation, and foster greater awareness of sustainability challenges, solutions and best practices among the workforce and greater population.

As information and communication technologies are applied with increasing intensity to address sustainability issues, there is a parallel challenge of sustainability of the computing and networking systems themselves. In particular, as computing and communication technologies - from mobile phones to massive data centers - proliferate, challenges in managing their consumption of energy, materials, and other resources have become a critical sustainability issue. CyberSEES also welcomes interdisciplinary research that addresses holistic, integrative approaches to sustainable computing and information technologies and systems, including consideration of design and use with impact across the lifecycle in mind.

As the SEES research community increases in number of researchers and breadth of scientific participation, the shared computing and communication infrastructure must allow easy and timely sharing of data and computational tools to advance interdisciplinary SEES research. CyberSEES welcomes cyberinfrastructure research that connects currently distinct SEES research activities.