The Laboratory for Environmental Computing and Decision Making (LECDM) is a collaborative, multi-disciplinary research effort aimed at creating cutting-edge computing technologies, using these technologies to build and integrate novel environmental models, and using those models to improve environmental decision-making in the public and private sectors. Our current research focus is on creating the technologies and models to support policy analysis in the transportation sector. Our overall research theme is modeling and cyberinfrastructure for sustainability policy analysis.
The research opportunity for the lab is driven by the scope of application domains that must be modeled, the diversity of modeling styles and technologies, and the need to flexibly and quickly integrate these diverse models into cooperating wholes and driving those models through a variety of policy scenarios. The application domain disciplines include:
- Geospatial modeling of multi-modal (truck, rail, water) freight transportation and its impact on greenhouse gas and other emissions
- Geospatial modeling of pollution emissions, emissions dispersion, and resulting impact on human health and the environment
- Materials flows life-cycle modeling of the emissions footprint of various products, such as conventional and alternative fuel vehicles
- Econometric modeling of consumer demand for vehicles
- Modeling of automobile producer decision-making, balancing consumer demand, competitive pressure, and public policy
- Geospatial and statistical modeling and analysis of energy resources on U.S. public lands
- Modeling of the interconnected system dynamics of material supply, vehicle production, vehicle purchase, vehicle use, and vehicle retirement (recycle and waste) under various possible future scenarios and public policy options (taxes, rebates, subsidies, research investment, etc.)
The diversity of modeling styles and technologies we currently use include:
- Analytical models (algebraic equations)
- Numerical methods (simulation, optimization, stochastic analysis)
- Agent-based simulation
- Game-theoretic models
- Discrete event simulation
From a computing cyberinfrastructure perspective, the research challenge is to integrate the very large scope and complexity of the domain discipline models and modeling styles and technologies into an integrated meta-system of models that are useable and understandable by decision-makers (producers, consumers, and policy makers). For model integration there are numerous existing approaches that solve part of the problem (such as, shared data using common, standard data models; service-oriented architectures; data integration and data mining; dynamic simulation) and numerous computing and programming technologies (we must integrate existing models developed in Matlab, Mathematica, ProModel, VenSim, Microsoft Excel, C, C++, Java, etc. and get them to run concurrently on distributed servers, grid clusters, etc.). However, no single approach solves the diversity of models and modeling styles we face, and there are no known combinations of computing architecture styles that address the integrated whole. Further, we need to discover ways for the decision-makers to interact with, configure, execute, and understand the results of this large model of models when they are not experts in any one of the models. Each model currently has its own visualization and human interaction approach, but there is no known way to provide an integrated visualization and human interaction across the models.
We have brought our research into the classroom through Public Policy courses, Environmental Science courses, Software Engineering undergraduate student projects, and Information Technology graduate student projects. We are discussing the possibility to provide additional courses and possibly tracks in the GCCIS Ph.D. and the Sustainability Ph.D. programs and other graduate and undergraduate programs.
Even though the LECDM has been in existence for only a year, and the initiative is only two years old, we have already built a strong team with a strong, multi-disciplinary, collaborative research mission and good momentum in accomplishing that mission. We already have achieved significant success in funding, publications, and education. We intend to continue that success and enhance it through collaboration with other RIT research efforts and external university, government, and corporate laboratories. We intend to become a recognized leader in research on modeling and cyberinfrastructure for sustainability policy analysis.