K. Max Zhang, Associate Professor in the Sibley School of Mechanical and Aerospace Engineering, led a team of Cornell Engineering and Computer Science graduate students—consisting of Ye Jiang (CS, M.Eng.), Vignesh Rao (CS, M.Eng.), and Jeff Sward (MAE, Ph.D.)—to win the Environmental Protection Agency (EPA)’s first-ever EmPOWER Air Data Challenge.
As part of their collaborative project, Zhang worked with Computer Science (CS) and Mechanical and Aerospace Engineering (MAE) students “to create a machine learning model to predict NOX, SO2, and CO2 emission rates of fossil fuel-fired electricity generating units.” As Zhang explains: “We showed in our submission that this model can be further improved to identify anomalies in CAMD data,” and added “I have enjoyed working with CS students, and am hoping to recruit more CS students to work on this problem in the next academic year.”
The EmPOWER Air Data Challenge aims to advance the knowledge, use, and understanding of the EPA’s Clean Air Markets Division (CAMD) data and related information to benefit the environment and the public through innovative and creative projects.
The EPA reports that “the majority of U.S. fossil fuel-fired electricity generating units submit quarterly reports on their hourly nitrogen oxides (NOX), sulfur dioxide (SO2), and carbon dioxide (CO2) emissions along with select operating parameters (e.g., hourly heat input, gross electricity generation).” So it is the EPA’s Clean Air Markets Division (CAMD) that “makes these data available through a variety of online reports and applications.” At present, the CAMD is looking to academic and research institutions to propose practical and replicable projects that benefit academic institutions, the environment, and the public. The EmPOWER Air Data Challengeis part of the EPA’s desire to support “innovative and creative uses of CAMD data” that can “advance the knowledge, use, and understanding of [the] CAMD data and related information."
See also Daniel Aloi's coverage in the Cornell Chronicle:
