Today, an unprecedented amount of data is available regarding how the world uses energy. This avalanche of data presents both challenges and opportunities for engaging in energy data analytics, understanding how individuals and companies make energy decisions, and improving efficiency.
Matthew Harding
In his complementary roles as an economist and a data scientist, Matthew Harding works to understand how people make energy consumption choices and to quantify how those choices impact individuals and the environment. Harding’s research is driven by the most significant modernization in history to the United States’ electric grid. Power companies, in collaboration with federal, state, and local governments, are installing advanced metering technologies, such as smart meters, on a large scale. Additionally, many new appliances include sensors that provide real-time, two-way communication of energy consumption and a wide range of other measures such as temperature or run-time data.
With this technological renaissance under way, Harding, an assistant professor in the Sanford School of Public Policy and a faculty fellow in the Duke Energy Initiative, can employ the massive amount of data captured by these meters and sensors to create predictive models of energy consumption. These models could be used, for example, to improve energy efficiency—what Harding describes as a “triple win”— an outcome that benefits individuals, companies, and society. This will help inform broad rollouts of smart meter technology.
In 2014, iiD provided seed funding to several data projects in collaboration with the Energy Initiative. Harding is working with Richard Newell, the director of the Energy Initiative and former head of the U.S. Energy Information Administration, to connect multiple Duke researchers, projects, data sets, and tools related to energy data analytics to form an Energy Data Analytics Lab. In one ongoing effort, faculty members are analyzing huge data sets on electricity consumption to investigate the factors that lead to load shifting—switching generation capacity between peak and off-peak periods or across geography to prevent expensive outages. At the same time, Duke researchers are exploring new ways to model, understand, and put to use the complex data layers that are created when numerous energy consuming and generating machines communicate with each other.
“Our Energy Data Analytics Lab, directed by Matt Harding, will provide new ways to connect Duke expertise in energy with external partners in the business and policy communities.” —Richard Newell, Director, Duke University Energy Initiative