AMHERST, Mass. – Researchers from the University of Massachusetts Amherst will share in a $3.8 million grant from The Rockefeller Foundation to support their work to use enhanced data modeling to transform electricity systems in emerging economies. The grant will be shared with a team of researchers at Columbia University, Carnegie Mellon University and Colorado School of Mines.
The team will use the funding to launch the Electricity Growth and Use In Developing Economics (e-GUIDE) Initiative, an effort to apply data science to electricity demand prediction in energy-poor emerging economies. The Rockefeller Foundation has a dedicated focus on ending energy poverty and improving livelihoods by leveraging breakthroughs in data science and decentralized energy to both accelerate the pace of electrification and dramatically decrease the cost.
“Better electricity consumption predictions enable clear-eyed, informed planning of future electricity systems,” said Jay Taneja of UMass Amherst’s department of Electrical and Computer Engineering, and the project lead. “If you can’t measure it, you can’t manage it. Better data helps governments and international donors to better direct investments to expand electricity access and grow emerging economies.
At UMass, the work will be led by the Systems Toward Infrastructure Measurement and Analytics (STIMA) laboratory, which Taneja directs. The lab studies infrastructure in the developed and developing world, including energy and building systems, but also transportation, water, and sanitation systems. The laboratory uses embedded and mobile devices, machine learning, and controls.
While there are a variety of advanced tools available in Europe and the United States to accurately estimate electricity demand, the tools and methodologies needed to discover the best-fit methods for expanding access to reliable, affordable power in emerging markets are lacking. The e-GUIDE Initiative will partner with electricity service providers across Sub-Saharan Africa and in Asia to enhance the data available for planning and operating their systems, and enabling the roll-out of integrated electrification strategies, especially in rural areas where unreliable demand prediction has hindered universal electrification efforts.
The consortium is developing an openly available application programming interface (API) that will enable data and insights on electricity consumption growth to flow across borders and throughout the sector. The tool is powered by applying new machine learning techniques to geospatial data from satellites in conjunction with real electricity billing and consumption data from hundreds of thousands of emerging market commercial and residential customers.