EEme, an Analytics-as-a-Service (AaaS) company, is pleased to announce a breakthrough in Advanced Metering Infrastructure (AMI) - based load disaggregation. Using AMI data collected in 1-second intervals from 10 homes over 77 weeks —without additional hardware, input or user training of algorithms — EEme achieved average hourly accuracy figures ranging between 90-99%. The results were validated by Pecan Street Inc., a 501(c)(3) applied research and commercialization institute. Having set a new industry standard that all utility industry stakeholders will benchmark against, this groundbreaking technology is available now from EEme as its “Disaggregation-as-a-Service (DaaS)” offering.
“Energy disaggregation technologies have suffered from a lack of meaningful validation studies - it is great to see EEme publicly report independently verified performance metrics for their algorithms on a dataset from the Pecan Street Project,” says Professor Mario Berges who directs the IBM Smart Infrastructure Analytics Laboratory and Intelligent Infrastructure Research Laboratory at Carnegie Mellon University. “With this unprecedented success using readily available 1-second data collected directly at the meter, load disaggregation will finally tap its full potential for near-real-time applications in the grid edge world,” asserts Dr. Enes Hosgor, Chief Executive Officer and Founder of EEme.
According to Dr. Hosgor, “This marks an extraordinary improvement over other load disaggregation solutions, and provides a new tool for energy companies who are working to improve energy efficiency and demand response programs. Scalable and accurate AMI-based load disaggregation will reduce operational uncertainties in the distribution system and provide value-added services to customers by extracting new and meaningful insights from existing AMI”.
This groundbreaking accomplishment will save considerable additional investment for energy companies, enabling accurate data-driven insights to support energy efficiency, demand response, and operational efficiency at a fraction of the cost over full-scale local end-use monitoring systems that rely on additional monitoring hardware or smart home devices and networks. It also leverages the existing AMI investment that has been made by most utilities, adding greater value to an already available data stream. Provisions must be made for the retrieval of 1-second data, but most systems are capable of providing such data through configuration settings at the head-end or through other means.
This breakthrough enables close to real-time functions that have strong advantages for today’s energy economy. 1-second-based load disaggregation enables use cases important to both energy companies and consumers, including appliance diagnostics; dynamic energy feedback, e.g., behavioral and physical demand response; user alerts in the broader home automation context, e.g., “You pool pump may be malfunctioning”; in-situ load-balancing in concert with distributed energy resources, among other uses. Moreover, the Evaluation, Measurement and Verification (EM&V) community can leverage highly accurate 1-second-based load disaggregation to streamline their activities in a software-driven manner and avoid deploying costly end-use monitoring hardware to validate energy savings. This breakthrough capability will also enable a smooth transition into full building control and automation in the transactive energy world.
Beyond its apparent improved accuracy benefits, leveraging 1-second-based disaggregation directly at the meter opens up possibilities in grid-edge computing that does not burden the basic functions of the Meter Data Management System’s (MDMS) daily data transfer cycles.
EEme, a Pittsburgh-based leading energy analytics company spun out of Carnegie Mellon University, is a scalable machine learning platform that converts raw smart meter data into appliance/equipment-level insights using proprietary algorithms. EEme’s proven Disaggregation-as-a-Service (DaaS) technology provides demand-side management stakeholders with appliance-level insights leveraging existing smart meter data and without relying on new hardware investments or user intervention. For additional information on EEme and its products contact Mr. Richard Huntley at email@example.com.
Located at The University of Texas at Austin, Pecan Street provides university researchers, utilities and technology companies with access to the world’s best original data on consumer energy and water consumption behavior, testing and verification of technology solutions, and commercialization services. Its network of over 1,200 research volunteers is the first of its kind on the planet. Its anonymized research database, the largest source of disaggregated customer energy data, is used by university researchers from over 175 universities in 38 nations, along with industry-leading companies around the world. Pecan Street is a 501(c)(3) applied research and commercialization institute. Learn more at pecanstreet.org. For additional information on Pecan Street contact Mr. Colin Rowan at firstname.lastname@example.org.