Concerns about the environmental footprint of machine learning are increasing. While studies of energy use and emissions of ML models are a growing subfield, most ML researchers and developers still do not incorporate energy measurement as part of their work practices. While measuring energy is a crucial step towards reducing carbon footprint, it is also not straightforward. This paper introduces the main considerations necessary for making sound use of energy measurement tools and interpreting energy estimates, including the use of at-the-wall versus on-device measurements, sampling strategies and best practices, common sources of error, and proxy measures. It also contains practical tips and real-world scenarios that illustrate how these considerations come into play. It concludes with a call to action for improving the state of the art of measurement methods and standards for facilitating robust comparisons between diverse hardware and software environments.

Electrification of vehicles is a potential way of reducing fossil fuel usage and thus lessening environmental pollution. Electric Vehicles (EVs) of various types for different transport modes (including air, water, and land) are evolving. Moreover, different EV user groups (commuters, commercial or domestic users, drivers) may use different charging infrastructures (public, private, home, and workplace) at various times. Therefore, usage patterns and energy demand are very stochastic. Characterizing and forecasting the charging demand of these diverse EV usage profiles is essential in preventing power outages. Previously developed data-driven load models are limited to specific use cases and locations. None of these models are simultaneously adaptive enough to transfer knowledge of day-ahead forecasting among EV charging sites of diverse locations, trained with limited data, and cost-effective. This article presents a location-based load forecasting of EV charging sites using a deep Multi-Quantile Temporal Convolutional Network (MQ-TCN) to overcome the limitations of earlier models. We conducted our experiments on data from four charging sites, namely Caltech, JPL, Office-1, and NREL, which have diverse EV user types like students, full-time and part-time employees, random visitors, etc. With a Prediction Interval Coverage Probability (PICP) score of 93.62\%, our proposed deep MQ-TCN model exhibited a remarkable 28.93\% improvement over the XGBoost model for a day-ahead load forecasting at the JPL charging site. By transferring knowledge with the inductive Transfer Learning (TL) approach, the MQ-TCN model achieved a 96.88\% PICP score for the load forecasting task at the NREL site using only two weeks of data.

Your smart phone can feel like a lifeline, helping you navigate a new town or delivering an urgent message to a friend. Many people have a funny or embarrassing anecdote about an autocorrected text message or a roundabout route to a destination. But these artificial intelligence (AI) flaws exist on a spectrum, from minor inconveniences to unfair treatment or even risk to human life. The people who create and use these AI technologies are also imperfect; we have our own biases, whether we are aware of them or not. Unconscious bias can influence our decisions and lead to unintended consequences; overt prejudice can result in our unethical and harmful exploitation of AI technologies.

Utilidata, an industry leading grid-edge software company, announced today that it is developing a software-defined smart grid chip in collaboration with NVIDIA. The chip will be powered by NVIDIA's AI platform and embedded in smart meters to enhance grid resiliency, integrate distributed energy resources (DERs) -- including solar, storage, and electric vehicles (EVs) -- and accelerate the transition to a decarbonized grid. The U.S. Department of Energy's (DOE's) National Renewable Energy Laboratory (NREL) will be among the first to test the software-defined smart grid chip as a way to scale and commercialize the lab's Real-Time Optimal Power Flow (RT-OPF) technology, with support from the Solar Energy Technologies Office Technology Commercialization Fund. Originally developed with funding from DOE's Advanced Research Projects – Energy (ARPA-E) program, RT-OPF enables highly localized load control to seamlessly integrate an increasing number of DERs while ensuring stable and efficient grid operations. "To date, the scalability and commercial potential of technologies like RT-OPF have been limited by single-use hardware solutions," said Santosh Veda, Group Manager for Grid Automation and Controls at NREL. "By developing a smart grid chip that can be embedded in one of the most ubiquitous utility assets – the smart meter – this approach will potential enable wider adoption and commercialization of the technology and redefine the role of edge computing for DER integration and resiliency. Enhanced situational awareness and visibility from this approach will greatly benefit both the end customers and the utility."

To discover materials for better batteries, researchers must wade through a vast field of candidates. New research demonstrates a machine learning technique that could more quickly surface ones with the most desirable properties. The study could accelerate designs for solid-state batteries, a promising next-generation technology that has the potential to store more energy than lithium-ion batteries without the flammability concerns. However, solid-state batteries encounter problems when materials within the cell interact with each other in ways that degrade performance. Researchers from the National Renewable Energy Laboratory (NREL), the Colorado School of Mines, and the University of Illinois demonstrated a machine learning method that can accurately predict the properties of inorganic compounds.

Researchers at the US Department of Energy's (DOE's) National Renewable Energy Laboratory (NREL) have developed a novel machine learning approach to quickly enhance the resolution of wind velocity data by 50 times and solar irradiance data by 25 times--an enhancement that has never been achieved before with climate data. The researchers took an alternative approach by using adversarial training, in which the model produces physically realistic details by observing entire fields at a time, providing high-resolution climate data at a much faster rate. This approach will enable scientists to complete renewable energy studies in future climate scenarios faster and with more accuracy. "To be able to enhance the spatial and temporal resolution of climate forecasts hugely impacts not only energy planning, but agriculture, transportation, and so much more," said Ryan King, a senior computational scientist at NREL who specializes in physics-informed deep learning. Recommended AI News: Interlink Electronics Welcomes Aboard Edward Suski As Chief Product Officer King and NREL colleagues Karen Stengel, Andrew Glaws, and Dylan Hettinger authored a new article detailing their approach, titled "Adversarial super-resolution of climatological wind and solar data," which appears in the journal Proceedings of the National Academy of Sciences of the United States …

HP Enterprise has partnered with the National Renewable Energy Laboratory (NREL), a unit of the Department of Energy, to create AI and machine learning-systems for greater data-center energy efficiency. The Department of Energy lab will provide HPE with multiple years' worth of historical data from sensors within its supercomputers and in its Energy Systems Integration Facility (ESIF) High-Performance Computing (HPC) Data Center, one of the world's most efficient data centers. This information will help other organizations to optimize their own operations, said NREL. The project, dubbed "AI Ops R&D collaboration," is expected to run over three years. Already NREL has 16 terabytes of data from the ESIF data center, collected from sensors in NREL's supercomputers, Peregrine and Eagle, and its facility.

A team at the US National Renewable Energy Laboratory (NREL) is working on autonomous energy grid (AEG) technology to ensure the electricity grid of the future can manage a growing base of intelligent energy devices, variable renewable energy, and advanced controls. "The future grid will be much more distributed too complex to control with today's techniques and technologies," said Benjamin Kroposki, director of NREL's Power Systems Engineering Center. "We need a path to get there--to reach the potential of all these new technologies integrating into the power system." The AEG effort envisions a self-driving power system - a very "aware" network of technologies and distributed controls that work together to efficiently match bi-directional energy supply to energy demand. This is a hard pivot from today's system, in which centralized control is used to manage one-way electricity flows to consumers along power lines that spoke out from central generators.

This paper is an appendix to the paper "Reasoning with Justifiable Exceptions in Contextual Hierarchies" by Bozzato, Serafini and Eiter, 2018 [2]. It provides further details on the language, the complexity results and the datalog translation introduced in the main paper.

Subnational policies pertaining to photovoltaic (PV) systems have increased in volume in recent years and federal incentives are set to be phased out over the next few. Understanding how subnational policies function within and across jurisdictions, thereby impacting PV market development, informs policy decision making. This report was developed for subnational policy-makers and researchers in order to aid the analysis on the function of PV system incentives within the emerging PV deployment market. The analysis presented is based on a ‘logic engine’, a database tool using existing state, utility, and local incentives allowing users to see the interrelationships between PV system incentives and parameters, such as geographic location, technology specifications, and financial factors. Depending on how it is queried, the database can yield insights into which combinations of incentives are available and most advantageous to the PV system owner or developer under particular circumstances. This is useful both for individual system developers to identify the most advantageous incentive packages that they qualify for as well as for researchers and policymakers to better understand the patch work of incentives nationwide as well as how they drive the market. In the case of the latter, findings from initial queries identify a limited connection between incentives and market development (based on current data) and point to differing complexities for system developers depending on system owner and size. The entire effort reveals (or possibly reiterates) a critical lack of data on both local policy environments and the structure of market penetration to be able to understand the impact of subnational incentives on the market.