Search NASA⌕ Search

SEARCH · Search NASA

Results for “Data reduction methods”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 217 records · Page 12

Capacity Design Method for Integrated Convective/Radiant Terminals to Guarantee Overall and Local Environment

To achieve energy efficiency in the operation of heating devices and ensure thermal comfort in indoor environments, the integrated convective/radiant terminals have become an important development direction for heating systems. This study selected a 14 m 2 bedroom hot summer and cold winter region in China as a case study. Computational Fluid Dynamics (CFD) methodology was used to investigate the design of integrated convective/radiant terminals to ensure thermal comfort for both the entire room and partial space. The results indicate that focusing on partial space during the steady-state stage can achieve energy savings of 31.1% compared to guaranteeing the entire room. Additionally, during the start-up stage, there is a significant reduction of convective unit capacity by 17.6% when the start-up time is 15 minutes. These findings provide data support for the design and engineering applications of products of integrated convective/radiant terminals. Besides, the excess convection ratios for the entire room and partial space are specified as 0.69 to 0.80 and 0.73 to 0.83, respectively.

Chi, Junjie↗

Two-phase flow numerical analysis of electrode geometry for alkaline water electrolyzers

Hydrogen is a promising component of a future energy-secure and efficient economy, but its competitiveness depends on reducing production costs. One strategy is to operate alkaline water electrolyzers at higher current densities to increase output. However, this intensifies performance losses due to gas bubble accumulation, which blocks transport pathways and deactivates electrochemically active surfaces. Enhancing bubble evacuation through electrode design is therefore essential. Previous studies have explored various approaches — such as modifying surface morphology, applying sonication or pressure modulation, and introducing surfactants — but these efforts have addressed a limited range of conditions due to the complexity of two-phase flow and electrode geometries. Experiments have also largely been focused on either cell level improvements, which lack the information necessary to isolate each contributing factor, or on modified geometries that are not relevant to practical cell operation. From a modeling perspective, conventional Eulerian multiphase models do not track the complex gas–liquid interfacial dynamics and often neglect surface tension and contact angle effects, reducing their predictive accuracy. To provide insights on the effects of different electrode geometries on the performance of alklaine water electrolyzers this work employs an immersed boundary volume-of-fluid method to simulate bubble behavior in 3D porous electrodes. Multiple base electrode geometries, typically used in practice, with varying porosity are evaluated under a constant surface gas generation rate. Simulation data is analyzed to quantify electrode gas coverage, bubble size dynamics and other relevant metrics. Results show that porosity strongly influences bubble accumulation on electrode surfaces, with higher porosity reducing gas coverage, and its not strictly dependent on the electrode geometry. However, the electrode’s base geometry significantly affects gas accumulation at the separator gap, independent of porosity. A foam electrode geometry resulted in the lowest gas coverage of all electrodes with a median volumetric gas coverage of 11%, but at the cost of a 70% reduction in active area compared with the largest surface area electrode, while gyroid electrodes showed the best trade-off between gas coverage, particularly at the separator surface, and electrochemically active area. In conclusion, the results highlight the need for holistic electrode design strategies.

Alkaline↗

Deciphering Reaction Mechanisms of Molecular Proton Reduction Catalysts with Cyclic Voltammetry: Kinetic vs Thermodynamic Control

The kinetics and thermodynamics of elementary reaction steps involved in the catalytic reduction of protons to hydrogen define the reaction landscape for catalysis. The mechanisms can differ in the order of the elementary proton transfer, electron transfer, and bond-forming steps and can be further differentiated by the sites at which protons and electrons localize. Access to fully elucidated mechanistic, kinetic, and thermochemical details of molecular catalysts is crucial to facilitate the development of new catalysts that operate with optimal efficiency, selectivity, and durability. The mechanism by which a catalyst operates, as well as the kinetics and thermodynamics associated with the individual steps, can often be accessed through electroanalytical studies. Here, this Account details the application of cyclic voltammetry to interrogate reaction mechanisms and quantify the kinetics and thermodynamics of elementary reaction steps for a series of molecular catalysts that mediate electrochemical proton reduction. I distinguish the limiting scenarios wherein a catalyst operates under kinetic control vs thermodynamic control, with a focus on detecting how cyclic voltammetry features shift with proton source strength and concentration, as well as scan rate. For systems that operate under kinetic control, catalytic currents are observed at, or slightly positive toward, the formal potential for the redox process that triggers catalysis. Under thermodynamic control, catalytic responses shift as a function of the proton source pKa and effective pH of the solution. After drawing this distinction, we introduce the appropriate voltammetry experiments and accompanying analytical expressions for extracting key metrics from the data. To illustrate analytical strategies to quantify elementary reaction steps of catalysts operating under kinetic control, I describe our studies of proton reduction catalysts Co(dmgBF 2 ) 2 (CH 3 CN) 2 (dmgBF 2 = difluoroboryl-dimethylglyoxime) and [Ni(P 2 Ph N 2 Ph ) 2 ] 2+ (P 2 Ph N 2 Ph = 1,5-phenyl-3,7-phenyl-1,5-diaza-3,7-diphosphacyclooctane). Here, peak shift analysis, foot-of-the-wave analysis, and plateau current analysis are applied to data sets wherein voltammetric response are recorded as a function of catalyst concentration, proton source concentration, proton source strength, and scan rate to quantify rate constants for elementary proton transfer and bond-forming steps in a catalytic cycle. Further, the case study of [Ni(P 2 Ph N 2 Ph ) 2 ] 2+ illustrates how complementary spectroscopic methods can bolster the mechanistic assignment. Collectively, these two studies showcase how detailed mechanistic studies inform on rate-limiting elementary steps in catalysis and other key processes underpinning catalysis. Second, I present analytical strategies to interrogate catalysts operating under thermodynamic control, centered on the case study of [Ni II (P 2 Ph N 2 Bn ) 2 ] 2+ (P 2 Ph N 2 Bn = 1,5-dibenzyl-3,7-diphenyl-1,5-diaza-3,7-diphosphacyclooctane). Here, the application of nonaqueous Pourbaix theory to extract thermodynamic information is introduced, and the construction of a coupled Pourbaix diagram is detailed. This study identifies ligand-based protonation as the key process that places catalysis under thermodynamic control and influences the reaction mechanism. Together, the work detailed in this Account showcases the utility of electroanalytical methods to disentangle complex reaction mechanisms and extract key thermochemical and kinetic parameters for elementary steps of catalysis. Through detailed presentation of the key analytical expressions that underpin these analyses, this Account seeks to facilitate the adoption of cyclic voltammetry by the community to fully extract kinetic, thermochemical, and mechanistic information on electrochemical small-molecule activation.

catalysts↗

Optimization of foreground moment deprojection for semi-blind CMB polarization reconstruction

Abstract Upcoming Cosmic Microwave Background (CMB) experiments, aimed at measuring primordial CMB polarization B-modes, require exquisite control of instrumental systematics and Galactic foreground contamination. Blind minimum-variance techniques, like the Needlet Internal Linear Combination (NILC), have proven effective in reconstructing the CMB polarization signal and mitigating foregrounds and systematics across diverse sky models without suffering from foreground mismodelling errors. Still, residual foreground contamination from NILC may bias the recovered CMB polarization at large angular scales when confronted with the most complex foreground scenarios.By adding constraints to NILC to deproject statistical moments of the Galactic emission, the Constrained Moment ILC (cMILC) method has been demonstrated to further enhance foreground subtraction, albeit with an associated increase in overall noise variance. Faced with this trade-off between foreground bias reduction and overall variance minimization, there is still no recipe on which moments to deproject and which are better suited for blind variance minimization. To address this, we introduce the optimized cMILC (ocMILC) pipeline, which performs full automated optimization of the required number and set of foreground moments to deproject, pivot parameter values, and deprojection coefficients across the sky and angular scales, depending on the actual sky complexity, available frequency coverage, and experiment sensitivity. The optimal number of moments for deprojection, before paying significant noise penalty, is determined through a data diagnosis inspired by the Generalized NILC (GNILC) method.Validated on B-mode simulations of thePICOspace mission concept with four challenging foreground models, ocMILC exhibits lower Galactic foreground contamination compared to NILC and cMILC at all angular scales, with limited noise penalty. This multi-layer optimization enables the ocMILC pipeline to achieve unbiased posteriors of the tensor-to-scalar ratio, regardless of foreground complexity.

Astronomy & Astrophysics↗

Optimal environmental and economic performance trade-offs for fifth generation district heating and cooling network topologies with waste heat recovery

Network topology greatly influences both the economic and environmental performance of fifth generation district heating and cooling (5GDHC) systems. In this study the optimal trade-offs between the environmental and economic performance of 5GDHC network topologies for a five-building district with waste heat recovery were explored. A life cycle assessment method was used to calculate the total life cycle CO 2 emissions (LCCO2) associated with the installation and operation of various network topologies. Twelve months of empirical data from a data center cooling system were analyzed to assess its suitability for integration into a 5GDHC system. The most suitable method for utilizing this waste heat was selected based on the ambient loop warm pipe setpoint, waste heat temperature, and district energy system configuration. A multi-objective optimization algorithm was used to select the 5GDHC network topology that provided the optimal trade-off between LCCO2 and life cycle cost (LCC). A trade-off parameter was employed to weigh the importance of each objective in the selection process. The results showed waste heat from the data center was suitable for integration into the 5GDHC system due to its availability and consistent temperature profile. When return temperatures of 25 °C or higher were available from the liquid-cooled system, direct pre-heating of the ambient loop warm pipe was found to be the most effective waste heat integration method. The selection of the network topology that provided the optimal trade-off between LCCO2 and LCC (optimal trade-off topology) was highly dependent on factors such as fuel prices, CO 2 prices, electricity CO 2 emissions factors, availability of waste heat, embodied CO 2 emissions associated with network installation and network infrastructure costs. Optimal trade-off topologies produced substantial LCCO2 reductions relative to corresponding LCC increases. LCCO2 reduction to LCC increase ratios from 5.78 to 117.79 were identified with CO 2 offset costs ranging from 4.77 to 60.08 ($/tCO 2 e).

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Data Science-Driven Discovery of Multimetallic Oxygen-cycle Electrocatalysts for Enhanced Energy Conversion

The overarching objective of this effort has been to combine state-of-the-art data science techniques, first principles analyses, and molecular-level characterization of electrocatalyst structure and reactivity to identify both in-situ mechanisms for degradation and transformation of electrocatalysts with highly complex catalytic structures and the impact of these transformations on catalytic activity. The primary catalysts of interest have been multielemental alloys, including high entropy alloys (HEA’s), which are characterized by a high degree of disorder and up to 20 different elements within a single nanoparticle. We have applied these strategies primarily to energy-critical oxygen cycle electrocatalytic reactions, including oxygen reduction (ORR), but we have also considered extensions to non-electrochemical chemistries such as ammonia synthesis and decomposition. We have made strong progress in the development of computational methods on both the level of machine learning methods development as well as first principles-based treatments of HEA’s, and we have leveraged these insights to propose promising HEA catalysts for the ORR. On the experimental side, we developed new HEA synthesis and characterization protocols relevant to these reactions and developed a database combining our experimental results with corresponding computational tools.

36 MATERIALS SCIENCE↗

Bias correcting regional scale Earth system model projections: novel approach using empirical mode decomposition

Bias correction is a crucial step in using Earth system model outputs for assessments, as it adjusts systematic errors by comparing the model to observations. However, standard methods – ranging from mean-based linear scaling to distribution-based quantile mapping typically treat bias correction as a single-scale process, overlooking the fact that biases can manifest differently across daily, seasonal, and annual timescales. In this study, we propose a novel, timescale-aware bias-correction approach built on Empirical Mode Decomposition. By decomposing the meteorological signal into multiple oscillatory components and aggregating them to represent distinct timescales, we apply targeted corrections to each component, thereby preserving both short- and long-term structure in the data. Experimental illustrations show that the timescale-aware EMDBC framework matches the performance of conventional quantile-delta mapping (QDM) at the native daily scale and achieves progressively larger bias reductions at bi-weekly, seasonal, and annual scales. As a result, the proposed approach offers a more robust path to accurate and reliable Earth system projections, strengthening their utility for resilience and adaptation planning.

Ganguli, Arkaprabha [Argonne National Laboratory (↗

FAIR Data and Interpretable AI Framework for Architectured Metamaterials (Final Report)

This research program established a transformative framework for the discovery and design of mechanical metamaterials, which are architected structures engineered to control physical phenomena like sound and vibration in ways natural materials cannot. To overcome the traditional reliance on trial-and-error, the project developed an interpretable Artificial Intelligence (AI) framework that moves beyond "black box" models to reveal the specific geometric patterns—such as "unit-cell templates"—that govern a material’s performance. A major breakthrough was the development of a hierarchical design method, which allows a single material to block vibrations across multiple frequency ranges simultaneously by layering patterns at different scales without them interfering with one another. This was further expanded to include irregular, graph-based designs that use spanning tree algorithms to ensure structural connectivity while allowing for customized, direction-dependent properties like stiffness and acoustic impedance. Beyond design, the project addressed the practicalities of real-world production by developing uncertainty quantification techniques that account for manufacturing defects and material variability, reducing the need for expensive physical testing by orders of magnitude. To speed up the discovery process, the team implemented Gaussian Process Regression and other surrogate models that provide accurate performance predictions at a fraction of the traditional computational cost. The AI-generated designs were successfully validated through fabrication of physical samples and wave propagation experiments, confirming their ability to accurately guide or reflect waves as predicted. By contributing these tools and high-quality FAIR benchmark datasets to the wider scientific community, this work provides a scalable foundation for advancing technologies in aerospace vibration control, medical imaging, and noise reduction.

36 MATERIALS SCIENCE↗

The detection of marine microseismic activity with the CUORE tonne-scale cryogenic experiment

Vibrations from experimental setups and the environment are a persistent source of noise for low-temperature calorimeters searching for rare events, including neutrinoless double beta ( 0νββ ) decay or dark matter interactions. Such noise can significantly limit experimental sensitivity to the physics case under investigation. Here, we report the detection of marine microseismic vibrations using mK-scale calorimeters. This study employs a multi-device analysis correlating data from CUORE, the leading experiment in the search for 0νββ decay with mK-scale calorimeters, and the Copernicus Earth Observation program, revealing the seasonal impact of Mediterranean Sea activity on CUORE’s energy thresholds, resolution, and sensitivity over four years. The detection of marine microseisms underscores the need to address faint environmental noise in ultra-sensitive experiments. Understanding how such noise couples to the detector and developing mitigation strategies is essential for next-generation experiments. We demonstrate one such strategy: a noise decorrelation algorithm implemented in CUORE using auxiliary sensors, which reduces vibrational noise and improves detector performance. Enhancing sensitivity to 0νββ decay and to rare events with low-energy signatures requires identifying unresolved noise sources, advancing noise reduction methods, and improving vibration suppression systems, all of which inform the design of next-generation rare event experiments.

experimental nuclear physics↗

Hybrid data-driven cement-stabilized soil design: An integration of machine learning, multi-objective optimization, and life cycle assessment

Soil stabilization is crucial in geotechnical engineering, yet conventional methods are often time-consuming, resource-intensive, and environmentally unsustainable. Despite growing interest in Machine Learning (ML) and optimization tools for mix design, few studies integrate these methods with decision-making techniques and environmental assessment to support practical implementation. This study proposes a hybrid data-driven framework for predicting strength, optimizing mix compositions, and evaluating environmental impacts via life cycle assessment of cement-stabilized soft soils. Six ML models were evaluated, and the top-performing eXtreme Gradient Boosting (XGB) model was further improved using the Grey Wolf Optimizer (GWO). The optimized XGB-GWO model, integrated with a polynomial cost function, served as the objective function in a multi-objective optimization problem solved via the Non-Dominated Sorting Genetic Algorithm II (NSGA-II), with final mix selection guided by the entropy-weighted TOPSIS method. Validation through a case study produced mix designs offering superior strength-cost trade-offs, with the optimal mix achieving 2243.2 kPa unconfined compressive strength and a 16.07 % reduction in carbon emissions compared to the highest-cost design. In conclusion, this study offers a sustainable, scalable approach to soil stabilization and supports informed decision-making in construction.

Life cycle assessment↗

Identifying Vehicle Signals in Continuous Seismic Data Using Unsupervised Machine-Learning Techniques

Seismic sensors deployed near roadways effectively capture ground vibrations generated by passing vehicles. Although both traditional and machine‐learning algorithms have been utilized for analyzing such signals, independent validation of detected vehicle events remains limited. We applied two unsupervised machine‐learning algorithms, uniform manifold approximation and projection for dimension reduction, and hierarchical density‐based spatial clustering of applications with noise, to continuous seismic data collected along a road on the main campus of Oak Ridge National Laboratory. The algorithms identified seven distinct cluster labels across the entire dataset. By comparing these cluster labels with precipitation records from a nearby weather station and image‐derived labels from a local camera system, we identified one cluster associated with rainfall and another with vehicle activity. Our algorithms identified a greater number of vehicle‐related labels compared to the camera‐derived labels because seismic data are unaffected by poor lighting conditions. The arrival times of the newly detected vehicle signals corresponded well with the road’s speed limit, supporting our findings. Our algorithm outperformed the short‐term average/long‐term average method and k‐means clustering. Our results suggest that seismic data, when analyzed with machine‐learning algorithms, can complement existing vehicle monitoring systems, particularly under challenging environmental conditions.

Chai, Chengping [Oak Ridge National Laboratory (OR↗

Robust Heat-Flux Sensors for Coal-Fired Boiler Extreme Environments

In this project, robust heat-flux measurement systems were developed. The heat-flux sensors utilize thermoelectric effects to directly transduce the heat-flux inputs to analog electrical voltage signals. They were constructed from dedicated materials that can withstand temperatures of at least 1000°C and maintain adequate performance at these conditions for prolonged periods of time. The proposed approaches took into account numerous considerations, including system cost, sensor head resilience, sensor footprint, data accuracy, response time, and maintenance requirements. Through modern thermoelectric materials design, methodical materials selection and rigorous testing in materials characterization labs and medium-scale fire research facilities, we have demonstrated functioning laboratory prototypes, upon which one could base industrial heat-flux sensing platforms capable of operating in the challenging high-temperature, corrosive environments of the boilers of coal-fired power plants. A distributed sensor array for heat-flux measurements throughout the furnace water-wall, the superheater area and the economizer coils can provide critical data for the power plant control systems to increase efficiency, improve safety and reduce down times. For example, the combined heat-flux sensor/control systems can contribute to the optimization of burner and boiler operations under flexible loads, the optimization of heat-exchange conditions and overall reduction of heat rate and emissions, the prediction of imminent overheating conditions, and the optimization of the soot-blowing protocols.

20 FOSSIL-FUELED POWER PLANTS↗

Scaling kinetic Monte-Carlo simulations of grain growth with combined convolutional and graph neural networks

Graph neural networks (GNN) have emerged as a promising machine learning method for microstructure simulations such as grain growth. However, accurate modeling of realistic grain boundary networks requires large simulation cells, which GNN has difficulty scaling up to. To alleviate the computational costs and memory footprint of GNN, we suggest a hybrid architecture combining a convolutional neural network (CNN) based bijective autoencoder to compress the spatial dimensions, and a GNN that evolves the microstructure in the latent space of reduced spatial sizes. Our results demonstrate that the new design significantly reduces computational costs with using fewer message passing layer (from 12 down to 3) compared with GNN alone. The reduction in computational cost becomes more pronounced as the spatial size increases, indicating strong computational scalability. For the largest mesh evaluated (160 3 ), our method reduces memory usage and runtime in inference by 117× and 115×, respectively, compared with GNN-only baseline. More importantly, it shows higher accuracy and stronger spatiotemporal capability than the GNN-only baseline, especially in long-term testing. Such combination of scalability and accuracy is essential for simulating realistic material microstructures over extended time scales. The improvements can be attributed to the bijective autoencoder’s ability to compress information losslessly from spatial domain into a high dimensional feature space, thereby producing more expressive latent features for the GNN to learn from, while also contributing its own spatiotemporal modeling capability. Training data are generated from stochastic grain growth simulations, providing realistic variability for learning robust microstructure evolution. Comprehensive system validation confirms that the model is accurate, robust, and scalable.

36 MATERIALS SCIENCE↗

Probing the influence of ion-pairing on ligand-field excited-state dynamics

Exploration of the photophysical and photochemical properties of transition metal complexes has driven ground-breaking advancements in solar energy conversion technologies, including photoredox catalysis. While significant research has been devoted to understanding excited state properties of second- and third-row transition metal complexes, earth-abundant first-row metal complexes have received comparatively little attention in this context until very recently. In particular, the role of ion-pairing – which has been identified as a potentially significant factor for Ir(III)-based photosensitizers – has not been examined with regard to its influence on the ligand-field excited states that dominate much of first-row photophysics. A key challenge in studying ion-pair interactions lies in quantifying the extent and nature of ion-pairing, particularly in non-aqueous media where the vast majority of photophysical studies are performed. Cobalt(III) polypyridyl complexes provide an attractive platform to address such questions due to their demonstrated potential for applications in photoredox catalysis involving ligand-field excited states. In the present study, we prepared a cobalt(III) polypyridyl complex, [Co(4,4′-OMebpy) 3 ](BAr F 4 ) 3 (where 4,4′-OMebpy is 4,4′-dimethoxy-2,2′-bipyridine and BAr F 4 is tetrakis(3,5-bis(trifluoromethyl)-phenyl)borate) to probe ion-pairing in non-aqueous solutions. Specifically, analysis of data acquired from both variable-temperature diffusion ordered spectroscopy (DOSY) NMR and 1-D rotating-frame nuclear Overhauser effect (ROE) experiments allowed us to identify and differentiate between solvent-separated ion pairs in high-dielectric media and contact ion pairs in a low-dielectric solvent. Time-resolved absorption spectroscopy was then used to measure ground-state recovery dynamics under these varying conditions of ion-pairing, the results of which revealed an increase in excited-state lifetime for contact ion-pairs that we suggest arises from a reduction in outer-sphere reorganization energy relative to conditions which favored solvent-separated ion pairs. We believe this study demonstrates that one can leverage broadly available NMR-based methods to understand ion-pairing in non-aqueous solutions, which in turn can provide a microscopic picture of intermolecular interactions that can impact the photophysical properties of transition metal-based chromophores.

Ghosh, Atanu [Michigan State University, East Lans↗

CO 2 storage site characterization using ensemble-based approaches with deep generative models

Estimating spatially distributed properties such as permeability from available sparse measurements is a great challenge in efficient subsurface CO 2 storage operations. In this paper, a deep generative model that can accurately capture complex subsurface structure is tested with an ensemble-based inversion method for accurate and accelerated characterization of CO 2 storage sites. We chose Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP) for its realistic reservoir property representation and Ensemble Smoother with Multiple Data Assimilation (ES-MDA) for its robust data fitting and uncertainty quantification capability. WGAN-GP are trained to generate high-dimensional permeability fields from a low-dimensional latent space and ES-MDA then updates the latent variables by assimilating available measurements. Several subsurface site characterization examples including Gaussian, channelized, and fractured reservoirs are used to evaluate the accuracy and computational efficiency of the proposed method and the main features of the unknown permeability fields are characterized accurately with reliable uncertainty quantification. Furthermore, the estimation performance is compared with a widely-used variational, i.e., optimization-based, inversion approach, and the proposed approach outperforms the variational inversion method in several benchmark cases. We explain such superior performance by visualizing the objective function in the latent space: because of nonlinear and aggressive dimension reduction via generative modeling, the objective function surface becomes extremely complex while the ensemble approximation can smooth out the multi-modal surface during the minimization. This suggests that the ensemble-based approach works well over the variational approach when combined with deep generative models at the cost of forward model runs unless convergence-ensuring modifications are implemented in the variational inversion.

42 ENGINEERING↗

Cognitive IoT and Edge Computing for Intrusion Detection with Federated TinyML

Internet of Things (IoT) and Edge Computing (EC) are rapidly becoming an integral part of the modern society. By 2030, there is estimated to be over 40 billion active and connected IoT devices [1]. This rapid progress also comes with a significant implication on cybersecurity. Back-end infrastructure and systems have a much broader attack than they did previously due to vulnerable IoT/EC devices being connected to wireless networks. This expanding attack surface is a growing concern because IoT/EC are increasingly being used in critical systems such as power grids, health care, and smart homes. To effectively address a problem of this scale, cognitive cyber methods—which can autonomously detect and react to cyber attacks as they develop—are needed. To address this, we bring Artificial Intelligence (AI) and Machine Learning (ML) to IoT/EC devices, using tinyML to monitor voluminous IoT data against cyber threats, and using Federated Learning (FL) to share local detection knowledge across the system while preserving privacy. We propose a novel three-layer architecture: (1) an IoT layer for tinyML-based inference, (2) an edge layer for ML model training, and (3) a cloud layer for FL operations. Using the publicly available 11-class N-BaIoT dataset [2], we demonstrate that this architecture mitigates resource constraints at the IoT layer while improving detection accuracy over standard two-layer designs. An outlier-resistant scaler, feature reduction, and quantization enable the tinyML model to maintain detection accuracy with a reduced model size. Additionally, federated learning that only utilizes the intersection (across heterogenous devices) of the reduced feature set achieves superior detection accuracy compared to locally trained models.

Li, Mingyan [ORNL] (ORCID:0009000569532640)↗

GreenThrift: Optimizing Carbon and Cost for Flexible Residential Loads

Reducing buildings’ carbon emissions is an important sustainability challenge. While scheduling flexible building loads has been previously used for a variety of grid and energy optimizations, carbon footprint reduction using such flexible loads poses new challenges since such methods need to balance both energy and carbon costs while also reducing user inconvenience from delaying such loads. This paper highlights the potential conflict between electricity prices and carbon emissions and the resulting trade-offs in carbon-aware and cost-aware load scheduling. To address this trade-off, we propose GreenThrift, a home automation system that leverages the scheduling capabilities of smart appliances and knowledge of future carbon intensity and cost to reduce both the carbon emissions and costs of flexible energy loads. At the heart of GreenThrift is an optimization technique that automatically computes schedules based on user configurations and preferences. We evaluate the effectiveness of GreenThrift using real-world carbon intensity data, electricity prices, and load traces from multiple locations and across different scenarios and objectives. Our results show that GreenThrift can replicate the offline optimal and retains 97% of the savings when optimizing the carbon emissions. Moreover, we show how GreenThrift can balance the conflict between carbon and cost and retain 95.3% and 85.5% of the potential carbon and cost savings, respectively.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗