Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “learning based systems”

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 1,765 records · Page 98
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

An interactive machine learning platform for analyzing multi-particle coincidence data from cold target recoil ion momentum spectroscopy

We present SCULPT (Supervised Clustering and Uncovering Latent Patterns with Training), a comprehensive software platform for analyzing tabulated high-dimensional multi-particle coincidence data from Cold Target Recoil Ion Momentum Spectroscopy (COLTRIMS) experiments. The software addresses critical challenges in modern momentum spectroscopy by integrating advanced machine learning techniques with physics-informed analysis in an interactive web-based environment. SCULPT implements uniform manifold approximation and projection for non-linear dimensionality reduction to reveal correlations in high-dimensional data. We also discuss potential extensions to deep autoencoders for feature learning and genetic programming for automated discovery of physically meaningful observables. A novel adaptive confidence scoring system provides quantitative reliability assessments by evaluating user-selected clustering quality metrics with predefined weights that reflect each metric’s robustness. The platform features configurable molecular profiles for different experimental systems, interactive visualization with selection tools, and comprehensive data filtering capabilities. Utilizing a subset of SCULPT’s capabilities, we analyze photo-double-ionization data measured using the COLTRIMS method for three-body dissociation of the D 2 O molecule, revealing distinct fragmentation channels and their correlations with physics parameters. The software’s modular architecture and web-based implementation make it accessible to the broader atomic and molecular physics community, significantly reducing the time required for complex multi-dimensional analyses. This opens the door to finding and isolating rare events exhibiting non-linear correlations on the fly during experimental measurements, which can help steer exploration and improve the efficiency of experiments.

Artificial neural networks↗
DOE OSTI2025

Energy storage planning for enhanced resilience of power systems against wildfires and heatwaves

Extreme weather events pose significant risks to power grid stability due to their severe consequences and potential for widespread failures. Energy storage systems hold great potential for enhancing grid resilience against such events by providing reliable power during peak demand periods. However, accurately quantifying the size, location, and investment costs of new energy storage assets is a complex task, as energy storage planning decisions depend on the investment choices of other generation technologies and the integration of new transmission projects. Here, this paper presents a novel capacity expansion planning framework that simultaneously optimizes investments in energy storage, generation, and transmission, determining their optimal size, location, and type, while incorporating extreme weather events into long-term planning. More specifically, our stress-event-informed planning framework integrates the impact of heatwaves and wildfires into the planning process, identifying least-cost investment solutions that comply with policy goals and enhance grid resilience. The proposed framework employs machine-learning-based modeling to project heatwave-induced loads and performance-based risk assessment to evaluate wildfire-driven transmission line derates. Using industry-standard datasets to accurately represent the transmission topology of the Western Interconnection (WI) system, the proposed framework is applied to the WI 40-zone system, with investment decisions reported for the years 2030, 2035, and 2040. Simulation results reveal that with just a 10% increase in investment costs, resilience against extreme events can be significantly improved, with investment decisions heavily favoring energy storage, particularly 4-hour energy storage systems.

25 ENERGY STORAGE↗
NASA NTRSDate not supplied

Using Machine Learning to Infer Material Properties of Debris Fragments from X-ray Images in the DebriSat Project

The DebriSat project is a collaboration effort with the NASA Orbital Debris Program Office, the U.S. Space Force Space Systems Command Center, The Aerospace Corporation, and the University of Florida. To date, over 200,000 fragments from this ground-based, hypervelocity impact experiment have been collected, and processing is underway to determine their physical characteristics, such as material, shape, color, characteristic length, and average cross-sectional area. The x-ray process is primarily used to identify the location of the fragments and estimated size for extraction, so that these physical characteristics can be assessed. This paper proposes a machine learning-based approach to characterize materials from x-ray images of debris fragments embedded in soft-catch foam used in the DebriSat project. The novel methodology discussed in this paper will highlight the use of x-ray imagery data to characterize these fragments without extraction or a human-in-the-loop. Both supervised and unsupervised machine learning techniques are utilized with this approach to infer the physical parameters of the fragments embedded in the soft-catch foam panels used in the impact experiment based on x-ray images of the foam panels. Additionally, 3D reconstructions of the extracted fragments are created with images taken from two different angles using the structure from motion (SfM) method. The characteristic lengths and shape from the 3D reconstruction, alongside the physical characteristics of the debris, are used in the inference of the material type. To develop and test the approach, a dataset of x-ray images of debris fragments of varying sizes and materials is collected. Supervised learning methods such as convolutional neural networks (CNNs), support vector machines (SVM), decision trees, and random forest classifiers are used due to the high-dimensional feature spaces of the debris and nonlinear decision boundaries for material categorization. Given the limited pre-labeled data of embedded debris materials smaller than 10 mm, unsupervised machine learning techniques such as clustering algorithms and autoencoders are used, in addition to supervised learning methods. The clustering algorithms group similar fragments together based on their physical properties, and autoencoders reduce the dimensionality of the x ray images and extract relevant features. The performance of the proposed approach's is analyzed using a range of statistical methods, including confusion matrices, receiver operating characteristic curves, and precision-recall curves. The results are compared with those obtained using a baseline approach that relies on manual identification and classification of debris fragments. To evaluate the effectiveness of different machine learning methods, statistical tests such as t-tests, ANOVA, and cross-validation are performed, comparing the performance of CNNs, SVMs, clustering algorithms, and autoencoders. Additional analysis needs to be conducted to identify any sources of bias or variability that may affect the results, such as variations in imaging conditions or fragmentation patterns. Other topics explored are limitations, refinements, and the potential use of semi-supervised learning techniques, such as self-training to label unlabeled datasets and co-training using x-ray images taken from two different angles as two different models.

Saik Anam Siam↗
NASA NTRS2023

Using Machine Learning to Infer Material Properties of Debris Fragments from X-ray Images in the DebriSat Project

The DebriSat project is a collaboration effort with the NASA Orbital Debris Program Office, the U.S. Space Force Space Systems Command Center, The Aerospace Corporation, and the University of Florida. To date, over 200,000 fragments from this ground-based, hypervelocity impact experiment have been collected, and processing is underway to determine their physical characteristics, such as material, shape, color, characteristic length, and average cross-sectional area. The x-ray process is primarily used to identify the location of the fragments and estimated size for extraction, so that these physical characteristics can be assessed. This paper proposes a machine learning-based approach to characterize materials from x-ray images of debris fragments embedded in soft-catch foam used in the DebriSat project. The novel methodology discussed in this paper will highlight the use of x-ray imagery data to characterize these fragments without extraction or a human-in-the-loop. Both supervised and unsupervised machine learning techniques are utilized with this approach to infer the physical parameters of the fragments embedded in the soft-catch foam panels used in the impact experiment based on x-ray images of the foam panels. Additionally, 3D reconstructions of the extracted fragments are created with images taken from two different angles using the structure from motion (SfM) method. The characteristic lengths and shape from the 3D reconstruction, alongside the physical characteristics of the debris, are used in the inference of the material type. To develop and test the approach, a dataset of x-ray images of debris fragments of varying sizes and materials is collected. Supervised learning methods such as convolutional neural networks (CNNs), support vector machines (SVM), decision trees, and random forest classifiers are used due to the high-dimensional feature spaces of the debris and nonlinear decision boundaries for material categorization. Given the limited pre-labeled data of embedded debris materials smaller than 10 mm, unsupervised machine learning techniques such as clustering algorithms and autoencoders are used, in addition to supervised learning methods. The clustering algorithms group similar fragments together based on their physical properties, and autoencoders reduce the dimensionality of the x ray images and extract relevant features. The performance of the proposed approach's is analyzed using a range of statistical methods, including confusion matrices, receiver operating characteristic curves, and precision-recall curves. The results are compared with those obtained using a baseline approach that relies on manual identification and classification of debris fragments. To evaluate the effectiveness of different machine learning methods, statistical tests such as t-tests, ANOVA, and cross-validation are performed, comparing the performance of CNNs, SVMs, clustering algorithms, and autoencoders. Additional analysis needs to be conducted to identify any sources of bias or variability that may affect the results, such as variations in imaging conditions or fragmentation patterns. Other topics explored are limitations, refinements, and the potential use of semi-supervised learning techniques, such as self-training to label unlabeled datasets and co-training using x-ray images taken from two different angles as two different models.

Saik Anam Siam↗
NASA NTRS2005

AC/DC Power Systems with Applications for future Lunar/Mars base and Crew Exploration Vehicle

ABSTRACT The Power Systems branch at JSC faces a number of complex issues as it readies itself for the President's initiative on future space exploration beyond low earth orbit. Some of these preliminary issues - those dealing with electric power generation and distribution on board Mars-bound vehicle and that on Lunar and Martian surface may be summarized as follows: Type of prime mover - Because solar power may not be readily available on parts of the Lunar/Mars surface and also during the long duration flight to Mars, the primary source of power will most likely be nuclear power (Uranium fuel rods) with a secondary source of fuel cell (Hydrogen supply). The electric power generation source - With nuclear power being the main prime mover, the electric power generation source will most likely be an ac generator at a yet to be determined frequency. Thus, a critical issue is whether the generator should generate at constant or variable frequency. This will decide what type of generator to use - whether it is a synchronous machine, an asynchronous induction machine or a switched reluctance machine. The type of power distribution system - the distribution frequency, number of wires (3- wire, 4-wire or higher), and ac/dc hybridization. Building redundancy and fault tolerance in the generation and distribution sub-systems so that the system is safe; provides 100% availability to critical loads; continues to operate even with faulted sub-systems; and requires minimal maintenance. This report descril_es results of a summer faculty fellowship spent in the Power Systems Branch with the specific aim of investigating some of the lessons learned in electric power generation and usage from the terrestrial power systems industry, the aerospace industry as well as NASA's on-going missions so as to recommend novel surface and vehicle-based power systems architectures in support of future space exploration initiatives. A hybrid ac/dc architecture with source side and load side redundancies and including emergency generators on both ac and dc sides is proposed. The generation frequency is 400 Hz mostly because of the technology maturity at this frequency in the aerospace industry. Power will be distributed to several ac load distribution buses through solid state variable speed, constant frequency converters on the ac side. A segmented dc ring bus supplied from ac/dc converters and with the capability of connecting/disconnecting the segments will supply power to multiple de load distribution buses. The system will have the capability of reverse flow from dc to ac side in the case of an extreme emergency on the main ac generation side.

Chowdhury, Badrul H.↗
NASA NTRS2013

Scheduling Spitzer: The SIRPASS Story

NASA's Spitzer Space Telescope was launched on August 25, 2003 from Florida's Cape Canaveral Air Force Base. Drifting in a unique Earth-trailing orbit around the Sun, Spitzer sees an optically invisible universe dominated by dust and stars. Since 1997, the Spitzer Integrated Resource Planning and Scheduling System (SIRPASS) has helped produce spacecraft activity plans for the Spitzer Space Telescope. SIRPASS is used by members of the Observatory Planning and Scheduling Team to plan, schedule and sequence the Telescope from data made available to them from the science and engineering community. Because of the volume of data that needs to be scheduled, SIRPASS offers a variety of automated assistants to aid in this task. This paper will describe the functional elements of the SIRPASS software system -- emphasizing the role that automation plays in the system -- and will highlight lessons learned for the software developer from a decade of Spitzer Space Telescope operations experience.

flight systems↗
DOE OSTI2020

Novel Fault Location Method for Power Systems Based on Attention Mechanism and Double Structure GRU Neural Network

Fault location is one of the most essential techniques to maintain the stable operation of power systems. A fast and accurate fault location allows operators to restore power grids faster and avoid economic losses. Conventional methods rely on expert knowledge to extract the necessary features (e.g. DWT, DFT). For large systems, more coupling effects of transmission lines require more complex feature engineering, and incomplete features can easily introduce large errors. To overcome this, a deep learning approach without manual feature extraction is introduced to the fault location model under big data application. Towards this end, in the proposed method, the attention mechanism, the Bi-GRU and a dual structure network are applied to analyze the current data from different perspectives. Complete information for the fault features is extracted for the fault location. Based on a time series model and benefit from the ability to internally acquire the information architecture of faulty line, the established model is adaptive to the power grids with very complex topologies. Simulation results indicate that the proposed double-structure model reduces the maximum error and is less affected by noise. In comparison with different structures and different models, the proposed method shows better performance in IEEE 39-bus system.

42 ENGINEERING↗
NASA NTRSDate not supplied

A Framework for the Analysis of Deep Neural Networks in Autonomous Aerospace Applications using Bayesian Statistics

Deep Neural Networks (DNNs) are considered to be key components in many autonomous systems. Applications range from vision-based obstacle avoidance to intelligent/learning control and planning. Safety-critical applications as found in the aerospace domain require that the behavior of the DNN is validated and tested rigorously for safety of the autonomous system (AUS). In this paper, we present a framework to support testing of DNNs and the analysis of the network structure. Our framework employs techniques from statistical modeling and active learning to effectively generate test cases for DNN safety testing and performance analysis. We will present results of a case study on a physics-based Deep recurrent residual neural network (DR-RNN), which has been trained to emulate the aerodynamics behavior of a fixed-wing aircraft.

Deep Neural networks↗
DOE OSTI2017

Considerations for Corridor and Community DC Fast Charging Complex System Design

The Idaho National Laboratory (INL), a U.S. Department of Energy (DOE) national laboratory located in Idaho Falls, Idaho, has been DOE’s lead laboratory for the field testing of light-duty plug-in electric vehicles (PEVs), charging infrastructure, and other alternative vehicle technologies for 25 years. INL has tested and collected data from more than 17,000 charging stations, documenting efficiencies, charge profiles, and PEV drivers’ behaviors when using charging infrastructure. Based on the experience of INL staff and support personnel, DOE requested that INL perform three tasks: • Identify pertinent lessons learned from previous direct current (DC) fast charger (DCFC) deployment and data collection activities in order to perform an analysis of the experience to date with DC fast charging systems • Establish the criteria for siting DC fast charging systems, including access to current or future sizeable penetrations of PEVs and existing power, the potential to be used by local (intracity) as well as pass-through (intercity) PEV drivers, expandability, and accessible to the federal highway system • Develop a draft system specification that lists the components that will be incorporated in a DC fast charging complexes.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
NASA NTRS1994

Goal Directed Model Inversion: A Study of Dynamic Behavior

Goal Directed Model Inversion (GDMI) is an algorithm designed to generalize supervised learning to the case where target outputs are not available to the learning system. The output of the learning system becomes the input to some external device or transformation, and only the output of this device or transformation can be compared to a desired target. The fundamental driving mechanism of GDMI is to learn from success. Given that a wrong outcome is achieved, one notes that the action that produced that outcome 0 "would have been right if the outcome had been the desired one." The algorithm then proceeds as follows: (1) store the action that produced the wrong outcome as a "target" (2) redefine the wrong outcome as a desired goal (3) submit the new desired goal to the system (4) compare the new action with the target action and modify the system by using a suitable algorithm for credit assignment (Back propagation in our example) (5) resubmit the original goal. Prior publications by our group in this area focused on demonstrating empirical results based on the inverse kinematic problem for a simulated robotic arm. In this paper we apply the inversion process to much simpler analytic functions in order to elucidate the dynamic behavior of the system and to determine the sensitivity of the learning process to various parameters. This understanding will be necessary for the acceptance of GDMI as a practical tool.

Colombano, Silvano P.↗
DOE OSTI2024

Structured Neural Network Modeling for Developing Digital Twins Models of Hydropower Generation Units

Dynamic modeling is a key part in the development of digital twin (DT) for dynamic systems. This is true for hydropower systems, where whole system modeling including penstock, turbine and generators, etc is important in realizing actuate modeling for the real systems. On the other hand, in response to the large variations of the power demand due to increased penetration of renewables such as wind and solar, hydropower systems are now required to operate in a large power generation range. This situation triggers the nonlinear characteristics of the generation unit with respect to its models. As such, it is imperative to use data driven modeling such as neural networks to learn the nonlinear dynamics of the hydropower generation unit. To achieve this objective, this study constructs a modeling and learning algorithm integrated with multiple structured neural network models for the modeling of turbine shaft speed, penstock pressure, and generator power output based on the generator power control setpoint, field current, and field voltage. In addition, the study uses the hydropower data from Tacoma Public Utilities to train and validate the proposed neural network algorithm. The results have shown that this structured neural network modeling approach can learn the system dynamics effectively by using the real-time data collected from the hydropower system with the desired modeling results.

Wang, Hong↗
DOE OSTI2024

Computational design of high entropy alloy coating for hydrogen turbine applications

This project aims to develop novel high entropy alloy (HEA)-based coatings to protect critical components in hydrogen-fueled turbine power system. The HEA-coatings will demonstrate superior performance in hydrogen combustion environment to commercial NiCoCrAlY coating in current natural gas turbine system. The HEA coating facilitates the formation of a protective scale of alpha-alumina to slow down the inward diffusion of oxidizing species and the outward diffusion of metal elements, and possesses ultrahigh corrosion and spallation resistance to prolong the service lifetime of critical components in hydrogen turbine power system. Aimed to accelerate the discovery of novel HEA coating compositions, high throughput computational modeling including CALPHAD and density functional theory and machine learning are performed to predict phase stability, oxygen permeability, oxidation rate constant, coefficient of thermal expansion, and mechanical properties. Based on the modeling and machine learning prediction, experimental validation is performed. Preliminary results will be presented and approaches to minimize oxidation will be discussed.

alloy design↗
NASA NTRS2004

Space Network Devices Developed

The NASA Glenn Research Center through a contract with Spectrum Astro, Inc., has been developing space network hardware as an enabling technology using open systems interconnect (OSI) standards for space-based communications applications. The OSI standard is a well-recognized layered reference model that specifies how data should be sent node to node in a communications network. Because of this research and technology development, a space-qualifiable Ethernet-based network interface card (similar to the type found in a networked personal computer) and the associated four-port hub were designed and developed to flight specifications. During this research and development, there also have been many lessons learned for determining approaches for migrating existing spacecraft architectures to an OSI-network-based platform. Industry has recognized the benefits of targeting hardware developed around OSI standards such as Transmission Control Protocol/Internet Protocol (TCP/IP) or similar protocols for use in future generations of space communication systems. Some of these tangible benefits include overall reductions in mission schedule and cost and in system complexity. This development also brings us a step closer to the realization of a principal investigator on a terrestrial Internet site being able to interact with space platform assets in near real time. To develop this hardware, Spectrum Astro first conducted a technology analysis of alternatives study. For this analysis, they looked at the features of three protocol specifications: Ethernet (IEEE 802.3), Firewire (IEEE 1394), and Spacewire (IEEE 1355). A thorough analysis was performed on the basis of criteria such as current protocol performance and suitability for future space applications. Spectrum Astro also projected future influences such as cost, hardware and software availability, throughput performance, and integration procedures for current and transitive space architectures. After a thorough analysis, Ethernet was chosen because it was seen as the best longer term fit because of the prevalent commercial market; the current and projected availability of hardware, software, and development tools; and the ease of architecture integration.

Jones, Robert E.↗
DOE OSTI2024

Towards verifiable cancer digital twins: tissue level modeling protocol for precision medicine

Cancer exhibits substantial heterogeneity, manifesting as distinct morphological and molecular variations across tumors, which frequently undermines the efficacy of conventional oncological treatments. Developments in multiomics and sequencing technologies have paved the way for unraveling this heterogeneity. Nevertheless, the complexity of the data gathered from these methods cannot be fully interpreted through multimodal data analysis alone. Mathematical modeling plays a crucial role in delineating the underlying mechanisms to explain sources of heterogeneity using patient-specific data. Intra-tumoral diversity necessitates the development of precision oncology therapies utilizing multiphysics, multiscale mathematical models for cancer. This review discusses recent advancements in computational methodologies for precision oncology, highlighting the potential of cancer digital twins to enhance patient-specific decision-making in clinical settings. We review computational efforts in building patient-informed cellular and tissue-level models for cancer and propose a computational framework that utilizes agent-based modeling as an effective conduit to integrate cancer systems models that encode signaling at the cellular scale with digital twin models that predict tissue-level response in a tumor microenvironment customized to patient information. Furthermore, we discuss machine learning approaches to building surrogates for these complex mathematical models. These surrogates can potentially be used to conduct sensitivity analysis, verification, validation, and uncertainty quantification, which is especially important for tumor studies due to their dynamic nature.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

Generative Vulnerability Assessment for Cyber-Physical Systems

Cyber-physical systems (CPS) are highly susceptible to malicious attacks due to their complex dynamics and interconnectivity. A comprehensive understanding of their vulnerabilities is essential for designing effective resilience measures. This paper presents a data-driven attack generative system for evaluating the vulnerability of CPS. The proposed approach formulates the vulnerability assessment problem as determining the feasibility of a specific attack set based on two boundary functions that represent the effectiveness and stealthiness of attacks. The attack generative model is trained using a custom loss function, with two universal approximators designed to learn the effectiveness and stealthiness functions simultaneously. Theoretical results for successful generation and asymptotic convergence of the resulting training algorithm are given. As a result, the proposed approach is evaluated via numerical simulation of an IEEE 14-bus system and gas pipeline systems, demonstrating its viability in learning how to attack nonlinear CPS and identify potential vulnerabilities.

Computer systems organization↗
DOE OSTI2024

Long-Term Statistical Process Monitoring of an Ultrafiltration Water Treatment Process

As water treatment technology has improved, the amount of available process data has substantially increased, making real-time, data-driven fault detection a reality. One shortcoming of the fault detection literature is that methods are usually evaluated by comparing their performance on hand-picked, short-term case studies, which yields no insight into long-term performance. In this work, we first evaluate multiple statistical and machine learning approaches for detrending process data. Then, we evaluate the performance of a PCA-based fault detection approach, applied to the detrended data, to monitor influent water quality, filtrate quality, and membrane fouling of an ultrafiltration membrane system for indirect potable reuse. Based on two short case studies, the adaptive lasso detrending method is selected, and the performance of the multivariate approach is evaluated over more than a year. The method is tested for different sets of three critical tuning parameters, and we find that for long-term, autonomous monitoring to be successful, these parameters should be carefully evaluated. However, in comparison with industry standards of simpler, univariate monitoring or daily pressure decay tests, multivariate monitoring produces substantial benefits in long-term testing.

ammonia↗
NASA NTRS1997

Towards a Self-Configuring Optimization System for Spacecraft Design

In this paper, we propose the use of a set of generic, metaheuristic optimization algorithms, which is configured for a particular optimization problem by an adaptive problem solver based on artificial intelligence and machine learning techniques. We describe work in progress on these principles.

Metaheuristic Optimization Algorithm OASIS↗
DOE OSTI2024

DS-GL: Advancing Graph Learning via Harnessing the Power of Nature within Dynamic Systems

With the rapid digitization of the world, an increasing number of real-world applications are turning to nonEuclidean data, modeled as graphs. Due to their intrinsic high complexity and irregularity, learning from graph data demands tremendous computational power. Recently, CMOS-compatible Ising machines, i.e., dynamic systems composed of CMOS components, have emerged as a new approach that harnesses the inherent power of natural annealing within dynamic systems to efficiently resolve binary optimization problems and have been adopted for traditional graph computation, such as max-cut. However, when performing complex Graph Learning (GL) tasks, Ising machines face significant hurdles: (i) they are inherently binary and thus ill-suited for real-valued problems; (ii) their expensive all-to-all coupling network that guarantees effective natural annealing poses daunting scalability concerns. To address these challenges, this paper proposes a nature-powered graph learning framework dubbed DS-GL, which is the first effort to transform the process of solving graph learning problems into the natural annealing process within a parameterized dynamic system embodied as a CMOS chip. To tackle the two major hurdles, DS-GL first augments the Ising machine architecture to modify the self-reaction term of its Hamiltonian function from linear to quadratic, effectively serving as an energy regulator. This adjustment maintains the system’s original physical interpretation while enabling it to process continuous, real-valued data. Second, to address the scaling issue, DS-GL further upgrades the real-valued dense Ising machine by decomposing it into a mesh-based multi-PE dynamic system that supports efficient distributed spatial-temporal co-annealing across different PEs through sparse interconnects. By exploiting the inherent sparsity and component structures in real-world graphs, DS-GL is able to map complex graph learning tasks onto the scalable dynamic system while maintaining high accuracy. Evaluations with three diverse GL applications across six real-world datasets, including traffic flow and COVID-19 prediction, show that DS-GL can deliver from 102× to 106× speedups and 500× energy reduction over Graph Neural Networks on GPUs, with 5% - 20% accuracy enhancement.

Song, Ruibing↗