Search NASA⌕ Search

SEARCH · Search NASA

Results for “Edge Computing”

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 127 records · Page 7

Efficient Anomaly Detection Driven By Different Machine Learning Architectures And Models

The rapid growth and ubiquitous adoption of the internet and cyber-physical systems (CPS) have fundamentally transformed modern communication, work, and human-system interactions. While networks now form the backbone of critical digital ecosystems, enabling seamless data transmission across diverse, interconnected systems, this increased connectivity also expands the attack surface, making real-time detection of network intrusions and anomalies a pressing challenge. Detecting unusual activities within network infrastructure requires advanced data traffic analysis to differentiate between legitimate and malicious interactions. Traditional approaches to network anomaly detectionâ??such as rule-based and signature-based systemsâ??often depend on predefined patterns to identify known anomalies, limiting their effectiveness against emerging, stealthy, or previously unseen threats. These conventional methods suffer from high false alarm rates and fail to adapt to the ever-evolving nature of network traffic, particularly in large-scale, decentralized environments where data volume, velocity, and variety are constantly increasing. This dissertation presents artificial intelligence (AI)-driven approaches to anomaly detection that leverage graphics processing unit (GPU)-enabled high-performance computing (HPC) platforms for processing massive network traffic data and monitoring the components of cyber-physical systems (CPS) for potentially hazardous conditions. The research advances several key contributions: (1) Designing efficient machine learning techniques for CPS condition monitoring and anomaly detection; (2) enabling federated learning (FL) frameworks that enable distributed detection while preserving data privacy and system resilience; (3) exploring graph-based methodologies combining graph neural networks (GNN) and graph machine learning (ML) approaches for the Internet of Things (IoT) and automotive network security, and (4) performing distributed edge computing optimizations that integrate FL with scalable technologies for reduced communication overhead. Through extensive experiments, these methodologies demonstrate that complex anomaly detection and condition monitoring tasks can be achieved while balancing computational efficiency and detection accuracy through fine-grained network information processing. The frameworks developed in this research establish a robust foundation for network anomaly detection, providing scalable, adaptive, and privacy-preserving solutions for safeguarding CPS and IoT networks in an increasingly interconnected digital landscape. The practical implications of these research findings are significant, as they can inform the development of next-generation network security systems and contribute to the protection of critical infrastructure against sophisticated cyber attacks.

Marfo, William↗

NPLOT: an Interactive Plotting Program for NASTRAN Finite Element Models

The NPLOT (NASTRAN Plot) is an interactive computer graphics program for plotting undeformed and deformed NASTRAN finite element models. Developed at NASA's Goddard Space Flight Center, the program provides flexible element selection and grid point, ASET and SPC degree of freedom labelling. It is easy to use and provides a combination menu and command driven user interface. NPLOT also provides very fast hidden line and haloed line algorithms. The hidden line algorithm in NPLOT proved to be both very accurate and several times faster than other existing hidden line algorithms. A fast spatial bucket sort and horizon edge computation are used to achieve this high level of performance. The hidden line and the haloed line algorithms are the primary features that make NPLOT different from other plotting programs.

Jones, G. K.↗

Learning from learning machines: improving the predictive power of energy-water-land nexus models with insights from complex measured and simulated data

Focal Area(s): Insights gleaned from complex data (observed and simulated) using AI; Predictive modeling through the use of AI techniques and AI-derived model components, including physics- and knowledge-informed models; Energy-water-land nexus and integrated energy systems – models of MultiSector dynamics. Science Challenge: Scientific communities are in need of tools for the computational integration of physics-based models, experimental data, and empirical/observational studies across a broad range of temporal and spatial scales to explore and meet EESSD grand challenges. We require AI algorithms for the discovery of process-drivers in the Earth-energy-human system and to link them with complex measured and simulated data. We aim to understand the energy-water-land nexus, particularly under extreme forcing scenarios and rare events, leveraging scale-aware AI process models, including probabilistic uncertainties, and benefiting from the emerging 5G-enabled landscape-scale sensing and edge computing capabilities. These models are needed to identify instabilities and tipping points that manifest extreme system behaviors with consequences for integrated energy systems and the environment. This work requires fundamental advances in uncertainty quantification, in particular, to identify and model unlikely but catastrophic outliers. Specifically, we need models that are interpretable to domain scientists and, essentially, explainable to public and private stake holders.

54 ENVIRONMENTAL SCIENCES↗

Enabling a Decentralized Smart Grid Using Autonomous Edge Control Devices

As a large number of distributed devices are connected to the modern smart grid, the traditional centralized connectivity models fail to provide economic value. These models have relied on sending data to the cloud for processing and receiving commands to exert control actions, resulting in an ‘on-demand system’ with high bandwidth, low latency and an overload of data on the cloud. For realizing a decentralized system, there is a strong need to embed intelligence at the ‘edge of the network’. These intelligent devices, capable of sensing, local data processing and exerting control actions, report only actionable information to the cloud, acting as an edge control node. The system can then function autonomously, without constant cloud inputs, tolerating longer delays in communication, making the overall system ultra-low cost. The Global Asset Monitoring, Management and Analytics (GAMMA) Platform is a novel ultralow- cost, secure platform that operates through a Bluetooth based delay tolerant network. It relies on so-called ‘data mules’ to bridge the last mile connectivity gap in an inherently secure way. Due to this model, the platform requires no in-country certifications, does not rely on a dedicated backhaul technology and is immune to technology migration. This architecture also addresses some gaps identified in traditional IoT-based solutions in remote areas and sparse connectivity. A functional unit of the edge computing node has been built, taking into account various constraints like costs, customizations, data storage, cybersecurity and power management. Here, the platform has been built, deployed and has demonstrated distributed smart grid applications like power quality sensing, automated metering infrastructure and utility asset monitoring.

42 ENGINEERING↗

Competitive docking model for prediction of the human nicotinic acetylcholine receptor α7 binding of tobacco constituents

The detrimental health effects associated with tobacco use constitute a major public health concern. The addiction associated with nicotine found in tobacco products has led to difficulty in quitting among users. Nicotinic acetylcholine receptors (nAChRs) are the targets of nicotine and are responsible for addiction to tobacco products. However, it is unknown if the other >8000 tobacco constituents are addictive. Since it is time-consuming and costly to experimentally assess addictive potential of such larger number of chemicals, computationally predicting human nAChRs binding is important for in silico evaluation of addiction potential of tobacco constituents and needs structures of human nAChRs. Therefore, we constructed three dimensional structures of the ligand binding domain of human nAChR α7 subtype and then developed a predictive model based on the constructed structures to predict human nAChR α7 binding activity of tobacco constituents. The predictive model correctly predicted 11 out of 12 test compounds to be binders of nAChR α7. The model is a useful tool for high-throughput screening of potential addictive tobacco constituents. These results could inform regulatory science research by providing a new validated predictive tool using cutting-edge computational methodology to high throughput screen tobacco additives and constituents for their binding interaction with the human α7 nicotinic receptor. The tool represents a prediction model capable of screening thousands of chemicals found in tobacco products for addiction potential, which improves the understanding of the potential effects of additives.

60 APPLIED LIFE SCIENCES↗

Autonomous nondestructive evaluation of resistance spot welded joints

The application of non-destructive evaluation approaches has attracted strong interests in modern automotive industries. Here, we present an autonomous deep-computing framework to analyze raw videos from infrared systems and to predict weld nugget shape and size with unprecedented accuracy and speed. In a comprehensive training and testing experiment with 90 videos (seven sets of welding material stack-ups), a new method was developed to assemble sufficient datasets for neural network training. Our framework successfully predicts all the nugget shapes with F1 scores that range from 0.84 to 0.92. The total training time on Nvidia DGX station takes less than 10 min for each set of welding material stack-up. The real inference time of an individual dataset (with 30 video frames) takes about 0.005 s. The procedure and methods developed in the study can be applied to other image-based weld property prediction, as well as other manufacturing processes. Furthermore, our well-trained neural networks take limited memory resources (2.3 MB) and are suitable for embedded microprocessors for in-situ welding quality control as edge computing within an intelligent welding framework.

42 ENGINEERING↗

Even Higher-Level Synthesis: An Exploration of AI Hardware Accelerators using HLS4ML

With the rise of artificial intelligence, the popularization of deep learning, and a constantly evolving industry, the demand for flexible and efficient tools has never been greater. As algorithms grow more complex, their runtime and energy consumption increase exponentially. Customized hardware accelerators, long used for specific mathematical operations, remain essential for managing modern applications' computational and power demands. Hardware accelerators can speed up complex computations by orders of magnitude, but their manual design and verification processes are often challenging and time-consuming. High-Level Synthesis (HLS) provides a solution by transforming high-level algorithm descriptions, typically written in C++ or SystemC, into synthesizable RTL suitable for hardware implementation. This approach reduces development time for RTL engineers while offering flexibility beyond what traditional handwritten RTL can provide. We extended this capability to the machine-learning domain with the open-source framework hls4ml, which allows neural networks trained in Python frameworks like Tensorflow or PyTorch to be synthesized into efficient hardware representations for the traditional FPGA and ASIC flows. This breakthrough addresses the growing need for reduced design turnaround and easy verification of ML hardware accelerators with low latency and power efficiency constraints. During this tutorial, we will demonstrate how Python complements HLS by simplifying the ML design process, bridging the gap between software and hardware development. Attendees will explore how we translate neural networks modeled in Python into fixed-point C++ models suitable for HLS workflows. We will dive into strategies like Value-Range Analysis and Quantization-Aware Training, which optimize these designs for deployment and evaluate their accuracy, power consumption, and energy efficiency. To exemplify these concepts, experts from Fermilab will share their experiences applying this technology to high-energy physics experiments, where real-time, low-latency processing is critical. Over the years, Fermilab engineers have demonstrated how deep neural networks, optimized for hardware using hls4ml, can meet the stringent requirements of trigger systems at the CERN Large Hadron Collider. These systems rely on rapid decision-making to process immense data volumes while retaining only the most relevant events for further analysis. The application of hls4ml has also been extended to innovative technologies like smart pixel arrays. These smart pixels integrate ML inference capabilities directly into sensor devices, enabling localized data processing at the pixel level. This approach drastically reduces the need to transmit raw data to external processing units, significantly decreasing power consumption and latency. By embedding neural networks within the pixel architecture, the smart pixels can identify and prioritize relevant data in real time, providing a highly efficient solution for edge computing in scenarios such as particle detectors and imaging systems. Fermilab's work highlights the potential of hardware-accelerated ML in scenarios where both speed and power efficiency are mission-critical. Through this tutorial, attendees will gain valuable insights into the challenges and solutions of deploying ML in hardware. Understanding how HLS and hls4ml streamline the development of neural network-based hardware accelerators is fundamental for the industry's future. Participants will learn how these technologies are shaping the future of AI and scientific computing.

Di Guglielmo, Giuseppe [Fermilab]↗

Cross Inference of Throughput Profiles Using Micro Kernel Network Method

Dedicated network connections are being increasingly deployed in cloud, centralized and edge computing and data infrastructures, whose throughput profiles are critical indicators of the underlying data transfer performance. Due to the cost and disruptions to physical infrastructures, network emulators, such as Mininet, are often used to generate measurements needed to estimate throughput profiles, typically expressed as a function of the connection round trip time. The profiles estimated using measurements from such emulated networks are usually inaccurate for high bandwidth and high latency connections, since they do not accurately reflect the critical network transport dynamics mainly due to computing and memory constraints of the host. We present a machine learning (ML) method to estimate the throughput profiles using emulation measurements to closely match the testbed and production network profiles. In particular, we propose a micro Kernel Network (mKN) that provides baseline throughput measurements on the host running Mininet emulations, which are used to learn a regression map that converts them to the corresponding testbed measurement estimates. Once initially learned, this map is applied to measurements from subsequent network emulations on the same host. We present experimental measurements to illustrate this approach, and derive generalization equations for the proposed mKN-ML method. Using a four-site scenario emulation, we show the effectiveness of this method in providing accurate concave throughput profiles from inaccurate convex or non-smooth ones indicated by Mininet emulation.

Rao, Nageswara↗

Steric Mapping, Ligand Dynamics, and Cycloisomerization Catalysis with Redox Robust Mn I/0/-I Dicarbenes

Manganese is perhaps the most electronically versatile element, yet the redox properties, reactivity, and catalytic applications of low-valent Mn 0 /Mn −I complexes remain underexplored due to the propensity for Mn 0 to dimerize, quenching highenergy metalloradicals. We report a series of redox-active monometallic Mn I , Mn 0 and Mn −I complexes containing a BH 2 - bridged dicarbene, characterized using a suite of experimental and cutting-edge computational (DFT) methods. Slow electron transfer kinetics at Mn I/0 are observed, with computations and electrochemical simulations in excellent agreement with experimental values. Despite the lack of steric bulk at the BH 2 -bridged Mn 0 , the t Bu groups at the dicarbene provide adequate steric protection to prevent dimerization, with percent buried volume (% V bur ) serving as a valuable steric ranking tool. We also show that a %V bur > 83% prevents dimerization for a diverse array of Mn 0 complexes from the literature. Ligand sterics of BPh 2 -and BH 2 -bridged complexes dictate reaction outcomes when Mn I and Mn −I are exposed to nucleophiles and electrophiles, respectively, while Mn0 facilitates the radical cycloisomerization catalysis of 6-iodo-1- hexene at room temperature. Furthermore, this work underscores the importance of ligand sterics in rationalizing reactivity patterns at Mn and provides valuable insights for designing chelating ligands that can selectively leverage Mn I/0/‑I states in redox-mediated catalytic reactions.

Ligands↗

Domain alignment within ferroelectric/dielectric PbTiO 3 /SrTiO 3 superlattice nanostructures

The ferroelectric domain pattern within lithographically defined PbTiO 3 /SrTiO 3 ferroelectric/dielectric heteroepitaxial superlattice nanostructures is strongly influenced by the edges of the structures. Synchrotron X-ray nanobeam diffraction reveals that the spontaneously formed 180° ferroelectric stripe domains exhibited by such superlattices adopt a configuration in rectangular nanostructures in which domain walls are aligned with long patterned edges. The angular distribution of X-ray diffuse scattering intensity from nanodomains indicates that domains are aligned within an angular range of approximately 20° with respect to the edges. Computational studies based on a time-dependent Landau–Ginzburg–Devonshire model show that the preferred direction of the alignment results from lowering of the bulk and electrostrictive contributions to the free energy of the system due to the release of the lateral mechanical constraint. This unexpected alignment appears to be intrinsic and not a result of distortions or defects caused by the patterning process. Thus, our work demonstrates how nanostructuring and patterning of heteroepitaxial superlattices allow for pathways to create and control ferroelectric structures that may appear counterintuitive.

36 MATERIALS SCIENCE↗

Preface for the Sherwood Fusion Theory 2023 special collection

Pursuing controlled fusion is a monumental endeavor that promises to unlock a transformational source of clean energy. Here, the International Sherwood Fusion Theory Conference is a crucial platform for researchers to discuss the physics governing plasmas in fusion-relevant configurations. From exploring the fundamental properties of plasma to cutting-edge computer simulation techniques, this conference fosters innovation and collaboration, propelling us closer to realizing the dream of a fusion-powered future.

Beidler, Matthew T. [Oak Ridge National Laboratory↗

Probe microscopy is all you need *

We pose that microscopy offers an ideal real-world experimental environment for the development and deployment of active Bayesian and reinforcement learning methods. Indeed, the tremendous progress achieved by machine learning (ML) and artificial intelligence over the last decade has been largely achieved via the utilization of static data sets, from the paradigmatic MNIST to the bespoke corpora of text and image data used to train large models such as GPT3, DALL·E and others. However, it is now recognized that continuous, minute improvements to state-of-the-art do not necessarily translate to advances in real-world applications. We argue that a promising pathway for the development of ML methods is via the route of domain-specific deployable algorithms in areas such as electron and scanning probe microscopy and chemical imaging. This will benefit both fundamental physical studies and serve as a test bed for more complex autonomous systems such as robotics and manufacturing. Favorable environment characteristics of scanning and electron microscopy include low risk, extensive availability of domain-specific priors and rewards, relatively small effects of exogenous variables, and often the presence of both upstream first principles as well as downstream learnable physical models for both statics and dynamics. Recent developments in programmable interfaces, edge computing, and access to application programming interfaces (APIs) facilitating microscope control, all render the deployment of ML codes on operational microscopes straightforward. We discuss these considerations and hope that these arguments will lead to create novel set of development targets for the ML community by accelerating both real world ML applications and scientific progress.

47 OTHER INSTRUMENTATION↗

Accelerating data acquisition with FPGA-based edge machine learning: a case study with LCLS-II

New scientific experiments and instruments generate vast amounts of data that need to be transferred for storage or further processing, often overwhelming traditional systems. Edge machine learning (EdgeML) addresses this challenge by integrating machine learning (ML) algorithms with edge computing, enabling real-time data processing directly at the point of data generation. EdgeML is particularly beneficial for environments where immediate decisions are required, or where bandwidth and storage are limited. In this paper, we demonstrate a high-speed configurable ML model in a fully customizable EdgeML system using a field programmable gate array (FPGA). Our demonstration focuses on an angular array of electron spectrometers, referred to as the ‘CookieBox,’ developed for the Linac Coherent Light Source II project. The EdgeML system captures 51.2 Gbps from a 6.4 GS s −1 analog to digital converter and is designed to integrate data pre-processing and ML inside an FPGA. Our implementation achieves an inference latency of 0.2 µs for the ML model, and a total latency of 0.4 µs for the complete EdgeML system, which includes pre-processing, data transmission, digitization, and ML inference. The modular design of the system allows it to be adapted for other instrumentation applications requiring low-latency data processing.

97 MATHEMATICS AND COMPUTING↗

High-temperature Majorana fermions in magnet-superconductor hybrid systems

Magnet-superconductor hybrid (MSH) structures represent one of the most promising platforms to realize, control, and manipulate Majorana modes using scanning tunneling methods. By depositing either chains or islands of magnetic atoms on the surface of a conventional, elemental superconductor such as Pb or Re, topological superconducting phases can emerge. They feature either localized Majorana bound states at the chain ends or dispersing chiral Majorana modes at the island's boundary. However, some of these experiments have not reached the spectral resolution to clearly distinguish between topological Majorana and trivial Shiba states due to very small superconducting gap sizes and experiments performed at sub-Kelvin temperatures. Here we consider superconducting substrates with unconventional spin-singlet pairing, including high-temperature d -wave and extended s -wave superconductors. We derive topological phase diagrams and compute edge states for cylinder and island geometries and discuss their properties. Several time-reversal invariant topological superconducting phases of the Zhang-Kane-Mele type are found and discussed. Addiotnally, we review one-dimensional MSH structures and show that parameters to realize topologically nontrivial magnetic chains embedded into a larger, two-dimensional substrate differ from the purely one-dimensional case. Quite generally we find that unconventional superconducting substrates work as well as the conventional s -wave substrates to realize topological phases. In particular, iron-based pnictide and chalcogenide superconductors are the most promising class of substrates for future high-temperature MSH systems.

36 MATERIALS SCIENCE↗

Contrasting advantages of learning with random weights and backpropagation in non-volatile memory neural networks

Recently, a Cambrian explosion of novel, non-volatile memory (NVM) devices known as memristive devices have inspired effort in building hardware neural networks that learn like the brain. Early experimental prototypes built simple perceptrons from nanosynapses, and recently, fully-connected multi-layer perceptron (MLP) learning systems have been realized. However, while backpropagating learning systems pair well with high-precision computer memories and achieve state-of-the-art performances, this typically comes with a massive energy budget. For future Internet of Things/peripheral use cases, system energy footprint will be a major constraint, and emerging NVM devices may fill the gap by sacrificing high bit precision for lower energy. In this work, we contrast the well known MLP approach with the Extreme Learning Machine (ELM) or NoProp approach, which uses a large layer of random weights to improve the separability of high-dimensional tasks, and is usually considered inferior in a software context. However, we find that when taking device non-linearity into account, NoProp manages to equal hardware MLP system in terms of accuracy. While also using a sign-based adaptation of the delta rule for energy-savings, we find that NoProp can learn effectively with four to six ’bits’ of device analog capacity, while MLP requires eight bit capacity with the same rule. This may allow the requirements for memristive devices to be relaxed in the context of online learning. By comparing the energy footprint of these systems for several candidate nanosynapses, as well as realistic peripherals, we confirm that memristive NoProp systems save energy compared to MLP systems. Lastly, we show that ELM/NoProp systems can achieve better generalization abilities than nanosynaptic MLP systems when paired with pre-processing layers (which do not require backpropagated error). Collectively, these advantages make such systems worthy of consideration in future accelerators or embedded hardware.

97 MATHEMATICS AND COMPUTING↗

Incremental Threshold Scheme Enabled IoT Group Key Management

Cyber landscape evolves rapidly. Internet of Things (IoT) and Edge Computing (EC) have rapidly become an integral part of the modern computing infrastructure. It is expected that there will be more than 50 billion active and connected IoT devices by 2025 [1]. Pervasive IoT/EC creates unprecedented opportunities bridging the gap between previously segregated cyber and physical spaces. However, this progress also brings along new security challenges. IoT devices typically have limited computation, communication, and storage resources. This leads to security architecture designs such as using symmetric keys for group communication. While secure and efficient in stable network settings, symmetric key solutions are ill-adapted for IoT's highly dynamic device mobility behavior and frequent group membership turnover. Whenever IoT members leave a group, the known symmetric keys cannot be made forgotten, posing a serious vulnerability. This leads to frequent re-groupings that require expensive re-authentication, key regeneration, and key redistribution in order to maintain IoT/EC security. We present a novel symmetric key management framework that integrate an Incremental Threshold Scheme (ITS) cryptographical function into communication protocol's key rotation mechanism to allow for secure and efficient symmetric key communication group member node revocation. This ITS-enabled key management framework alleviates the need of frequent and expensive re-grouping and re-keying needed by today's large and dynamic IoT/EC operations. We further applied this ITS-enabled key management framework to a distributed IoT/EC-integrated publish and subscribe framework for applicability validation.

Li, Mingyan↗

Fast Vehicle Turning-Movement Counting using Localization-based Tracking

Despite the high utility of traffic volume and turning movement data, such data is still hard to come by for the vast majority of roadways and intersections in nearly ev- ery city. Edge computing devices offer a promising tool for recording turning movement data if lightweight algorithms can be designed to run in real-time with relatively modest computational complexity. To that end, this work presents Vehicle Turning-Movement Counting using Localization- based Tracking (LBT-Count). This method is fast because it never performs detection on a full frame. Instead, only a few portions of the image are cropped and used to de- tect objects within the frame. The method achieves com- petitive performance on the public evaluation server for Track 1 of the AI City Challenge (7th overall on the first 50% of data). Furthermore, we show that LBT-Count is 52% faster than an analogous counting algorithm utilizing a traditional tracking-by-detection framework on available challenge data.

42 ENGINEERING↗

Engineering Privacy at the Edge: A Practical Guide to Differential Privacy in System Architectures

The rapid expansion of distributed and edge computing platforms—spanning autonomous vehicles, IoT sensors, and healthcare monitors—has heightened concerns about data privacy. Differential Privacy (DP) offers a rigorous mathematical framework to protect sensitive information while retaining analytical utility. This tutorial introduces the foundations of DP for both numerical and categorical datasets and extends the discussion to correlation-aware techniques tailored for structured and high-dimensional data. Hands-on demonstrations will begin with the PETINA (Privacy prEservaTIoN Algorithms) package for numerical data and continue with MIC-DP (Maximum Information Correlated Differential Privacy) for tabular data. Designed for researchers and practitioners in secure systems, embedded architectures, and AI accelerators, the tutorial emphasizes practical and scalable methods for integrating DP into real-world system designs.

Kotevska, Olivera [ORNL] (ORCID:0000000316772243)↗