Search NASA⌕ Search

SEARCH · Search NASA

Results for “hardware neural networks”

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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
NASA NTRS1991

Neural Architectures for Control

The cerebellar model articulated controller (CMAC) neural architectures are shown to be viable for the purposes of real-time learning and control. Software tools for the exploration of CMAC performance are developed for three hardware platforms, the MacIntosh, the IBM PC, and the SUN workstation. All algorithm development was done using the C programming language. These software tools were then used to implement an adaptive critic neuro-control design that learns in real-time how to back up a trailer truck. The truck backer-upper experiment is a standard performance measure in the neural network literature, but previously the training of the controllers was done off-line. With the CMAC neural architectures, it was possible to train the neuro-controllers on-line in real-time on a MS-DOS PC 386. CMAC neural architectures are also used in conjunction with a hierarchical planning approach to find collision-free paths over 2-D analog valued obstacle fields. The method constructs a coarse resolution version of the original problem and then finds the corresponding coarse optimal path using multipass dynamic programming. CMAC artificial neural architectures are used to estimate the analog transition costs that dynamic programming requires. The CMAC architectures are trained in real-time for each obstacle field presented. The coarse optimal path is then used as a baseline for the construction of a fine scale optimal path through the original obstacle array. These results are a very good indication of the potential power of the neural architectures in control design. In order to reach as wide an audience as possible, we have run a seminar on neuro-control that has met once per week since 20 May 1991. This seminar has thoroughly discussed the CMAC architecture, relevant portions of classical control, back propagation through time, and adaptive critic designs.

Peterson, James K.↗
NASA NTRS1994

Photonics: From target recognition to lesion detection

Since 1989, Martin Marietta has invested in the development of an innovative concept for robust real-time pattern recognition for any two-dimensioanal sensor. This concept has been tested in simulation, and in laboratory and field hardware, for a number of DOD and commercial uses from automatic target recognition to manufacturing inspection. We have now joined Rose Health Care Systems in developing its use for medical diagnostics. The concept is based on determining regions of interest by using optical Fourier bandpassing as a scene segmentation technique, enhancing those regions using wavelet filters, passing the enhanced regions to a neural network for analysis and initial pattern identification, and following this initial identification with confirmation by optical correlation. The optical scene segmentation and pattern confirmation are performed by the same optical module. The neural network is a recursive error minimization network with a small number of connections and nodes that rapidly converges to a global minimum.

Henry, E. Michael↗
NASA NTRS1994

Transition From NASA Space Communication Systems to Commerical Communication Products

Transitioning from twenty-five years of space communication system architecting, engineering and development to creating and marketing of commercial communication system hardware and software products is no simple task for small, high-tech system engineering companies whose major source of revenue has been the U.S. Government. Yet, many small businesses are faced with this onerous and perplexing task. The purpose of this talk/paper is to present one small business (LinCom) approach to taking advantage of the systems engineering expertise and knowledge captured in physical neural networks and simulation software by supporting numerous National Aeronautics and Space Administration (NASA) and the Department of Defense (DoD) projects, e.g., Space Shuttle, TDRSS, Space Station, DCSC, Milstar, etc. The innovative ingredients needed for a systems house to transition to a wireless communication system products house that supports personal communication services and networks (PCS and PCN) development in a global economy will be discussed. Efficient methods for using past government sponsored space system research and development to transition to VLSI communication chip set products will be presented along with notions of how synergy between government and industry can be maintained to benefit both parties.

Ghazvinian, Farzad↗
DOE OSTI2026

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multiple efforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680,000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

Hawks, Benjamin [Fermilab] (ORCID:0000000157000288↗
DOE OSTI2025

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multipleefforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of synthesized ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680 000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

Hawks, Benjamin G. [Fermilab]↗
DOE OSTI2026

Generative Thermodynamic Computing

Here, we introduce a generative modeling framework for thermodynamic computing, in which structured data are synthesized from noise by the natural time evolution of a physical system governed by Langevin dynamics. While conventional diffusion models use neural networks to perform denoising, here the information needed to generate structure from noise is encoded by the dynamics of a thermodynamic system. Training proceeds by maximizing the probability with which the computer generates the reverse of a noising trajectory, which ensures that the computer generates data with minimal heat emission. We demonstrate this framework within a digital simulation of a thermodynamic computer. If realized in analog hardware, such a system would function as a generative model that produces structured samples without the need for artificially injected noise or active control of denoising.

Whitelam, Stephen [Lawrence Berkeley National Labo↗
DOE OSTI2025

jaxhps: An elliptic PDE solver built with machine learning in mind

Elliptic partial differential equations (PDEs) can model many physical phenomena, such as electrostatics, acoustics, wave propagation, and diffusion. In scientific machine learning settings, a high-throughput PDE solver may be required to generate a training dataset, run in the inner loop of an iterative algorithm, or interface directly with a deep neural network. To provide value to machine learning users, such a PDE solver must be compatible with standard automatic differentiation frameworks, scale efficiently when run on graphics processing units (GPUs), and maintain high accuracy for a large range of input parameters. We have designed the jaxhps package with these use-cases in mind by implementing a highly efficient and accurate solver for elliptic problems with native hardware acceleration and automatic differentiation support.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1991

Neurocontrol and fuzzy logic: Connections and designs

Artificial neural networks (ANNs) and fuzzy logic are complementary technologies. ANNs extract information from systems to be learned or controlled, while fuzzy techniques mainly use verbal information from experts. Ideally, both sources of information should be combined. For example, one can learn rules in a hybrid fashion, and then calibrate them for better whole-system performance. ANNs offer universal approximation theorems, pedagogical advantages, very high-throughput hardware, and links to neurophysiology. Neurocontrol - the use of ANNs to directly control motors or actuators, etc. - uses five generalized designs, related to control theory, which can work on fuzzy logic systems as well as ANNs. These designs can copy what experts do instead of what they say, learn to track trajectories, generalize adaptive control, and maximize performance or minimize cost over time, even in noisy environments. Design tradeoffs and future directions are discussed throughout.

Werbos, Paul J.↗
NASA NTRS1996

Development of Collaborative Research Initiatives to Advance the Aerospace Sciences-via the Communications, Electronics, Information Systems Focus Group

The primary goal of the Adaptive Vision Laboratory Research project was to develop advanced computer vision systems for automatic target recognition. The approach used in this effort combined several machine learning paradigms including evolutionary learning algorithms, neural networks, and adaptive clustering techniques to develop the E-MOR.PH system. This system is capable of generating pattern recognition systems to solve a wide variety of complex recognition tasks. A series of simulation experiments were conducted using E-MORPH to solve problems in OCR, military target recognition, industrial inspection, and medical image analysis. The bulk of the funds provided through this grant were used to purchase computer hardware and software to support these computationally intensive simulations. The payoff from this effort is the reduced need for human involvement in the design and implementation of recognition systems. We have shown that the techniques used in E-MORPH are generic and readily transition to other problem domains. Specifically, E-MORPH is multi-phase evolutionary leaming system that evolves cooperative sets of features detectors and combines their response using an adaptive classifier to form a complete pattern recognition system. The system can operate on binary or grayscale images. In our most recent experiments, we used multi-resolution images that are formed by applying a Gabor wavelet transform to a set of grayscale input images. To begin the leaming process, candidate chips are extracted from the multi-resolution images to form a training set and a test set. A population of detector sets is randomly initialized to start the evolutionary process. Using a combination of evolutionary programming and genetic algorithms, the feature detectors are enhanced to solve a recognition problem. The design of E-MORPH and recognition results for a complex problem in medical image analysis are described at the end of this report. The specific task involves the identification of vertebrae in x-ray images of human spinal columns. This problem is extremely challenging because the individual vertebra exhibit variation in shape, scale, orientation, and contrast. E-MORPH generated several accurate recognition systems to solve this task. This dual use of this ATR technology clearly demonstrates the flexibility and power of our approach.

Knasel, T. Michael↗
DOE OSTI2020

GRANT: Ground-Roaming Autonomous Neuromorphic Targeter

In this work we describe the design, implementation, and testing of the first neuromorphic robot capable of obstacle avoidance, grid coverage, and targeting controlled by the second generation Dynamic Adaptive Neural Network Array (DANNA2) digital spiking neuromorphic processor. The simplicity of the DANNA2 processor along with the TENNLab hardware/software co-design framework allows for compact spiking networks that can run efficiently on a small, resource-constrained, platform such as a Xilinx Artix-7 field-programmable gate array. Additionally, we present the dynamic reconfigurability of DANNA2 arrays as a method of realizing complex, multi-objective tasks on hardware that is restricted to relatively small networks.

Ambrose, Jonathan↗
NASA NTRS2003

Neural-net Processed Electronic Holography for Rotating Machines

This report presents the results of an R&D effort to apply neural-net processed electronic holography to NDE of rotors. Electronic holography was used to generate characteristic patterns or mode shapes of vibrating rotors and rotor components. Artificial neural networks were trained to identify damage-induced changes in the characteristic patterns. The development and optimization of a neural-net training method were the most significant contributions of this work, and the training method and its optimization are discussed in detail. A second positive result was the assembly and testing of a fiber-optic holocamera. A major disappointment was the inadequacy of the high-speed-holography hardware selected for this effort, but the use of scaled holograms to match the low effective resolution of an image intensifier was one interesting attempt to compensate. This report also discusses in some detail the physics and environmental requirements for rotor electronic holography. The major conclusions were that neural-net and electronic-holography inspections of stationary components in the laboratory and the field are quite practical and worthy of continuing development, but that electronic holography of moving rotors is still an expensive high-risk endeavor.

Decker, Arthur J.↗
NASA NTRS2021

Microwave Radiometer RFI Detection Using Deep Learning

Radio frequency interference (RFI) is a risk for microwave radiometers due to their requirement of very high sensitivity. The Soil Moisture Active Passive (SMAP) mission has an aggressive approach to RFI detection and filtering using dedicated spaceflight hardware and ground processing software. As more sensors push to observe at larger bandwidths in unprotected or shared spectrum, RFI detection continues to be essential. This article presents a deep learning approach to RFI detection using SMAP spectrogram data as input images. The study utilizes the benefits of transfer learning to evaluate the viability of this method for RFI detection in microwave radiometers. The well-known pretrained convolutional neural networks, AlexNet, GoogleNet, and ResNet-101 were investigated. ResNet-101 provided the highest accuracy with respect to validation data (99%), while AlexNet exhibited the highest agreement with SMAP detection (92%).

Microwave radiometry↗
DOE OSTI2023

SwitchX : Gmin-Gmax Switching for Energy-efficient and Robust Implementation of Binarized Neural Networks on ReRAM Xbars

Memristive crossbars can efficiently implement Binarized Neural Networks (BNNs) wherein the weights are stored in high-resistance states (HRS) and low-resistance states (LRS) of the synapses. We propose SwitchX mapping of BNN weights onto ReRAM crossbars such that the impact of crossbar non-idealities, that lead to degradation in computational accuracy, are minimized. Essentially, SwitchX maps the binary weights in such a manner that a crossbar instance comprises of more HRS than LRS synapses. We find BNNs mapped onto crossbars with SwitchX to exhibit better robustness against adversarial attacks than the standard crossbar mapped BNNs, the baseline. Finally, we combine SwitchX with state-aware training (that further increases the feasibility of HRS states during weight mapping) to boost the robustness of a BNN on hardware. We find that this approach yields stronger defense against adversarial attacks than adversarial training, a state-of the-art software defense. We perform experiments on a VGG16 BNN with benchmark datasets (CIFAR-10, CIFAR-100 and TinyImagenet) and use Fast Gradient Sign Method (ϵ = 0.05 to 0.3) and Projected Gradient Descent (ϵ = $\frac{2}{255}$ to $\frac{32}{255}$, α = $\frac{2}{255}$) adversarial attacks. We show that SwitchX combined with state-aware training can yield upto ~35% improvements in clean accuracy and ~6–16% in adversarial accuracies against conventional BNNs. Furthermore, an important by-product of SwitchX mapping is increased crossbar power savings, owing to an increased proportion of HRS synapses, which is furthered with state-aware training. We obtain upto ~21–22% savings in crossbar power consumption for state-aware trained BNN mapped via SwitchX on 16 × 16 and 32 × 32 crossbars using the CIFAR-10 and CIFAR-100 datasets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Real Space Imaging of Field-Driven Decision-Making in Nanomagnetic Galton Boards

A possible spintronic route to hardware implementation for decision-making involves injecting a domain wall into a bifurcated magnetic nanostrip resembling a Y-shaped junction. A decision is made when the domain wall chooses a particular path through the bifurcation. Recently, it was shown that a structure like a nanomagnetic Galton board, which is essentially an array of interconnected Y-shaped junctions, produces outcomes that are stochastic and therefore relevant to artificial neural networks. However, the exact mechanism leading to the robust nature of randomness is unknown. Here, in this study, we directly image the decision-making process in nanomagnetic Galton boards using Lorentz transmission electron microscopy. We identify that the stochasticity in nanomagnetic Galton boards arises as a culmination of (1) the topology of the injected domain wall, (2) dissimilarly sized vertices, and (3) the strength of the applied field. Our results pave the way to a detailed understanding of stochasticity in nanomagnetic networks.

lorentz microscopy↗
DOE OSTI2025

TrioSim: A Lightweight Simulator for Large-Scale DNN Workloads on Multi-GPU Systems

Deep Neural Networks (DNNs) have become increasingly capable of performing tasks ranging from image recognition to content generation. The training and inference of DNNs heavily rely on GPUs, as GPUs' massively parallel architecture delivers extremely high computing capability. With the growing complexity of DNNs and the size of training datasets, training DNNs with a large number of GPUs is becoming a prevalent strategy. Researchers have been exploring how to design software and hardware systems for GPU farms to achieve the best utilization, efficiency, and DNN accuracy during training or inference. However, when designing and deploying such systems, designers usually rely on testing on physical hardware platforms equipped with many GPUs, incurring high costs that are almost prohibitive for system designers to test different configurations and designs, even for highly resourceful companies. While an alternative solution is to test on GPU simulators, they are often too slow for these l

Li, Ying [William & Mary, Williamsburg, VA, USA] (↗
DOE OSTI2023

Digital Modeling on Large Kernel Metamaterial Neural Network

Deep neural networks (DNNs) utilized recently are physically deployed with computational units (e.g., CPUs and GPUs). Such a design might lead to a heavy computational burden, significant latency, and intensive power consumption, which are critical limitations in applications such as Internet of Things (IoT), edge computing, and usage of drones. Recent advances in optical computational units (e.g., metamaterial) have shed light on energy-free and light-speed neural networks. However, the digital design of the metamaterial neural network (MNN) is fundamentally limited by its physical limitations, such as precision, noise, and bandwidth during fabrication. Moreover, the unique advantages of MNN’s (e.g., light-speed computation) are not fully explored via standard 3×3 convolution kernels. In this paper, we propose a novel large kernel metamaterial neural network (LMNN) that maximizes the digital capacity of the state-of-the-art (SOTA) MNN with model re-parametrization and network compression, while also considering the optical limitation explicitly. The new digital learning scheme can maximize the learning capacity of MNN while modeling the physical restrictions of meta-optics. With the proposed LMNN, the computation cost of the convolutional front-end can be offloaded to fabricated optical hardware. The experimental results on two publicly available datasets demonstrate that the optimized hybrid design improved classification accuracy while reducing computational latency. In conclusion, the development of the proposed LMNN is a promising step towards the ultimate goal of energy-free and light-speed AI.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1995

Cascade Error Projection: An Efficient Hardware Learning Algorithm

A new learning algorithm termed cascade error projection (CEP) is presented. CEP is an adaption of a constructive architecture from cascade correlation and the dynamical stepsize of A/D conversion from the cascade back propagation algorithm.

learning algorithm pattern recognition cascade err↗
NASA NTRS2005

Artificial Neural Network Test Support Development for the Space Shuttle PRCS Thrusters

A significant anomaly, Fuel Valve Pilot Seal Extrusion, is affecting the Shuttle Primary Reaction Control System (PRCS) Thrusters, and has caused 79 to fail. To help address this problem, a Shuttle PRCS Thruster Process Evaluation Team (TPET) was formed. The White Sands Test Facility (WSTF) and Boeing members of the TPET have identified many discrete valve current trace characteristics that are predictive of the problem. However, these are difficult and time consuming to identify and trend by manual analysis. Based on this exhaustive analysis over months, 22 thrusters previously delivered by the Depot were identified as high risk for flight failures. Although these had only recently been installed, they had to be removed from Shuttles OV103 and OV104 for reprocessing, by directive of the Shuttle Project Office. The resulting impact of the thruster removal, replacement, and valve replacement was significant (months of work and hundreds of thousands of dollars). Much of this could have been saved had the proposed Neural Network (NN) tool described in this paper been in place. In addition to the significant benefits to the Shuttle indicated above, the development and implementation of this type of testing will be the genesis for potential Quality improvements across many areas of WSTF test data analysis and will be shared with other NASA centers. Future tests can be designed to incorporate engineering experience via Artificial Neural Nets (ANN) into depot level acceptance of hardware. Additionally, results were shared with a NASA Engineering and Safety Center (NESC) Super Problem Response Team (SPRT). There was extensive interest voiced among many different personnel from several centers. There are potential spin-offs of this effort that can be directly applied to other data acquisition systems as well as vehicle health management for current and future flight vehicles.

Lehr, Mark E.↗