SEARCH · Search NASA

Results for “DL”

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.

SourceResearch connectionFrom yearThrough year

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

At least 109 records · Page 6

All sources NASA NTRS DOE OSTI

DOE OSTI2021

Predicting Nugget Size of Resistance Spot Welds Using Infrared Thermal Videos With Image Segmentation and Convolutional Neural Network

Resistance spot welding (RSW) is a widely adopted joining technique in automotive industry. Recent advancement in sensing technology makes it possible to collect thermal videos of the weld nugget during RSW using an infrared (IR) camera. The effective and timely analysis of such thermal videos has the potential of enabling in situ nondestructive evaluation (NDE) of the weld nugget by predicting nugget thickness and diameter. Deep learning (DL) has demonstrated to be effective in analyzing imaging data in many applications. However, the thermal videos in RSW present unique data-level challenges that compromise the effectiveness of most pre-trained DL models. We propose a novel image segmentation method for handling the RSW thermal videos to improve the prediction performance of DL models in RSW. The proposed method transforms raw thermal videos into spatial-temporal instances in four steps: video-wise normalization, removal of uninformative images, watershed segmentation, and spatial-temporal instance construction. The extracted spatial-temporal instances serve as the input data for training a DL-based NDE model. The proposed method is able to extract high-quality data with spatial-temporal correlations in the thermal videos, while being robust to the impact of unknown surface emissivity. Overall, our case studies demonstrate that the proposed method achieves better prediction of nugget thickness and diameter than predicting without the transformation.

42 ENGINEERING↗

DOE OSTI2021

Reconfigurable Framework for Resilient Semantic Segmentation for Space Applications

Deep learning (DL) presents new opportunities for enabling spacecraft autonomy, onboard analysis, and intelligent applications for space missions. However, DL applications are computationally intensive and often infeasible to deploy on radiation-hardened (rad-hard) processors, which traditionally harness a fraction of the computational capability of their commercial-off-the-shelf counterparts. Commercial FPGAs and system-on-chips present numerous architectural advantages and provide the computation capabilities to enable onboard DL applications; however, these devices are highly susceptible to radiation-induced single-event effects (SEEs) that can degrade the dependability of DL applications. In this article, we propose Reconfigurable ConvNet (RECON), a reconfigurable acceleration framework for dependable, high-performance semantic segmentation for space applications. In RECON, we propose both selective and adaptive approaches to enable efficient SEE mitigation. In our selective approach, control-flow parts are selectively protected by triple-modular redundancy to minimize SEE-induced hangs, and in our adaptive approach, partial reconfiguration is used to adapt the mitigation of dataflow parts in response to a dynamic radiation environment. Combined, both approaches enable RECON to maximize system performability subject to mission availability constraints. We perform fault injection and neutron irradiation to observe the susceptibility of RECON and use dependability modeling to evaluate RECON in various orbital case studies to demonstrate a 1.5–3.0× performability improvement in both performance and energy efficiency compared to static approaches.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

Topological Interpretability for Deep Learning

With the growing adoption of AI-based systems across everyday life, the need to understand their decision-making mechanisms is correspondingly increasing. The level at which we can trust the statistical inferences made from AI-based decision systems is an increasing concern, especially in high-risk systems such as criminal justice or medical diagnosis, where incorrect inferences may have tragic consequences. Despite their successes in providing solutions to problems involving real-world data, deep learning (DL) models cannot quantify the certainty of their predictions. These models are frequently quite confident, even when their solutions are incorrect. This work presents a method to infer prominent features in two DL classification models trained on clinical and non-clinical text by employing techniques from topological and geometric data analysis. We create a graph of a model's feature space and cluster the inputs into the graph's vertices by the similarity of features and prediction statistics. We then extract subgraphs demonstrating high-predictive accuracy for a given label. These subgraphs contain a wealth of information about features that the DL model has recognized as relevant to its decisions. We infer these features for a given label using a distance metric between probability measures, and demonstrate the stability of our method compared to the LIME and SHAP interpretability methods. This work establishes that we may gain insights into the decision mechanism of a DL model. This method allows us to ascertain if the model is making its decisions based on information germane to the problem or identifies extraneous patterns within the data.

Spannaus, Adam↗

DOE OSTI2024

Flexible silicon photonic architecture for accelerating distributed deep learning

The increasing size and complexity of deep learning (DL) models have led to the wide adoption of distributed training methods in datacenters (DCs) and high-performance computing (HPC) systems. However, communication among distributed computing units (CUs) has emerged as a major bottleneck in the training process. In this study, we propose Flex-SiPAC, a flexible silicon photonic accelerated compute cluster designed to accelerate multi-tenant distributed DL training workloads. Flex-SiPAC takes a co-design approach that combines a silicon photonic hardware platform with a tailored collective algorithm, optimized to leverage the unique physical properties of the architecture. The hardware platform integrates a novel wavelength-reconfigurable transceiver design and a micro-resonator-based wavelength-reconfigurable switch, enabling the system to achieve flexible bandwidth steering in the wavelength domain. The collective algorithm is designed to support reconfigurable topologies, enabling efficient all-reduce communications that are commonly used in DL training. The feasibility of the Flex-SiPAC architecture is demonstrated through two testbed experiments. First, an optical testbed experiment demonstrates the flexible routing of wavelengths by shuffling an array of input wavelengths using a custom-designed spatial-wavelength selective switch. Second, a four-GPU testbed running two DL workloads shows a 23% improvement in job completion time compared to a similarly sized leaf-spine topology. We further evaluate Flex-SiPAC using large-scale simulations, which show that Flex-SiPAC is able to reduce the communication time by 26% to 29% compared to state-of-the-art compute clusters under representative collective operations.

Wu, Zhenguo (ORCID:0000000322847985)↗

DOE OSTI2023

A Comparative Study of Deep Learning Models for Fracture and Pore Space Segmentation in Synthetic Fractured Digital Rocks

This study focuses on the comparative study of deep learning (DL) models for pore space and discrete fracture networks (DFNs) segmentation in synthetic fractured digital rocks, specifically targeting low-permeability rock formations, such as shale and tight sandstones. Accurate characterization of pore space and DFNs is critical for subsequent property analysis and fluid flow modeling. Four DL models, SegNet, U-Net, U-Net-wide, and nested U-Net (i.e., U-Net++), were trained, validated, and tested using synthetic datasets, including input and label image pairs with varying properties. The model performance was assessed regarding pixel-wise metrics, including the F1 score and pixel-wise difference maps. In addition, the physics-based metrics were considered for further analysis, including sample porosity and absolute permeability. Particularly, We first simulated the permeability of porous media containing only pore space and then simulated the permeability of porous media with DFNs added. The difference between these two values is used to quantify the connectivity of segmented DFNs, which is an important parameter for low-permeability rocks. The pixel-wise metrics showed that the nested U-Net model outperformed the rest of the DL models in pore space and DFNs segmentation, with the SegNet model exhibiting the second-best performance. Particularly, nested U-Net enhanced segmentation accuracy for challenging boundary pixels affected by partial volume effects. The U-Net-wide model achieved improved accuracy compared to the U-Net model, which indicated the influence of parameter numbers. Similarly, nested U-Net has the closest match to the ground truth of physics-based metrics, including the porosity of pore space and DFNs, and the permeability difference quantifying the connectivity of DFNs. The findings highlight the effectiveness of DL models, especially the U-Net++ model with nested architecture and redesigned skip connections, in accurately segmenting pore spaces and DFNs, which are crucial for pore-scale fluid flow and transport simulation in low-permeability rocks.

Wang, Hongsheng↗

DOE OSTI2023

PRIMED for the Future: Purposing Raw Intake for Machine Learning-Enabled Detection (Final Report)

The COVID-19 pandemic demonstrated how a novel, elusive, and diffuse biological threat can engender uncertainty and misinformation, and it underscored the need for flexible analytical modalities agnostic to the identity of biological material. Yet even before the pandemic, recognition of the limitations of the current, list-based approach, which focuses on known pathogens and biotoxins, and of the importance of agent-agnostic biodetection was growing within the biosecurity community. In a 2018 report on “Biodefense in the Age of Synthetic Biology,” for example, the National Academy of Sciences stated that “an overreliance on the Select Agent List is a systematic weakness affecting many aspects of the United States’ current biodefense mitigation capability”. More recently, a group of biodefense researchers proposed the identification and adoption of “bioagent-agnostic signatures (BASs)” as a way of detecting and characterizing not only existing agents but also novel ones, an approach they believe will “enable a more flexible and resilient biodefense posture”. Indeed, the future of biodetection requires us to begin developing novel analytics that can identify anomalies and/or characteristics that indicate a potential threat, whether known or unknown, without looking for a specific signature that has been identified previously. To assess potential threats more rapidly, it is critical to develop agnostic artificial intelligence (AI)/machine learning (ML) systems that can be employed for real-time assessment of the nature and source of a perturbation. Such systems should be multi scale and multi-dimensional, integrating sensor data from a range of biological, chemical, and physical application spaces. Emerging deep learning (DL) models demonstrate exceptional promise for identification of discriminatory features within multi-dimensional datasets. DL models have the capacity to recognize and encode highly complex patterns in a wide range of input data modalities, including images, text, and biological/chemical/physical spectra. As such, they can execute a wide range of assessments and determinations that have traditionally required a human operator. The promise of advances in DL is apparent in the realm of human health and medicine. DL models have been validated for evaluating a variety of clinical threats to human health in a range of contexts, including infection and cancer, and they demonstrated improved performance in predicting stroke relative to human neurologists in some categories of data. Continuously evolving advances in AI/ML are expected to support more efficient evaluation of raw sequence, spectroscopy, and spectrometry data. For instance, recent advances and deployment of large language models (LLM) such as Generative Pre training Transformer (GPT) and Bidirectional Encoder Representations from Transformers (BERT) have already motivated application of these models for biological function prediction. As frameworks such as LLMs become larger and more complex in their representations, their capacity to serve as pre-trained models that can be fine-tuned for biological/biodetection purposes will similarly be amplified. While existing and emerging AI/ML have found broad applicability and use cases in the clinical sciences, development for environmental evaluation and biodetection has been limited. Functionalizing such capabilities for this purpose requires an understanding of the existing technical landscape and how the respective tools and algorithms are currently being employed. This landscape awareness then allows an assessment of the current practical capabilities of existing models and the anticipated requirements and development efforts that will be needed to adapt available algorithms for biodetection applications relevant to DHS. Leveraging expertise in biodetection, ML, and operational biodetection, the effort described in this report is comprised of a systematic landscape assessment (Subtask 2.1), comparative evaluation (Subtask 2.2), and formulation of a value proposition (Subtask 2.3) for the prospect of ML-enabled, agnostic biodetection from raw, or minimally-processed, datasets.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2024

Non-conformity Scores for High-Quality Uncertainty Quantification from Conformal Prediction

High-quality uncertainty quantification (UQ) is a critical component of enabling trust in deep learning (DL) models and is especially important if DL models are to be deployed in high-consequence applications. Conformal prediction (CP) methods represent an emerging nonparametric approach for producing UQ that is easily interpretable and, under weak assumptions, provides a guarantee regarding UQ quality. This report describes the research outputs of an Exploratory Express Laboratory Directed Research and Development (LDRD) project at Sandia National Laboratories. This project focused on how best to implement CP methods for DL models. This report introduces new methodology for obtaining high-quality UQ from DL models using CP methods, describes a novel system of assessing UQ quality, and provides experimental results that demonstrate the quality of the new methodology and utility of the UQ quality assessment system. Avenues for future research and discussion of potential impacts at Sandia and in the wider research community are also given.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Deep Learning-based Surrogate Model for Efficient Reservoir Simulation in Large-scale Geological Carbon Storage: Application in IBDP Dataset

This project introduces an advanced deep learning (DL)-based surrogate modeling approach to enhance the efficiency and accuracy of large-scale geological carbon storage (GCS) simulations. Using the Illinois Basin Decatur Project (IBDP) dataset as training data, the study employs a residual U-Net architecture to predict critical state variables such as pressure and CO₂ saturation, as well as CO₂ plume migration. By incorporating key geological parameters (e.g., porosity, permeability, and rock facies) and physics-informed inputs like the diffusive time of flight and time step, the DL model effectively reduces computational complexity while maintaining robust physical constraints. Compared to traditional simulators like Eclipse, the DL model achieves remarkable accuracy, with a root mean square error (RMSE) of 1.57 psi for pressure and 0.007 for saturation, and dramatically reduces computational time from hours to just 69.9 seconds for 50-step simulations. These results demonstrate the potential of innovative DL methodologies to improve the predictivity and operational efficiency of GCS simulations, providing a reliable foundation for decision-making in CCS operations. Supported by the SMART initiative, this project underscores the success of leveraging computational innovations to advance CCS technologies.

advanced deep learning↗

DOE OSTI2025

You Only Look Once v5 and Multi-Template Matching for Small-Crack Defect Detection on Metal Surfaces

This paper compares the performance of Deep Learning (DL) and multi-template matching (MTM) models for detecting small defects. DL models extract distinguishing features of objects but require a large dataset of images. In contrast, alternative computer vision techniques like MTM need a relatively small dataset. The lack of large datasets for small metal-surface defects has inhibited the adoption of automation in small-defect detection in remanufacturing settings. This motivated this preliminary study to compare template-based approaches, like MTM, with feature-based approaches, such as DL models, for small-defect detection on an initial laboratory and remanufacturing industry dataset. This study used You Only Look Once v5 (YOLOv5) as the DL model and compared its performance against the MTM model for small-crack detection. The findings of our preliminary investigation are as follows: (i) YOLOv5 demonstrated higher performance than MTM in detecting small cracks; (ii) an extra-large variant of YOLOv5 outperformed a small-size variant; (iii) the size and object variety of the data are crucial in achieving robust pre-trained weights for use in transfer learning; and (iv) enhanced image resolution contributes to precise object detection.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2024

Yet Another Discriminant Analysis (YADA): A Probabilistic Model for Machine Learning Applications

This paper presents a probabilistic model for various machine learning (ML) applications. While deep learning (DL) has produced state-of-the-art results in many domains, DL models are complex and over-parameterized, which leads to high uncertainty about what the model has learned, as well as its decision process. Further, DL models are not probabilistic, making reasoning about their output challenging. In contrast, the proposed model, referred to as Yet Another Discriminate Analysis(YADA), is less complex than other methods, is based on a mathematically rigorous foundation, and can be utilized for a wide variety of ML tasks including classification, explainability, and uncertainty quantification. YADA is thus competitive in most cases with many state-of-the-art DL models. Ideally, a probabilistic model would represent the full joint probability distribution of its features, but doing so is often computationally expensive and intractable. Hence, many probabilistic models assume that the features are either normally distributed, mutually independent, or both, which can severely limit their performance. YADA is an intermediate model that (1) captures the marginal distributions of each variable and the pairwise correlations between variables and (2) explicitly maps features to the space of multivariate Gaussian variables. Numerous mathematical properties of the YADA model can be derived, thereby improving the theoretic underpinnings of ML. Validation of the model can be statistically verified on new or held-out data using native properties of YADA. However, there are some engineering and practical challenges that we enumerate to make YADA more useful.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1978

Experimental observations of strong double layers

A computer simulation is applied to the production of strong electric potential double layers (DL) in a triple plasma device. The simulation is intended to represent DL in the low magnetosphere above the auroral zones. The DL are described as standing electrostatic shocks with different energy coefficients in their strong and weak forms. The strong DL was generally found to be unstable, but stability could be imparted if a population of trapped electrons was presented. Stability increased with the length of the system. A schematic for the system is presented, and a phase-space plot of electrons (indicating system stability) is graphed.

Coakley, P.↗

NASA NTRS1986

Creep-rupture and fractographic analysis of Stirling engine superalloys tested in air and 15 MPa hydrogen

A brief comparative analytical and microstructural evaluation of creep-rupture performance of two iron-base superalloys in air and 15 MPa of hydrogen, is presented. Creep rupture data are presented for the sheet alloy 19-9DL and the cast alloy XF-818, including temperature, initial stress, rupture life, minimum creep rate, time to reach one percent creep strain, and total elongation. In 19-9DL, both rupture life and minimum creep rate are more sharply dependent on small stress changes than in XF-818 in the given environment, and 19-9DL appears to become a more creep-resistant material with increasing Q (apparent activation energy) while the opposite is noted for XF-818. There appears to be no environmental effect on minimum creep rate for 19-9DL, whereas Q becomes less negative for XF-818 for 15 MPa of H2. Multiple cracks leading to rupture are observed on the fracture surfaces, with sheet specimens showing many more cracks close to the fracture surface than cast specimens.

Bhattacharyya, S.↗

NASA NTRS1993

Atmospheric measurements of carbonyl sulfide, dimethyl sulfide, and carbon disulfide using the electron capture sulfur detector

Measurements of atmospheric dimethyl sulfide (DMS), carbonyl sulfide (COS), and carbon disulfide (CS2) were conducted over the Atlantic Ocean on board the NASA Electra aircraft during the Chemical Instrumentation Test and Evaluation (CITE 3) project using the electron capture sulfur detector (ECD-S). The system employed cryogenic preconcentration of air samples, gas chromatographic separation, catalytic fluorination, and electron capture detection. Samples collected for DMS analysis were scrubbed of oxidants with NaOH impregnated glass fiber filters to preconcentration. The detection limits (DL) of the system for COS, DMS, and CS2 were 5, 5, and 2 ppt, respectively. COS concentrations ranged from 404 to 603 ppt with a mean of 489 ppt for measurements over the North Atlantic Ocean (31 deg N to 41 deg N), and from 395 to 437 ppt with a mean of 419 ppt for measurements over the Tropical Atlantic Ocean (11 deg S to 2 deg N). DMS concentrations in the lower marine boundary layer, below 600-m altitude, ranged from below DL to 150 ppt from flights over the North Atlantic, and from 9 to 104 ppt over the Tropical Atlantic. CS2 concentrations ranged from below DL to 29 ppt over the North Atlantic. Almost all CS2 measurements over the Tropical Atlantic were below DL.

Johnson, James E.↗

NASA NTRS1997

Building Multi-Discipline, Multi-Format Digital Libraries Using Clusters and Buckets

Our objective was to study the feasibility of extending the Dienst protocol to enable a multi-discipline, multi-format digital library. We implemented two new technologies: cluster functionality and publishing buckets. We have designed a possible implementation of clusters and buckets, and have prototyped some aspects of the resultant digital library. Currently, digital libraries are segregated by the disciplines they serve (computer science, aeronautics, etc.), and by the format of their holdings (reports, software, datasets, etc.). NCSTRL+ is a multi-discipline, multi-format digital library (DL) prototype created to explore the feasibility of the design and implementation issues involved with created a unified, canonical scientific and technical information (STI) DL. NCSTRL+ is based on the Networked Computer Science Technical Report Library (NCSTRL), a World Wide Web (WWW) accessible DL that provides access to over 80 university departments and laboratories. We have extended the Dienst protocol (version 4.1.8), the protocol underlying NCSTRL, to provide the ability to cluster independent collections into a logically centralized DL based upon subject category classification, type of organization, and genre of material. The concept of buckets provides a mechanism for publishing and managing logically linked entities with multiple data formats.

Nelson, Michael L.↗

NASA NTRSDate not supplied

Robust and Mass Efficient Thermal Protection Systems for Future Venus Missions

Several international Venus missions, both orbiter, and in-situ probe are in the works after a prolonged absence. Future missions are expected to be long-duration lander missions and balloon missions to investigate the cloud layer. The entry mass for these in-situ missions will be significantly higher, and the aeroshell will be larger, than previous missions. Two cardinal requirements govern the selection and use of thermal protection systems (TPS), namely, robustness to assure mission safety during entry and mass efficiency so that the useful mass for science is maximized. One cannot trade mission safety for mass when it comes to TPS. The robustness of the TPS is a paramount requirement as it is a single point of system failure. At the same time, TPS mass is carried for the entire mission duration prior to entry, and any excess TPS mass is at the cost of science payload. Future missions will benefit enormously from TPS options with mass and performance benefits far beyond advanced carbon-carbon (ACC) that is currently proposed for the DAVINCI mission. NASA STMD, and SMD jointly invested in the development of 3-D woven thermal protection systems in the last decade, and as a result, the Heat-shield for Extreme Entry Environment Technology (HEEET) TPS has been matured to TRL 6. It is ready for future mission use including Venus. The HEEET project focused on developing a broad technology base, applicable for missions not only to Venus, but also to Saturn, the Ice Giants, and higher speed sample return missions such as Mars Sample Return requiring extreme robustness due to bio-hazards. The dual-layer HEEET (DL-HEEET) is proven to be extremely robust. DL-HEEET performed well in arc jet and laser testing conditions where heritage carbon-phenolic failed. HEEET also proved to be more mass efficient compared to Carbon-Phenolic. Recently, the principal author was invited to present his perspectives to the 2023-2032 Planetary Sciences Decadal Venus subcommittee on the current state of TPS for future Venus missions [1]. In addition to pointing out successful TPS and other developments in the last decade in support of entry systems, the principal author made two findings: 1) importance of and need to sustain TPS capabilities that took nearly a decade to develop and 2) the opportunity to further optimize TPS mass without sacrificing robustness to further enable future in-situ missions. This talk will focus on the above two recommendations and provide the rationale for them. The intent of the talk is to seek advocacy from the VEXAG community for the criticality of TPS sustainment and the benefits of 3D woven TPS optimization. TPS Capability Sustainment: HEEET development was necessitated by the atrophy of heritage carbon-phenolic. Atrophy has impacted other TPS materials as well. Avcoat, the TPS that successfully allowed astronauts to explore the Moon in the 1960s and 70s, took nearly a decade and $10’sM to recover in support of Orion/Artemis. One of the most used TPS materials, SLA 561V, had to be recovered prior to its use on Mars Pathfinder. Phenolic Impregnated Carbon Ablator (PICA), the TPS that replaced SLA 561V on MSL, has undergone multiple replacements and recovery due to constituent rayon changes. Recently, NASA invested in a domestic rayon replacement program and also invested in FMI to consolidate PICA capability for NASA missions, as a result of FMI’s decision to discontinue commercial FiberForm, which is needed for PICA. TPS used for planetary missions are unique and have no other commercial or defense use. In addition, low mission cadence is also a driver for TPS atrophy. Hence, NASA, as the steward, must take steps to ensure and sustain TPS capabilities. In this talk, we outline steps NASA can take to keep abreast of emerging risks and target risk mitigation steps to ensure TPS capability sustainment for Venus and other extreme environment missions. Next Generation of Mass efficient and Robust TPS: NASA invested in an alternate TPS to PICA based on felt-technology called Conformal-PICA which has the potential to save 30% - 50% mass over PICA. The development was discontinued at a Technology Readiness Level (TRL) of ~ 5. 3D Mid-Density Carbon Phenolic (3MDCP) is a single layer variant of HEEET, based on the insulating layer only (SL-HEEET) and currently baselined for the Mars Sample Return Mission Earth Entry System due to its mass efficiency (30% more mass efficient than DL- HEEET). Currently, SL-HEEET is limited to aeroshell diameters of < 1.3m. The SL-HEEET was compared to DL-HEEET in the recent ADVENT flag-ship class mission study in support of the Planetary Science Decadal. SL HEEET was the recommended TPS based on 30% mass savings for both the balloon and lander missions. Given C-PICA and SL-HEEET have superior reliability and mass efficiency, advocacy from VEXAG is sought for completing their further development to TRL 5/6 in this decade so as to enable TPS readiness for future missions.

Thermal Protection Systems↗

NASA NTRSDate not supplied

The Application of Artificial Intelligence Deep Learning to Visually Identify Micrometeoroid and Orbital Debris Impacts

Recent advances in Artificial Intelligence (AI) are changing the World. Novel approaches to training AI systems have led to dramatic reductions in the amount of time required. Training an AI system could take years and teams of people using traditional methods, but with the advancements of Deep Learning (DL) models this training can now be accomplished by an individual in a matter of minutes. The development of “fast AI” libraries has delivered AI to essentially everyone. Democratization of AI power has inspired many to revisit past problems that will benefit from DL approaches. For example, the application of AI has improved detection of breast cancer by 20% compared to traditional detection methods. Computer vision and machine learning are being used to identify soil deficiencies and provide planting recommendations to farmers. Success stories like these and many others have provided inspiration to see if AI can help improve one of our needed capabilities – that of visually identifying micrometeoroid and orbital debris (MMOD) impact damage to spacecraft from images of the spacecraft exterior. The need to visually locate and characterize spacecraft MMOD impact damage has been present since the early days of space travel. This is often done by either having a crew member take photographs of the spacecraft through a window using a hand-held camera or ground personnel directing externally-mounted cameras. The photographs are then transmitted back to Earth for visual analysis. This method of MMOD damage inspection works well and has been used on various spacecraft including the Space Shuttle and the International Space Station (ISS). One of the issues with the current method that we believe AI could improve is the speed and possibly the accuracy in identifying MMOD impacts. Note that detecting MMOD impacts in images can be very difficult. The visual appearance of an MMOD impact can change dramatically with lighting conditions, size of impact, depth of penetration, material types, surface waviness, fabric coverings, camera & lens, distance to surface, spacecraft orientation, analyst experience, and many other factors. Currently, this takes a team of highly-experienced specialists in both the fields of Image Analysis and MMOD impacts. This paper documents our initial research in training an AI DL model using the fast-AI library to identify actual and simulated MMOD impacts and perforations into exposed flat surfaces. While we recognize that this initial goal seems modest, it must be noted that what we have done would have taken teams of individuals and years of training just ten years ago. Our long-term goal is to add complexity and use-cases to the DL model being trained to expand the capabilities of this model so that it can be used to identify MMOD impacts on all types of spacecraft surfaces.

Cameron M Collins↗

NASA NTRSDate not supplied

Advancement of Deep Learning and Geometric Methods for Active Terrain Relative Navigation

To enhance NASA’s precision landing capabilities, in conjunction with the development of a novel active terrain relative navigation (ATRN) and terrain mapping system, denoted SHERIF, this work performed a comparative analysis between both deep-learning (DL) based and geometric approaches to hazard detection (HD) and safe-site-identification (SSI) through hardware-in-the loop testing on the Six degree-of-freedom Tendon Actuated Robot (STAR). The Standalone Hazard Evaluation and Refinement using Instrument Findings (SHERIF) system is capable of ingesting sensor data at an asynchronous rate, stitching successive terrain scans together to yield a high-resolution digital elevation map (DEM), performing absolute and relative localization using novel 3D feature extraction and matching methods, and HD/SSI activities. The DL-based HD/SSI algorithm provides a modular alternative to classical geometric approaches which have performance times that scale with map resolution. As the adoption of AI solutions become more prevalent for autonomous system decision making, it is prudent to explore the utility of such solutions in applications where they traditionally excel, such as image classification. Along with the development of a DL-based HD system, this work performed the first comparative analysis between DL and geometric approaches to HD/SSI using real sensor data from real-time testing in a relevant environment.

Davis Adams↗

DOE OSTI2025

Combining Deep Learning and scatterControl for High-Throughput X-ray CT Based Non-Destructive Characterization of Large-Scale Casted Metallic Components

X-ray computed tomography (XCT) is essential for nondestructive evaluation and quality control of large-scale metal components. XCT imaging, however, faces significant challenges from metal artifacts, particularly those caused by Compton scattering, which degrade image quality and obscure critical details. Hardware-based solutions (e.g. scatterControl) offer advancements by intercepting scattered photons and reducing artifacts, but they can be time-consuming and require additional processing. Here, we propose modifying and leveraging a novel deep learning (DL) framework, Simurgh, to enhance and accelerate scatter correction in XCT. By combining scatterControl with DL-based artifact removal, we demonstrate significant reduction in scan time while producing high-quality reconstructions. Through extensive evaluation on industrial XCT data, we show that our methods reduce scan time by up to more than 10 x while preserving flaw detectability. Quantitative analysis across multiple segmentation techniques confirms that Simurgh-based reconstructions consistently outperform traditional Feldkamp-Davis-Kress, model-based iterative reconstruction, and commercial DL models in both pixel-level and task-specific evaluations, enabling scalable, high-throughput XCT workflows for characterization of large scale components in applications such as casting and metal additive manufacturing.

Complex metal parts↗