SEARCH · Engineering Papers
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.
EEG baselines covering wide range of states of wakefulness and sleep in astronaut candidates estimated by computation and pattern recognition techniques
Input/output model of system with finite settling time, using error correcting technique employed in pattern recognition
VIGILANTE is an ultrafast smart sensor testbed for generic Automatic Target Recognition (ATR) applications with a series of capability demonstration focussed on cruise missile defense (CMD). VIGILANTE's sensor/processor architecture is based on next-generation UV/visible/IR sensors and a tera-operations per second sugar-cube processor, as well as supporting airborne vehicle. Excellent results of efficient ATR methodologies that use an eigenvectors/neural network combination and feature-based precision tracking have been demonstrated in the laboratory environment.
Fingerprint recognition is typically done by matching graphs or templates of extracted minutia from whole fingerprint images.
A bottom-up visual attention model (the saliency model) is tested to enhance the performance of Automated Target Recognition (ATR). JPL has developed an ATR system that identifies regions of interest (ROI) using a trained OT-MACH filter, and then classifies potential targets as true- or false-positives using machine-learning techniques. In this project, saliency is used as a pre-processing step to reduce the space for performing OT-MACH filtering. Saliency parameters, such as output level and orientation weight, are tuned to detect known target features. Preliminary results are promising and future work entails a rigrous and parameter-based search to gain maximum insight about this method.
Several feature extraction and selection methods for an existing automatic target recognition (ATR) system using JPLs Grayscale Optical Correlator (GOC) and Optimal Trade-Off Maximum Average Correlation Height (OT-MACH) filter were tested using MATLAB. The ATR system is composed of three stages: a cursory region of-interest (ROI) search using the GOC and OT-MACH filter, a feature extraction and selection stage, and a final classification stage. Feature extraction and selection concerns transforming potential target data into more useful forms as well as selecting important subsets of that data which may aide in detection and classification. The strategies tested were built around two popular extraction methods: Principal Component Analysis (PCA) and Independent Component Analysis (ICA). Performance was measured based on the classification accuracy and free-response receiver operating characteristic (FROC) output of a support vector machine(SVM) and a neural net (NN) classifier.
The task for this project was to design, develop, test, and deploy a facial recognition system for the Kennedy Space Center Augmented/Virtual Reality Lab. This system will serve as a means of user authentication as part of the NUI of the lab. The overarching goal is to create a seamless user interface that will allow the user to initiate and interact with AR and VR experiences without ever needing to use a mouse or keyboard at any step in the process.
Increasingly, unmanned aircraft systems (UAS) are being applied to wildfire incidents for tasks such as mapping, aerial ignition, and delivery. As a result, incident reporting systems for wildfires are beginning to accumulate data related to UAS mishaps in wildfire response. In this research, we apply state-of-the-art natural language processing (NLP) techniques to develop a custom Named Entity Recognition (NER) model which extracts a Failure Modes and Effects Analysis (FMEA)-style survey of wildfire UAS mishaps reported in SAFECOM. The custom NER model is built by fine-tuning an existing (BERT) model, resulting in a generalizable NER model that can extract engineering relevant entities including failure modes, causes, effects, control processes, and recommendations from any failure-relevant text. Similar mishaps are clustered and reported as single rows within the FMEA. For each cluster, frequency, severity, and overall risk are computed. The methodology can be applied as part of a broader safety management system to track trends in mishaps and discover knowledge that can be utilized to improve safety outcomes and system performance.
Increasingly, unmanned aircraft systems (UAS) are being applied to wildfire incidents for tasks such as mapping, aerial ignition, and delivery. As a result, aviation incident reporting systems for wildfires are beginning to accumulate data related to UAS mishaps in wildfire response. In this research, we apply state-of-the-art natural language processing (NLP) techniques to develop a custom Named Entity Recognition (NER) model which extracts entities relevant to safety analysts. The custom NER model is built by fine-tuning an existing Bidirectional Encoder Representations from Transformers (BERT) model, resulting in a generalizable NER model that can extract engineering relevant entities including failure modes, causes, effects, control processes, and recommendations from failure-relevant text. This model performs passably, with a weighted average f1 score of 0.33 across entity types, indicating more labeled training data is needed. Extracted entities are used to form a Failure Modes and Effects Analysis (FMEA)-style survey of wildfire UAS mishaps reported using the SAFECOM system. Similar mishaps are manually clustered and reported as single rows within an FMEA. Foreach cluster, we compute frequency, severity, and overall riskin accordance with FAA standards. This methodology can beapplied as part of a broader safety management system totrack trends in mishaps (e.g., likelihood, severity) and discoverknowledge (e.g., causes, effects) that can be utilized to improvesafety outcomes and system performance.
Increasingly, unmanned aircraft systems (UAS) are being applied to wildfire incidents for tasks such as mapping, aerial ignition, and delivery. As a result, incident reporting systems for wildfires are beginning to accumulate data related to UAS mishaps in wildfire response. In this research, we apply state-of-the-art natural language processing (NLP) techniques to develop a custom Named Entity Recognition (NER) model which extracts a Failure Modes and Effects Analysis (FMEA)-style survey of wildfire UAS mishaps reported in SAFECOM. The custom NER model is built by fine-tuning an existing (BERT) model, resulting in a generalizable NER model that can extract engineering relevant entities including failure modes, causes, effects, control processes, and recommendations from any failure-relevant text. Similar mishaps are clustered and reported as single rows within the FMEA. For each cluster, frequency, severity, and overall risk are computed. The methodology can be applied as part of a broader safety management system to track trends in mishaps and discover knowledge that can be utilized to improve safety outcomes and system performance.
Mindin (spondin-2) is an extracellular matrix protein of unknown structure that is required for efficient T-cell priming by dendritic cells. Additionally, mindin functions as a pattern recognition molecule for initiating innate immune responses. These dual functions are mediated by interactions with integrins and microbial pathogens, respectively. Mindin comprises an N-terminal F-spondin (FS) domain and C-terminal thrombospondin type 1 repeat (TSR). We determined the structure of the FS domain at 1.8-A resolution. The structure revealed an eight-stranded antiparallel beta-sandwich motif resembling that of membrane-targeting C2 domains, including a bound calcium ion. We demonstrated that the FS domain mediates integrin binding and identified the binding site by mutagenesis. The mindin FS domain therefore represents a new integrin ligand. We further showed that mindin recognizes lipopolysaccharide (LPS) through its TSR domain, and obtained evidence that C-mannosylation of the TSR influences LPS binding. Through these dual interactions, the FS and TSR domains of mindin promote activation of both adaptive and innate immune responses.
We developed multiple machine learning methods for the detection and classification of new wireless communication waveforms, which is critical for targeted attacks in wireless networks and electronic warfare. Our machine learning models are capable of dynamically detecting security threats in near real time through our advanced open set recognition (OSR) approach. This model has demonstrated significant improvements in the detection of unknown waveforms, thereby enhancing the security and reliability of mission critical communications. Our approach to detecting uncertain security threats is novel; we advanced OSR techniques by incorporating domain knowledge of wireless signals. Specifically, we combined time and frequency domain model features to enhance the model’s performance. Utilizing an OSR approach eliminates the need for training data to be distributed similarly to the deployment environment and removes the requirement for the training set to contains all possible threat classes. This is crucial because it is often infeasible to determine and characterize all potential security threats in advance. Our model were trained on simulated data, generated in partnership with the University at Albany, State of New York. The data set contained a diverse array of wireless signals, including those with additive white Gaussian noise and multipath signals, with and without line of sight. This comprehensive training set allowed us to optimize our models to detect unknown waveforms under various challenging scenarios, such as low signal-to-noise ratios. By training on various waveforms, varying signal-to-noise ratio, and different sample sizes under normal conditions, our models were fine tuned to perform effectively in challenging environments.
This presentation applies open set recognition to classify unknown waveforms, enabling systems to not only identify known types but also reliably detect when waveforms fall outside the training distribution. This approach enhances robustness by avoiding forced misclassification of novel or anomalous signals.
The role of the conserved residue Tyr105 in class A β‐lactamases has been the subject of investigation using both structural studies and saturation mutagenesis. Both have shown that while it does not need to be strictly conserved for activity, it is important for substrate recognition. With this in mind we determined the crystal structure of Toho1 β‐lactamase at 15 K to 1.10 Å resolution in complex with penicillin. As expected a ring‐opened penicillin molecule bound to Ser70 the catalytic nucleophile, can clearly be seen in electron density in the active site. In addition to the trapped penicillin, however, are two additional intact ring‐closed penicillin molecules, captured by the enzyme through noncovalent interactions at the edge of the active site.
Abstract The heterogeneous nuclear ribonucleoprotein (hnRNP) R‐like family is a class of RNA binding proteins in the hnRNP superfamily with diverse functions in RNA processing. Here, we present the 1.90 Å X‐ray crystal structure and solution NMR studies of the first RNA recognition motif (RRM) of human hnRNPR. We find that this domain adopts an extended RRM (eRRM1) featuring a canonical RRM with a structured N‐terminal extension (N ext ) motif that docks against the RRM and extends the β‐sheet surface. The adjoining loop is structured and forms a tryptophan cage motif to position the N ext motif for docking to the RRM. Combining mutagenesis, solution NMR spectroscopy, and thermal denaturation studies, we evaluate the importance of residues in the N ext –RRM interface and adjoining loop on eRRM folding and conformational dynamics. We find that these sites are essential for protein solubility, conformational ordering, and thermal stability. Consistent with their importance, mutations in the N ext –RRM interface and loop are associated with several cancers in a survey of somatic mutations in cancer studies. Sequence and structure comparison of the human hnRNPR eRRM1 to experimentally verified and predicted hnRNPR‐like proteins reveals conserved features in the eRRM.
Abstract Musashi‐2 (MSI2) belongs to Musashi family of RNA binding proteins (RBP). Like Musashi‐1 (MSI1), it is overexpressed in a variety of cancers and is a promising therapeutic target. Both MSI proteins contain two N‐terminal RNA recognition motifs and play roles in posttranscriptional regulation of target mRNAs. Previously, we have identified several inhibitors of MSI1, all of which bind to MSI2 as well. In order to design MSI2‐specific inhibitors and compare the differences of binding mode of the inhibitors, we set out to solve the structure of MSI2‐RRM1, the key motif that is responsible for the binding. Here, we report the crystal structure and the first NMR solution structure of MSI2‐RRM1, and compare these to the structures of MSI1‐RBD1 and other RBPs. A high degree of structural similarity was observed between the crystal and solution NMR structures. MSI2‐RRM1 shows a highly similar overall folding topology to MSI1‐RBD1 and other RBPs. The structural information of MSI2‐RRM1 will be helpful for understanding MSI2‐RNA interaction and for guiding rational drug design of MSI2‐specific inhibitors.
Abstract While there is a growing body of work that examines disciplinary identity development, unlike qualitative work in this area, quantitative research has not fully incorporated the importance of different contexts, nor has it uniquely focused on underrepresented groups (in this case, women in physics). This study examines how the constructs posited by prior work as important for physics identity, as well as an additional theorized construct, may interrelate and affect female students' physics identity differently depending on the context. Context in this study refers to two different experiential levels in college. The constructs examined include performance/competence, recognition, and interest, as well as sense of belonging. In particular, we used structural equation modeling to examine the effect that these constructs have on the physics identity of two groups of female physics undergraduates: first year students and senior year students. The results reveal that the relationship of the theorized constructs with physics identity vary between the two groups as well as compared to prior research with broad college student populations (not just physics majors). Unlike broad college student populations, for our sample of female physics undergraduates, interest did not have a direct effect on physics identity while sense of belonging was significant only for senior year students. These results exemplify the importance of examining context or different types of student experiences when studying disciplinary identity development rather than generalizing previous frameworks to all contexts.
Scientific Named Entity Referent Extraction is often more complicated than traditional Named Entity Recognition (NER). For example, in polymer science, chemical structure may be encoded in a variety of nonstandard naming conventions, and authors may refer to polymers with conventional names, commonly used names, labels (in lieu of longer names), synonyms, and acronyms. As a result, accurate scientific NER methods are often based on task-specific rules, which are difficult to develop and maintain, and are not easily generalized to other tasks and fields. Machine learning models require substantial expert-annotated data for training. Here we propose polyNER: a semi-automated system for efficient identification of scientific entities in text. PolyNER applies word embedding models to generate entity-rich corpora for productive expert labeling, and then uses the resulting labeled data to bootstrap a context-based word vector classifier. Evaluation on materials science publications shows that the polyNER approach enables improved precision or recall relative to a state-of-the-art chemical entity extraction system at a dramatically lower cost: it required just two hours of expert time, rather than extensive and expensive rule engineering, to achieve that result. This result highlights the potential for human-computer partnership for constructing domain-specific scientific NER systems.