Search for: All records

Additive manufacturing enables the fabrication of complex designs while minimizing waste, but faces challenges related to defects and process anomalies. This study presents a novel multimodal Retrieval-Augmented Generation-based framework that automates anomaly detection across various Additive Manufacturing processes leveraging retrieved information from literature, including images and descriptive text, rather than training datasets. This framework integrates text and image retrieval from scientific literature and multimodal generation models to perform zero-shot anomaly identification, classification, and explanation generation in a Laser Powder Bed Fusion setting. The proposed framework is evaluated on four L-PBF manufacturing datasets from Oak Ridge National Laboratory, featuring various printer makes, models, and materials. This evaluation demonstrates the framework's adaptability and generalizability across diverse images without requiring additional training. Comparative analysis using Qwen2-VL-2B and GPT-4o-mini as MLLM within the proposed framework highlights that GPT-4o-mini outperforms Qwen2-VL-2B and proportional random baseline in manufacturing anomalies classification. Additionally, the evaluation of the RAG system confirms that incorporating retrieval mechanisms improves average accuracy by 12% by reducing the risk of hallucination and providing additional information. The proposed framework can be continuously updated by integrating emerging research, allowing seamless adaptation to the evolving landscape of AM technologies. This scalable, automated, and zero-shot-capable framework streamlines AM anomaly analysis, enhancing efficiency and accuracy. 

Safety-critical data, such as crash and near-crash records, are crucial to improving autonomous vehicle (AV) design and development. Sharing such data across AV companies, academic researchers, regulators, and the public can help make all AVs safer. However, AV companies rarely share safety-critical data externally. This paper aims to pinpoint why AV companies are reluctant to share safety-critical data, with an eye on how these barriers can inform new approaches to promote sharing. We interviewed twelve AV company employees who actively work with such data in their day-to-day work. Findings suggest two key, previously unknown barriers to data sharing: (1) Datasets inherently embed salient knowledge that is key to improving AV safety and are resource-intensive. Therefore, data sharing, even within a company, is fraught with politics. (2) Interviewees believed AV safety knowledge is private knowledge that brings competitive edges to their companies, rather than public knowledge for social good. We discuss the implications of these findings for incentivizing and enabling safety-critical AV data sharing, specifically, implications for new approaches to (1) debating and stratifying public and private AV safety knowledge, (2) innovating data tools and data sharing pipelines that enable easier sharing of public AV safety dataand knowledge; (3) offsetting costs of curating safety-critical data and incentivizing data sharing. 

Abstract We study the black hole mass–host galaxy stellar mass relation,M_BH–M_*, for a sample of 706z ≲ 1.5 andi ≲ 24 optically variable active galactic nuclei (AGNs) in three Dark Energy Survey (DES) Deep Fields: C3, X3, E2, which partially cover Chandra Deep Field-South, XMM Large Scale Structure survey, and European Large Area ISO Survey, respectively. The parent sample was identified by optical variability from the DES supernova survey program imaging. Using publicly available spectra and photometric catalogs, we consolidate their spectroscopic redshifts, estimate their black hole masses using broad line widths and luminosities, and obtain improved stellar masses using spectral energy distribution fitting from X-ray to mid-infrared wavelengths. Our results confirm previous work from Hyper-Suprime Camera imaging that variability searches with deep, high-precision photometry can reliably identify AGNs in low-mass galaxies up toz ∼ 1. However, we find that the hosted black holes are more massive than predicted by the local AGN relation, fixing host galaxy stellar mass. Instead,z ∼ 0.1–1.5 variability-selected AGNs lie in between theM_BH–M_*relation for local inactive early-type galaxies and local active galaxies. This result agrees with most previous studies of theM_BH–M_*relation for AGNs at similar redshifts, regardless of the selection technique. We demonstrate that studies of variability-selected AGN provide critical insights into the low-mass end of theM_BH–M_*relation, shedding light on the occupation fraction of that provides constraints on early black hole seeding mechanisms and self-regulated feedback processes during their growth and coevolution with their hosts. 

High-throughput screening enabled by structure-property prediction models is a powerful approach for accelerating materials discovery. However, while machine learning of structure-property models have become widespread, its application to mixtures remains limited due to increased complexity and the scarcity of available data. Machine learning methods for high-throughput screening of eutectic mixtures have been proposed in recent years, but there remain challenges due to the lack of diverse, open-access datasets and the need for feature engineering based on chemical knowledge. To overcome these limitations, we propose a method using Siamese graph neural networks trained solely on structural information, without requiring any prior chemical descriptors, to predict eutectic melting temperatures. We demonstrate on a dataset of molten salt eutectics that this approach can reach similar performance to chemistry-based models that require significantly more prior knowledge. We show that lower-order mixtures may be used to augment data on higher-order mixtures. Interestingly, our model trained on inorganic molten salts seems to learn information about the ideal mixture model. We also evaluate the efficacy of using our inorganic molten salt model for transfer learning with a variety of organic eutectic mixtures. 

Abstract Additive manufacturing enables the fabrication of complex designs while minimizing waste, but faces challenges related to defects and process anomalies. This study presents a novel multimodal Retrieval-Augmented Generation-based framework that automates anomaly detection across various Additive Manufacturing processes leveraging retrieved information from literature, including images and descriptive text, rather than training datasets. This framework integrates text and image retrieval from scientific literature and multimodal generation models to perform zero-shot anomaly identification, classification, and explanation generation in a Laser Powder Bed Fusion setting. The proposed framework is evaluated on four L-PBF manufacturing datasets from Oak Ridge National Laboratory, featuring various printer makes, models, and materials. This evaluation demonstrates the framework’s adaptability and generalizability across diverse images without requiring additional training. Comparative analysis using Qwen2-VL-2B and GPT-4o-mini as MLLM within the proposed framework highlights that GPT-4o-mini outperforms Qwen2-VL-2B and proportional random baseline in manufacturing anomalies classification. Additionally, the evaluation of the RAG system confirms that incorporating retrieval mechanisms improves average accuracy by 12% by reducing the risk of hallucination and providing additional information. The proposed framework can be continuously updated by integrating emerging research, allowing seamless adaptation to the evolving landscape of AM technologies. This scalable, automated, and zero-shot-capable framework streamlines AM anomaly analysis, enhancing efficiency and accuracy. 

MultiTaskDeltaNet (MTDN) reframes semantic segmentation as change detection, enabling data-efficient, automated operando ETEM analysis for spatially-resolved carbon gasification kinetics with superior performance on small, ambiguous features. 

We combine graph neural networks and an upper bound energy minimization framework to explore >90 000 candidates, discovering 1810 stable Zintl phases with high precision and revealing the key role of ionic bonding in their stability.