Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Adversarial examples”

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.

116 records · Page 7

A Data Processing Pipeline To Extract A Knowledge Graph From Heterogeneous Data For Socio-technical Analysis Of Critical Infrastructure Influence

The code is written in Python and consists of the following pipeline that is implemented in Apache Airflow. This pipeline intends to understand the companies that are directly or indirectly involved with a type of critical infrastructure system at some point in that system's lifecycle. The pipeline takes a configuration file that specifies a list of initial companies to consider, a geographic region of interest, and a set of SEC form types as well as other data sources (e.g. CrunchBase) from which to extract entities and relations. There are four main components to this pipeline as currently implemented: Entity Extraction, Network Construction, Analysis, and Visualization. First, Entity Extraction, is implemented as the `topear-extract_organizations` Apache Airflow workflow. Given an initial query that specifies a geographic region of interest and a time interval, the software will extract CI facilities of interest and organizations that have a direct influence relationship to those facilities (e.g. ownership). During the course of the LDRD, we focused on Electric Vehicle charging stations and this information is available via the Department of Energy (DOE) database on fueling stations maintained by NREL. Within the context of the DOE CESER project, we have focused on Battery Energy Storage Systems (BESS). Second, the Network Extraction component will iteratively construct a social network graph given the set of organizations and people extracted in the previous step. Organizations (and eventually People if desired) are then fed as a query to the `topgear-construct_social_network` Apache Airflow workflow which given a set of initial companies and data sets (e.g. SEC EDGAR form types, OpenCorporates, Crunchbase). This Airflow workflow will iteratively query such data sources to discover relationships with new organizations and people. For example, this module can iteratively query SEC EDGAR for metadata that documents the number of each type of form for the given set of companies and their location. This forms metadata represents a catalog of data sources from SEC EDGAR for the extracted social network knowledge graph. The pipeline then downloads these forms from the website and saves them in a build directory for further processing. These documents are then parsed for entities and relations. Again, we note that in additional to SEC data sources, this step can also pull in information on organizations via API services such as CrunchBase and OpenCorporates or bulk data sources. At the end of this step, the resultant social network, the Critical Infrastructure network, and the edges that encode relationships between organizations and CI facilities, form the Adversarial Socio-Technical Network (ASTN) that informs the analysis. Third, the Analysis component processes these generated ASTN. Previously, that has included the ability to compare prevalence of different vendors for a given infrastructure component type across different regions as well as identify common public and private investors across those vendors. This was demonstrated for EV Charging Stations across several different metropolitan areas within an IEEE PES GridEdge publication. More recently, we have looked at ways to identify infrastructure owners and operators of BESS with the most nameplate capacity across different states as well as other indictors of risk resulting from changes in ownership over time. Finally, the Visualization component consists of an HTML/CSS/JS framework by which users can interact geospatial, operational, and organizational relationships across a given portfolio of Critical Infrastructure facilities. The objective is to provide a library of UI/UX modules that can be repurposed for stakeholder-specific dashboards. All of the modules are related via a common event model that enables UI actions in one view to percolate across the other views.

Weaver, Gabriel [Idaho National Laboratory (INL), ↗

Resilient Distributed Parameter Estimation With Heterogeneous Data

This paper studies resilient distributed estimation under measurement attacks. A set of agents each makes successive local, linear, noisy measurements of an unknown vector field collected in a vector parameter. The local measurement models are heterogeneous across agents and may be locally unobservable for the unknown parameter. An adversary compromises some of the measurement streams and changes their values arbitrarily. The agents' goal is to cooperate over a peer-to-peer communication network to process their (possibly compromised) local measurements and estimate the value of the unknown vector parameter. We present SAGE, the Saturating Adaptive Gain Estimator, a distributed, recursive, consensus + innovations estimator that is resilient to measurement attacks. We demonstrate that, as long as the number of compromised measurement streams is below a particular bound, then, SAGE guarantees that all of the agents' local estimates converge almost surely to the value of the parameter. The resilience of the estimator - i.e., the number of compromised measurement streams it can tolerate - does not depend on the topology of the inter-agent communication network. Finally, we illustrate the performance of SAGE through numerical examples.

42 ENGINEERING↗

Wrap-up session

Listening to the case studies that were presented it became quite evident that the best data management systems were the ones where data managers and research scientists worked as a team developed in the early stages of project planning. Examples that were given included WOCE Data Assembly Centres e.g., Drifters, the Global Temperature Salinity Pilot Project (GTSPP) collaboration with Joint Analyses Centres in the U.S. and Australia, and JGOFS/BOFS development of Topical Centres. While each of these has some elements unique to the project, each had brought together 'teams' of Principal Investigators (PI's) and data management experts at an early stage of project development. Conversely, projects which had considered data management as a totally separate activity with lower priority often failed to provide the service required to meet scientific objectives. Therefore, the following actions should be brought to the attention of relevant groups within the IOC and other international organizations: (1) Publicize, at the national and international level, underway data/scientist collaborations that may be used as models in planning for the future. (2) Reduce adversarial situations where data managers and scientists appear to be in competition. (3) Colocation and other forms of collaboration often results in very high quality data sets and more timely data submission. Improved timeliness of data submissions was a common theme throughout the workshop and must be considered an important element in all future plans. Improvements in timely submission of data were noted. In order to continue this trend the advantages of timely submission of data must be stressed to those planning new ocean science projects.

Holland, Geoffrey↗

GPLADD: Quantifying Trust in Government and Commercial Systems A Game-Theoretic Approach

Trust in a microelectronics-based system can be characterized as the level of confidence that a system is free of subversive alterations made during system development, or that the development process of a system has not been manipulated by a malicious adversary. Trust in systems has become an increasing concern over the past decade. This report introduces a novel game-theoretic framework, called GPLADD (Graph-based Probabilistic Learning Attacker and Dynamic Defender), for analyzing and quantifying system trustworthiness at the end of the development process, through the analysis of risk of development-time system manipulation. GPLADD represents attacks and attacker-defender contests over time. Here, time is an explicit constraint and allows incorporating the informational asymmetries between the attacker and defender into analysis. GPLADD includes an explicit representation of attack steps via multi-step attack graphs, attacker and defender strategies, and player actions at different times. GPLADD allows quantifying the attack success probability over time and the attacker and defender costs based on their capabilities and strategies. This ability to quantify different attacks provides an input for evaluation of trust in the development process. We demonstrate GPLADD on an example attack and its variants. We develop a method for representing success probability for arbitrary attacks and derive an explicit analytic characterization of success probability for a specific attack. We present a numeric Monte Carlo study of a small set of attacks, quantify attack success probabilities, attacker and defender costs, and illustrate the options the defender has for limiting the attack success and improving trust in the development process.

37 attack↗

Effectiveness of denoising diffusion probabilistic models for fast and high-fidelity whole-event simulation in high-energy heavy-ion experiments

Artificial intelligence (AI) generative models, such as generative adversarial networks (GANs), variational autoencoders, and normalizing flows, have been widely used and studied as efficient alternatives for traditional scientific simulations. However, they have several drawbacks, including training instability and inability to cover the entire data distribution, especially for regions where data are rare. This is particularly challenging for whole-event, full-detector simulations in high-energy heavy-ion experiments, such as sPHENIX at the Relativistic Heavy Ion Collider and Large Hadron Collider experiments, where thousands of particles are produced per event and interact with the detector. This work investigates the effectiveness of denoising diffusion probabilistic models (DDPMs) as an AI-based generative surrogate model for the sPHENIX experiment that includes the heavy-ion event generation and response of the entire calorimeter stack. DDPM performance in sPHENIX simulation data is compared with a popular rival, GANs. Results show that both DDPMs and GANs can reproduce the data distribution where the examples are abundant (low-to-medium calorimeter energies). Nonetheless, DDPMs significantly outperform GANs, especially in high-energy regions where data are rare. Additionally, DDPMs exhibit superior stability compared to GANs. The results are consistent between both central and peripheral centrality heavy-ion collision events. Moreover, DDPMs offer a substantial speedup of approximately a factor of 100 compared to the traditional Geant4 simulation method.

42 ENGINEERING↗

Deep Generative Models for Materials Discovery and Machine Learning-Accelerated Innovation

Machine learning and artificial intelligence (AI/ML) methods are beginning to have significant impact in chemistry and condensed matter physics. For example, deep learning methods have demonstrated new capabilities for high-throughput virtual screening, and global optimization approaches for inverse design of materials. Recently, a relatively new branch of AI/ML, deep generative models (GMs), provide additional promise as they encode material structure and/or properties into a latent space, and through exploration and manipulation of the latent space can generate new materials. These approaches learn representations of a material structure and its corresponding chemistry or physics to accelerate materials discovery, which differs from traditional AI/ML methods that use statistical and combinatorial screening of existing materials via distinct structure-property relationships. However, application of GMs to inorganic materials has been notably harder than organic molecules because inorganic structure is often more complex to encode. In this work we review recent innovations that have enabled GMs to accelerate inorganic materials discovery. We focus on different representations of material structure, their impact on inverse design strategies using variational autoencoders or generative adversarial networks, and highlight the potential of these approaches for discovering materials with targeted properties needed for technological innovation.

36 MATERIALS SCIENCE↗

LLNL Contribution to Sandia Used Fuel Disposition - Security March 2011 Deliverable

Cleary [2007] divides the proliferation pathway into stages: diversion, facility misuse, transportation, transformation, and weapons fabrication. King [2010], using Cleary's methodology, compares a deepburn fusion-driven blanket containing weapons-grade plutonium with a PWR burning MOX fuel enrichments of 5-9%. King considers the stages of theft, transportation, transformation, and nuclear explosive fabrication. In the current study of used fuel storage security, a similar approach is appropriate. First, one must consider the adversary's objective, which can be categorized as on-site radionuclide dispersion, theft of material for later radionuclide dispersion, and theft of material for later processing and fabrication into a nuclear explosive. For on-site radionuclide dispersion, only a single proliferation pathway stage is appropriate: dispersion. That situation will be addressed in future reports. For later radionuclide dispersion, the stages are theft, transportation, and transformation (from oxide spent fuel containing both fission products and actinides to a material size and shape suitable for dispersion). For later processing and fabrication into a nuclear explosive, the stages are theft (by an outsider or by facility misuse by an insider), transportation, transformation (from oxide spent fuel containing both fission products and actinides to a metal alloy), and fabrication (of the alloy into a weapon). It should be noted that the theft and transportation stages are similar, and possibly identical, for later radionuclide dispersion and later processing and fabrication into a nuclear explosive. Each stage can be evaluated separately, and the methodology can vary for each stage. For example, King starts with the methodology of Cleary for the theft, transportation, transformation, and fabrication stages. Then, for each stage, King assembles and modifies the attributes and inputs suggested by Cleary. In the theft (also known as diversion) stage, Cleary has five high-level categories (material handling during diversion, difficulty of evading detection by the accounting system, difficulty of evading detection by the material control system, difficulty of conducting undeclared facility modifications for the purpose of diverting nuclear material, and difficulty of evading detection of the facility modifications for the purposes of diverting nuclear material). Each category has one or more subcategories. For example, the first category includes mass per significant quantity (SQ) of nuclear material, volume/SQ of nuclear material, number of items/SQ, material form (solid, liquid, powder, gas), radiation level in terms of dose, chemical reactivity, heat load, and process temperature. King adds the following two subcategories to that list: SQs available for theft, and interruptions/changes (normal and unexpected) in material stocks and flows. For the situation of an orphaned surface storage facility, this approach is applicable, with some of the categories and subcategories being modified to reflect the static situation (no additions or removals of fuel or containers). In addition, theft would require opening a large overpack and either removing a full container or opening that sealed container and then removing one or more spent nuclear fuel assemblies. These activities would require time without observation (detection), heavy-duty equipment, and some degree of protection of the thieves from radiological dose. In the transportation stage, Cleary has two high-level categories (difficulty of handling material during transportation, and difficulty of evading detection during transport). Each category has a number of subcategories. For the situation of an orphaned surface storage facility, these categories are applicable. The transformation stage of Cleary has three high-level categories (facilities and equipment needed to process diverted materials; knowledge, skills, and workforce needed to process diverted materials; and difficulty of evading detection of transformation activities). Again, there are subcategories. King [2007] adds a fourth high-level category: time required to transform the materials. For the situation of an orphaned surface storage facility, the categories are applicable, but the evaluations of each category and subcategory will be significantly different for later radionuclide dispersion than for later processing and fabrication into a nuclear explosive. The fabrication stage of Cleary has three high-level categories (difficulty associated with design, handling difficulties, and knowledge and skills needed to design and fabricate). King replaces the first two high-level categories with the Figure of Merit for Nuclear Explosives Utility (FOM), with subcategories of bare critical mass, heat content of transformed material, dose rate of transformed material, and SQs available for theft. The next section of this report describes the FOM in more detail.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Implications for advanced safeguards derived from PR&PP case study results

The proliferation resistance and physical protection (PR and PP) working group produced a case study on the Example Sodium Fast Reactor (ESFR). The ESFR is a hypothetical nuclear energy system consisting of four sodium-cooled fast reactors of medium size collocated with an on-site dry fuel storage facility and a spent fuel reprocessing facility using pyroprocessing technology. This study revealed how safeguards would be applied at such site consisting of integrated multiple fuel cycle facilities and the implications of what safeguards technology and safeguards concepts would need to be adapted and developed to safeguard successfully this Generation IV nuclear energy system concept. The major safeguards concepts driving our safeguards analysis are timeliness goals and material quantity goals. Because the fresh transuranic (TRU) fuel to be produced in the ESFR fuel fabrication facility contains plutonium, the ESFR will be reprocessing, using in the reactor, and storing material on site that will have IAEA defined 'direct-use material' in it with stringent timeliness goals and material quantity goals that drive the safeguards implementation. Specifically, the TRU fresh fuel, pyroprocessing in process material, LWR spent fuel sent to the ESFR, and TRU spent fuel will contain plutonium. This material will need to be verified at interim intervals four times per year because the irradiated direct-use material, as defined previously, has three-month timeliness goals and 8 kg material quantity goals for plutonium. The TRU in-process material is, of course, irradiated direct-use material as defined by the IAEA. Keeping the plutonium and uranium together with TRu products should provide a radiation barrier. this radiation barrier slows down the ability to reprocess the fuel. Furthermore, the reprocessing technique, if it has some intrinsic proliferation resistance, will need major modifications to be able to separate plutonium from the uranium and TRU mixture. The ESFR design should have such features in it if it is seen to have intrinsic proliferation resistance. The technical difficulty in diverting material from the ESFR is at least as strongly impacted by the adversaries overall technical capabilities as it is by the effort required to overcome those barriers intrinsic to the nuclear fuel cycle. The intrinsic proliferation resistance of the ESFR will affect how extrinsic measures in the safeguards approach for the ESFR will provide overall proliferation resistance.

99 GENERAL AND MISCELLANEOUS↗