Engineering topics

Robinson, Sean M

Publications and source records attributed to Robinson, Sean M.

5 records

DOE OSTI2012

Statistical Methods Applied to Gamma-ray Spectroscopy Algorithms in Nuclear Security Missions

In a wide range of nuclear security missions, gamma-ray spectroscopy is a critical research and development priority. One particularly relevant challenge is the interdiction of special nuclear material for which gamma-ray spectroscopy supports the goals of detecting and identifying gamma-ray sources. This manuscript examines the existing set of spectroscopy methods, attempts to categorize them by the statistical methods on which they rely, and identifies methods that have yet to be considered. Our examination shows that current methods effectively estimate the effect of counting uncertainty but in many cases do not address larger sources of decision uncertainty—ones that are significantly more complex. We thus explore the premise that significantly improving algorithm performance requires greater coupling between the problem physics that drives data acquisition and statistical methods that analyze such data. Untapped statistical methods, such as Bayes Modeling Averaging and hierarchical and empirical Bayes methods have the potential to reduce decision uncertainty by more rigorously and comprehensively incorporating all sources of uncertainty. We expect that application of such methods will demonstrate progress in meeting the needs of nuclear security missions by improving on the existing numerical infrastructure for which these analyses have not been conducted.

detection algorithm.↗

DOE OSTI2012

Imaging for dismantlement verification: information management and analysis algorithms

The level of detail discernible in imaging techniques has generally excluded them from consideration as verification tools in inspection regimes. An image will almost certainly contain highly sensitive information, and storing a comparison image will almost certainly violate a cardinal principle of information barriers: that no sensitive information be stored in the system. To overcome this problem, some features of the image might be reduced to a few parameters suitable for definition as an attribute, which must be non-sensitive to be acceptable in an Information Barrier regime. However, this process must be performed with care. Features like the perimeter, area, and intensity of an object, for example, might reveal sensitive information. Any data-reduction technique must provide sufficient information to discriminate a real object from a spoofed or incorrect one, while avoiding disclosure (or storage) of any sensitive object qualities. Ultimately, algorithms are intended to provide only a yes/no response verifying the presence of features in the image. We discuss the utility of imaging for arms control applications and present three image-based verification algorithms in this context. The algorithms reduce full image information to non-sensitive feature information, in a process that is intended to enable verification while eliminating the possibility of image reconstruction. The underlying images can be highly detailed, since they are dynamically generated behind an information barrier. We consider the use of active (conventional) radiography alone and in tandem with passive (auto) radiography. We study these algorithms in terms of technical performance in image analysis and application to an information barrier scheme.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

DOE OSTI2011

Low-Intrusion Techniques and Sensitive Information Management for Warhead Counting and Verification: FY2011 Annual Report

Future arms control treaties may push nuclear weapons limits to unprecedented low levels and may entail precise counting of warheads as well as distinguishing between strategic and tactical nuclear weapons. Such advances will require assessment of form and function to confidently verify the presence or absence of nuclear warheads and/or their components. Imaging with penetrating radiation can provide such an assessment and could thus play a unique role in inspection scenarios. Yet many imaging capabilities have been viewed as too intrusive from the perspective of revealing weapon design details, and the potential for the release of sensitive information poses challenges in verification settings. A widely held perception is that verification through radiography requires images of sufficient quality that an expert (e.g., a trained inspector or an image-matching algorithm) can verify the presence or absence of components of a device. The concept of information barriers (IBs) has been established to prevent access to relevant weapon-design information by inspectors (or algorithms), and has, to date, limited the usefulness of radiographic inspection. The challenge of this project is to demonstrate that radiographic information can be used behind an IB to improve the capabilities of treaty-verification weapons-inspection systems.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

DOE OSTI2011

Non-Invertible Transforms for Image-Based Verification

Imaging may play a unique role in verifying the presence and distribution of warhead components in warhead counting and dismantlement settings where image information content can distinguish among shapes, forms, and material composition of items. However, a major issue with imaging is the high level of intrusiveness, and in particular, the possible need to store sensitive comparison images in the inspection system that would violate information barrier (IB) principles. Reducing images via transformations or feature extraction can produce image features (e.g. attributes) for verification, but with enough prior information about structure the reduced information itself may be sufficient to deduce sensitive details of the original image. Further reducing resolution of the transformed image information is an option, but too much reduction destroys the quality of the attribute. We study the possibility of a one-way transform that allows storage of non-sensitive reference information and analysis to enable comparison of transformed images within IB constraints. In particular, we consider the degree to which images can be reconstructed from image intensity histograms depending on the number of pixel intensity bins and the degree of frequency data quantization, as well as assumed knowledge of configuration of objects in the images. We also explore the concept of a 'perceptual hash' as a class of transforms that may enable verification with provable non-invertibility, leading to an effective one-way transform that preserves the nature of the image feature data without revealing sufficient information to reconstruct the original image.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2011

IMAGE-BASED VERIFICATION: SOME ADVANTAGES, CHALLENGES, AND ALGORITHM-DRIVEN REQUIREMENTS

ABSTRACT Imaging technologies may be a particularly useful technique that supports monitoring and verification of deployed and stockpiled nuclear weapons and dismantlement components. However, protecting the sensitive design information requires processing the image behind an information barrier and reporting only non-sensitive attributes related to the image. Reducing images to attributes may destroy some sensitive information, but the challenge remains. For example, reducing the measurement to an attribute such as defined shape and X-ray transmission of an edge might reveal sensitive information relating to shape, size, and material composition. If enough additional information is available to analyze with the attribute, it may still be possible to extract sensitive design information. In spite of these difficulties, the implementation of new treaty requirements may demand image technology as an option. Two fundamental questions are raised: What (minimal) information is needed from imaging to enable verification, and what imaging technologies are appropriate? PNNL is currently developing a suite of image analysis algorithms to define and extract attributes from images for dismantlement and warhead verification and counting scenarios. In this talk, we discuss imaging requirements from the perspective of algorithms operating behind information barriers, and review imaging technologies and their potential advantages for verification. Companion talks will concentrate on the technical aspects of the algorithms.

IMAGE-BASED VERIFICATION: SOME ADVANTAGES, CHALLEN↗