Search NASA⌕ Search

Engineering topics

Langner, Diana

Publications and source records attributed to Langner, Diana.

The attribute measurement technique

Any verification measurement performed on potentially classified nuclear material must satisfy two seemingly contradictory constraints. First and foremost, no classified information can be released. At the same time, the monitoring party must have confidence in the veracity of the measurement. An information barrier (IB) is included in the measurement system to protect the potentially classified information while allowing sufficient information transfer to occur for the monitoring party to gain confidence that the material being measured is consistent with the host's declarations, concerning that material. The attribute measurement technique incorporates an IB and addresses both concerns by measuring several attributes of the nuclear material and displaying unclassified results through green (indicating that the material does possess the specified attribute) and red (indicating that the material does not possess the specified attribute) lights. The attribute measurement technique has been implemented in the AVNG, an attribute measuring system described in other presentations at this conference. In this presentation, we will discuss four techniques used in the AVNG: (1) the 1B, (2) the attribute measurement technique, (3) the use of open and secure modes to increase confidence in the displayed results, and (4) the joint design as a method for addressing both host and monitor needs.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Random selection as a confidence building tool

Any verification measurement performed on potentially classified nuclear material must satisfy two seemingly contradictory constraints. First and foremost, no classified information can be released. At the same time, the monitoring party must have confidence in the veracity of the measurement. The first concern can be addressed by performing the measurements within the host facility using instruments under the host's control. Because the data output in this measurement scenario is also under host control, it is difficult for the monitoring party to have confidence in that data. One technique for addressing this difficulty is random selection. The concept of random selection can be thought of as four steps: (1) The host presents several 'identical' copies of a component or system to the monitor. (2) One (or more) of these copies is randomly chosen by the monitors for use in the measurement system. (3) Similarly, one or more is randomly chosen to be validated further at a later date in a monitor-controlled facility. (4) Because the two components or systems are identical, validation of the 'validation copy' is equivalent to validation of the measurement system. This procedure sounds straightforward, but effective application may be quite difficult. Although random selection is often viewed as a panacea for confidence building, the amount of confidence generated depends on the monitor's continuity of knowledge for both validation and measurement systems. In this presentation, we will discuss the random selection technique, as well as where and how this technique might be applied to generate maximum confidence. In addition, we will discuss the role of modular measurement-system design in facilitating random selection and describe a simple modular measurement system incorporating six small {sup 3}He neutron detectors and a single high-purity germanium gamma detector.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗