Материал: Крючков Фундаменталс оф Нуцлеар Материалс Пхысицал Протецтион 2011

∙monitoring of quite zones on both sides of the barcode symbol;

∙identification of the symbol type by analyzing some of the initial and closing bars;

∙comparison of the signal with memorized samples and conversion of same to ASCII codes;

∙performance of check calculations;

∙transmission of data to the computer.

Data is exchanged between the reader (or the data capture terminal device) and the PC through an interface which supports electric and logic reader-PC mating.

Data exchange between the barcode hardware and the РС is done via:

∙COM ports (RS 232С);

∙LPT ports (Centronics);

∙Net ports (Ethernet).

RS 232C (RS 422/485) is the standard recommended for serial transmission of data and communication facilities. This is the most popular, relatively simple and system-independent interface supported by all hardware makers. However, RS 232C has a low noise resistance and requires extra components (modems) for long-range communications.

Centronics Parallel is a parallel data transmission standard. Data is transmitted by bytes.

Ethernet is a network data transmission standard. Brief descriptions of some barcode readers:

Trakker Reader 9445 (INTERMEC), a portable self-contained programmable barcode reader with data initial handling and interim storing features. Includes a LED emitter, an LCD, a standard RS 232 interface and a 64Kb program and data memory.

Trakker ANTARES 2420 (INTERMEC), a self-contained portable barcode reader. Includes a laser scanner, an LCD, a standard RS 232 interface and a 2Mb program and data memory.

JANUS J2020 (INTERMEC), a portable self-contained barcode reader. Includes a 386-processor DOS minicomputer. Features a multi-disk memory, one of the disks (1 Мb) allocated for applications and data. The instrument is fitted with a laser scanner, an LCD, a near-eye display and a standard RS 232 interface.

Three programmable barcode readers deserve a special note here. PictoRL (Pictorial Reader Language), a generator of barcode reading

applications. So PictoRL’s software package includes a pictorial language for readers. The software package enables applications to be made by

416

barcode hardware users with no special skills. The PictoRL package includes an ICW application that supports reader-PC data exchange. PictoRL’s software package is intended for use with Intermec readers, including Trakker Reader 9445, Trakker ANTARES 2420 and JANUS J2020.

EZBuilder, a simple and graphic instrument to create applications for Intermec readers. Intended for particular reader types, including Trakker ANTARES and JANUS. The reader can be used to create menus, displays, data fields, tags and so on. Features built-in PC-reader data exchange and does not require operators with SI programming skills. Includes a highpower HELP–file for rapid user self-training. A use r simply needs to select some of the options and preset the system of parameters for EZBuilder to create the application for the above reader types. There is also a built-in simulator enabling tests of applications for operability without being downloaded into the reader.

LabelView, an application package enabling users to design labels and tags. LabelView-designed labels may include texts, barcode symbols, graphic images, fields and more.

Efficiency of barcode technology in NM physical inventory takings

Barcode technology enables automation of marking check processes to cut so drastically the labor expenditures in periodic nuclear material inventory taking and control.

Experiments were conducted by the Kurchatov Institute to identify different nuclear material in containers or in fuel assemblies as individual samples and individual fuel elements. The identification procedure was to check their respective names against what was given in certificate data. This is a standard procedure used in physical inventory takings.

Statistically, visual comparisons turned out to entail the following personnel errors: trained, skilled, competent and motivated personnel had one error per one thousand of marking check operations on the average. The process was certain to involve rather a great scatter of uncertainties in a range of a factor of 10 both upwardly and downwardly. This means there was someone who made one error per 100 regulated actions and another one making one error per 10000 actions. On the average, however, there was one error per one thousand regulated actions. That is, if a person has a task to check one thousand markers, there is at least one case when he or she says they are definitely identical though actually these differ. After one

417

thousand markers, there is one case (on the average) when markers are different and he/she says they are absolutely identical.

It should be noted now that if inventoried items are one thousand named articles to be checked visually in a physical inventory taking (particular articles are checked against respective certificate data), one can imagine how many errors this will entail.

Barcode application effect. Checking is a relatively time-taking procedure. It takes about 30 seconds to check markers of ten symbols (certificate data and the item). This stems from the preparation of the item and its respective positioning to make markers readable by two or three readings since markers are long and cannot be forthwith memorized by humans. This means these should be compared in a by-symbol manner.

It was ultimately found that use of barcode markers, with a respective computer support in place, involved a marker misreading probability, this, as experimentally shown by the Kurchatov, found to be at a level of about 10–6 . There can be approximately one such error per one million markers. Still, the most important thing is that comparison here takes 1.5 to 2 s instead of 30 seconds needed by humans (the probability of an error being rather certain).

Conclusion. Barcode technologies offer a radical solution to problems involved in checks of articles with documented data for physical inventory taking.

9.2. Automation of NM measurements

Automated measurements of nuclear material form an essential component of automated NM accounting. Fortunately, the equipment used nowadays to measure and control nuclear material is either computerized or require communication to computer-based data handling systems.

Examples of NDA equipment

Until recently, up to the advent of electronic scales into nuclear industry, analytical laboratory scales were used to measure nuclear material. These are exclusively accurate lever-type scales capable of weighing nuclear material of up to 5 kg with an accuracy of 0.001 g. The scales normally have a seismic-proof base. Just imagine the length of a weighing session on such a scale. It takes much time, occasionally hours, because multiple measurements are required. Electronic scales, in turn, enable weighing to an accuracy of 0.1 g done virtually in seconds.

418

Electronic scales are also fitted with an RS–232 in terface for connection to a PC.

Gamma spectrometry suggests application of not only detectors and digital-to-analog transducers but also of a computer built in the system to process experimental data. This leads to a statistical data capture in 5 to 10 minutes and a computer processing time of 1.5 to 2 min. Depending on respective accuracy requirements and use of nuclear material samples, a measurement session may take from several minutes to an hour or several hours. The situation is nearly the same with neutron measurements based, say, on neutron wells. A switch to instrumentation with built-in statistical handling of measurement data has led to significantly less labor-consuming processing of measurement data and a much shorter time needed to get the final result.

The order of the day now is therefore the mating of accounting and control systems with measuring equipment that generates statistically handled data ready to be put in computerized NM A&C systems.

Program (hardware-based) interfaces have been developed and are used by the Kurchatov Institute for measurements in the КИ– МАКС system. Similar interfaces have been designed and employed by the Institute of Physics and Power Engineering (FEI) in Obninsk, the All-Russian Research Institute of Experimental Physics (VNIIEF) in Sarov and other organizations. These interfaces have made it possible to automate data transmission in NM A&C systems and decrease (by approximately three orders of magnitude) the probability of instrument data to be read erroneously. This leads to a current measurement automation system built in an accounting and control systems enabling instrument operators to make decisions competently and in a statistically sound manner and not to rewrite instrument readings (estimate to what extent the measurements performed fit or do not fit).

9.3. Systems approach to construction of enterprise NM A&C systems

Accounting and control of nuclear material is a complex man-machine system in which machines (devices, instruments, standards, computer networks) represent tamper indication, identification, measurement, data input and handling functions, while humans represent analyzing, decisionmaking and implementation functions. Normally, such systems incorporate a great deal of hardware support arrangements and features. Accounting and control of nuclear material is being technologically upgraded primarily to convert to computerized accounting and automated identification, control

419

and measurement of NM. So evolution of NM A&C nowadays largely needs these to be combined efficiently in a coupled system of arrangements and features. This is achieved through a systems approach to building an NM A&C system model (Fig. 9.5).

ОКОНЧАТЕЛЬНЫЙFinal systemПРОЕКТdesignСИСТЕМЫ

Fig. 9.5. Construction and assessment of an NM accounting and control system model

The systems approach primarily suggests definition of the systemsolved tasks. For NM A&C, this means determination of NM inventories and flows, as well as timely detection of unauthorized NM handling or potential NM identification and measurement errors. The latter gives significance to the listing of potential NM onsite handling or storing anomalies as a part of the system task definition.

The next step is to build a model of the system. This is the system model version evolution phase. The construction of an NM A&C system model needs to structure the site so as dictated by the NM disposition and flows, lay out hardware within this structure, define the sequence of major

420