Viewing Strategy and Window Effects on Dynamic Visual-Inspection Accuracy

Equally-spaced product units, moving on a belt at a uniform velocity, were exposed for visual inspection over a fixed viewing window. Viewing windows were equal to the product unit interspacing and twice this distance to give respectively 1 and 2 product units within the window. Three different belt speeds were used. Subjects performed short runs of visual inspection using their self-directed strategies of viewing the sequence of product units and then these subjects repeated all conditions of this task using a “directed viewing strategy.” Improved viewing discipline was expected from this “directed” strategy and eye motion monitoring was employed to verify that the subjects used the “directed” strategy. Inspection errors of each type were observed under twelve experimental conditions. Both types of errors were found to be greater with: 1) the smaller viewing window, 2) self-directed viewing strategies in contrast to the “directed” strategy, and 3) higher belt velocities. Contrasts were also made between error types and between high- and low-performing subject groups. Inferences on the apparent cost of each error type were made from these data and compared with rational objectives. Economic implications are shown for these viewing-strategy and viewing window effects.     

Contemporary Approaches to Paced Visual Inspection

An approach to the study of inspection task design has been investigated. Twenty female inspectors did a simulated paced visual inspection task for 90 minutes each. The experimental variables were display size, product pacing, and fault information as a measure of defect probability. They were tested in a 15 point design with 3 replications of the center point. The criterion of performance was the fraction of total products correctly identified as being defect free. From the experimental data quantitative estimates of performance increments or decrements associated with changes in inspection task variables were determined. The implications of observations and their importance in the design of an inspection task have been discussed.     

An on-road assessment of cognitive distraction: impacts on drivers' visual behavior and braking performance

In this on-road experiment, drivers performed demanding cognitive tasks while driving in city traffic. All task interactions were carried out in hands-free mode so that the 21 drivers were not required to take their visual attention away from the road or to manually interact with a device inside the vehicle. Visual behavior and vehicle control were assessed while they drove an 8 km city route under three conditions: no additional task, easy cognitive task and difficult cognitive task. Changes in visual behavior were most apparent when performance between the No Task and Difficult Task conditions were compared. When looking outside of the vehicle, drivers spent more time looking centrally ahead and spent less time looking to the areas in the periphery. Drivers also reduced their visual monitoring of the instruments and mirrors, with some drivers abandoning these tasks entirely. When approaching and driving through intersections, drivers made fewer inspection glances to traffic lights compared to the No Task condition and their scanning of intersection areas to the right was also reduced. Vehicle control was also affected; during the most difficult cognitive tasks there were more occurrences of hard braking. Although hands-free designs for telematics devices are intended to reduce or eliminate the distraction arising from manual operation of these units, the potential for cognitive distraction associated with their use must also be considered and appropriately assessed. These changes are captured in measures of drivers' visual behavior.     

VISUAL INSPECTION STRATEGY

Considerable attention has already been devoted to the many aspects of the inspection task and findings suggest that the factors affecting the efficiency of inspection are subtle and not easily controlled. However, observation of the visual inspection techniques used in most factories suggests that this is not generally recognised and that there is an entirely misplaced confidence in the effectiveness of 100% visual inspection

In this study, consideration is given to some of the factors affecting visual inspection, and trials have been made, which, whilst showing inspection decisions to be inconsistent, suggest ways of achieving a progressive improvement. It is recognised, however, that variability is inevitable in the making of subjective judgements.     

A Search Performance Test for Visual Lobe Size

Visual search is a very common form of inspection and monitoring activity. This paper reviews work by the author that demonstrates a relationship between visual lobe size and search performance. A card sorting task was found to correlate significantly with visual lobe dimensions and it is suggested that this relationship may be used as the basis of a test for screening individuals for certain types of inspection and monitoring activities. The test is considered to be a more useful measure of visual scanning ability than tests of foveal acuity.     

Sequencing n Products Involving m Independent Jobs on m Machines

This article considers the problem of scheduling n products over m distinct machines. Every product consists of a set of jobs, each requiring a known processing time on a designated machine. There are no precedence constraints, and simultaneous processing of jobs requiring different machines within a product is allowed. The object of scheduling is to minimize a regular measure of performance associated with the products. It is shown that there exists an optimal schedule with the “no passing property.” Branch and bound routines are developed for finding the optimal solution for the two measures of performance: (1) total penalty cost; and (2) sum of product completion times. Comparisons between the optimal solution and solutions obtained using dispatching rules are given in the penalty cost case.     

ATTRIBUTES FOR EXPEDIENT COMPUTER VISION INSPECTION

The visual inspection of parts as they progress through die manufacturing process is an important task in all industries. Visual inspection, when performed by humans is a tedious task and is prone to error. This is precisely what makes it a good candidate for automation. Although computer vision systems have been around for over 30 years, die industrial applications of vision systems have become practical only in die last decade. Image processing and pattern recognition algorithms used in industrial vision systems are built upon a broad body of knowledge in vision research. But the use of computer vision systems in quality control has been limited to replicating die visual inspection tasks as they would be performed by a human operator. It is die contention of this study that when computerized inspection is employed, quality control inspection plans suitable for computerized inspection should also be employed to assure cost-effectiveness.

We study a simple gauging inspection task and propose a quality control plan that exploits die characteristics of computer vision systems in order to improve cost-effectiveness.     

Identifying cognitive distraction using steering wheel reversal rates

The influence of driver distraction on driving performance is not yet well understood, but it can have detrimental effects on road safety. In this study, we examined the effects of visual and non-visual distractions during driving, using a high-fidelity driving simulator. The visual task was presented either at an offset angle on an in-vehicle screen, or on the back of a moving lead vehicle. Similar to results from previous studies in this area, non-visual (cognitive) distraction resulted in improved lane keeping performance and increased gaze concentration towards the centre of the road, compared to baseline driving, and further examination of the steering control metrics indicated an increase in steering wheel reversal rates, steering wheel acceleration, and steering entropy. We show, for the first time, that when the visual task is presented centrally, drivers’ lane deviation reduces (similar to non-visual distraction), whilst measures of steering control, overall, indicated more steering activity, compared to baseline. When using a visual task that required the diversion of gaze to an in-vehicle display, but without a manual element, lane keeping performance was similar to baseline driving. Steering wheel reversal rates were found to adequately tease apart the effects of non-visual distraction (increase of 0.5° reversals) and visual distraction with offset gaze direction (increase of 2.5° reversals). These findings are discussed in terms of steering control during different types of in-vehicle distraction, and the possible role of manual interference by distracting secondary tasks.     

Qualitative Ordinal Scales: The Concept of Ordinal Range

Many practical problems of quality control involve the use of ordinal scales. Questionnaires planned to collect judgments on qualitative or linguistic scales, whose levels are terms such as “good,” “bad,” “medium,” etc., are extensively used both in evaluating service quality and in visual controls for manufacturing industry. In an ordinal environment, the concept of distance between two generic levels of the same scale is not defined. Therefore, a population (universe) of judgments cannot be described using “traditional” statistical distributions since they are based on the notion of distance. The concept of “distribution shape” cannot be defined as well. In this article, we introduce a new statistical entity, the so-called ordinal distribution, to describe a population of judgments expressed on an ordinal scale. We also discuss which of the traditional location and dispersion measures can be used in this context and we briefly analyze some of their properties. A new dispersion measure, the ordinal range, as an extension of the cardinal range to ordinal scales, is then proposed. A practical application in the field of quality is developed throughout the article.     

An Elementary Model of Human Performance on Paced Visual Inspection Tasks

An analysis of type one and type two error committed by human inspectors on a paced visual inspection task was conducted. The analysis was accomplished by development of a two-phase quantitative model describing the inspector's behavior when performing such inspection tasks: (1) the observe-not observe decision model described the inspector's decision to attend or not attend his task at any particular time; (2) the accept-reject decision model, which was based on signal detection theory, described the inspector's decision to classify a product unit as good or defective, given that he has made the decision to observe. Experimental evaluation supported the model, but did not provide complete validation.     

The contribution of inhibitory deficits to dangerous driving among young people

A recent theory of adolescent risk taking that may be applicable to young drivers proposed that young people engage in more risks because they are more impulsive. While past research has found that problematic drivers do tend to score higher on measures of impulsivity, most of this research has relied on self-reported behaviours and attitudes. The present study investigates the role of impulsivity using computer-based measures of inhibitory functioning. Young drivers who had been caught speeding by the police were compared with non-offenders on two inhibitory measures: the Stop-signal task and the Go/no-go task. While the two groups did not differ in their performance on the Stop-signal task, there were significant differences between the groups on the Go/no-go task with the offender group displaying lower inhibitory skills. The results of the Go/no-go task were not entirely unambiguous as offenders were also found to have responded to go trials with a faster reaction-time. The implications of these results both for the impulsivity theory of adolescent risk taking and for the more general issue of adolescent risk taking are discussed.     

Economic development of specifications for 100% inspection based on asymmetric quality loss functions

As a means of providing protection for both the producer and the consumer, for example as in airport security, 100% inspection plays an important role in today's environment. The same is true for some industrial or other decision-making processes where the consequences of excessive deviations from target values are very high. Motivated by Deming's “all or none” inspection rule and associated philosophy for quality improvement, we develop an inspection strategy by considering the costs to both the producer and the customer, and thus to the whole system. A general optimization model is formulated to determine whether 100% inspection is to be performed or no inspection to be performed. If complete inspection is chosen then the optimal specification limits can be determined at the same time. Specifically, we consider two types of asymmetric quality loss functions for the “target the best” quality characteristic: the asymmetric quadratic quality loss function, and the asymmetric piecewise linear loss function. For each loss function, we propose an optimization model to determine the optimal process mean and specification limits. Numerical examples are given to illustrate the proposed models that can be used to improve the quality of the output of any process.     

Minimum Cost 100% Inspection Systems with Inspection Error

A field failure in certain products has the potential to cause human suffering and/or loss of life. To the manufacturer, these failures will result in a significant external failure cost. In this paper, cost minimization models are developed to determine an optimal inspection system for multiple 100% attribute inspections with inspection error. An example involving the visual inspection of a braze joint in a cardiac biopsy device is presented to illustrate the use of the models. A key result is that a system requiring an item to pass every inspection in order to be shipped to a customer is not always the best alternative. Examples illustrating the sensitivity of the model results to key cost and operational parameters are shown.     

Performance Effects of Variables in Dynamic Visual Inspection

Three task variables of dynamic visual inspection were examined in laboratory experiments: (V) conveyor-belt velocities, (T) target exposure time and (P) the viewing position of the inspector. Measurements were obtained on both error types in inspection of equally-spaced targets for 36 experimental condition. Task variables V and T provided significant main effects on inspection accuracy and variables T and P had an interacting effect. The probability of correctly identifying a target was found to be describable as l-2.58e^-a where a = 10.587T- 0.0.75/V-0.012/[T+ P²] and the coefficient of determination, R² = 0.88. Inspector's eye motions were recorded and gross eye-motion patterns were identified but none appeared to be directly associated with inspection accuracy. Shifts in the inspection error criterion were observed and described for changes in these task variables. Both inspection speed and accuracy are combined in a cost model. This study shows that different conveyor speeds have better economics with different error costs but that shorter exposure times were never economically better.     

Time-redundant system reliability under randomly constrained time resources

The paper considers the time-redundant system where the system total task is a sequence of n phases and the total task must be executed during constrained time. For every phase, there is its own server, which executes the phase task during randomly distributed time. The server is not perfectly reliable and two types of failure (“open” and “closed”) are possible. Redundant servers may be used in any phase. The time–probability characteristics are introduced for any task, based on which the system reliability is treated as a probability that the system total task will be correctly completed during a corresponding time resource, which also may be randomly distributed. The adequate model is presented and a semi-Markov process is used as a mathematical technique. The closed-form solution was derived based on an acyclic Semi-Markov process. The numerical example of the elaborated approach is presented.     

Double-stage inspection for screening multi-characteristic items

Final product inspection of a multi-attribute product, such as in the electronic assembly industry, involves expensive facilities. The correlation among attributes may be used for reducing the efforts needed for screening the products; however, engineers without an appropriate statistical-economical analysis tools do not take risks, and they designate full inspection of each item. We propose a double stage inspection program for reducing inspection efforts. Assuming that the joint distribution is known, the conditional probability that a product is “good” may be evaluated conditional upon the observation of the product's first-stage inspected quality attributes. Then, an expected cost minimization is implemented in order to decide whether a second inspection stage is required or a classification should be based solely on the first inspection stage. The cost factors include inspection and false classification. The method is illustrated on a real data set from a particular electronic product of Motorola-Arad Ltd.     

Why use noise?

Measuring the dependence of visual sensitivity on parameters of the visual stimulus is a mainstay of vision science. However, it is not widely appreciated that visual sensitivity is a product of two factors that are each invariant with respect to many properties of the stimulus and task. By estimating these two factors, one can isolate visual processes more easily than by using sensitivity measures alone. The underlying idea is that noise limits all forms of communication, including vision. As an empirical matter, it is often useful to measure the human observer's threshold with and without a noise background added to the display, to disentangle the observer's ability from the observer's intrinsic noise. And when we know how much noise there is, it is often useful to calculate ideal performance of the task at hand, as a benchmark for human performance. This strips away the intrinsic difficulty of the task to reveal a pure measure of human ability. Here we show how to do the factoring of sensitivity into efficiency and equivalent noise, and we document the invariances of the two factors.     

Optimum speed of visual inspection using a systematic search strategy

In an earlier paper, the authors presented models of human functioning in visual inspection and derived optimal working speeds to balance the cost of time and the cost of errors. Both self-paced and externally paced models assumed a random search strategy. It was noted that choosing a standard speed for visual inspection is difficult with standard work measurement techniques and that the data required for the suggested search model are easily measurable. The purpose of this paper is to complement the earlier work by deriving similar models employing a systematic search strategy. The two models can be considered as bounds on actual performance; thus the current paper complements the previous one by establishing an upper bound rather than a lower bound on performance in search tasks. Comparisons between the two strategies are presented.     

CONWIP-based production lines with multiple bottlenecks: performance and design implications

This research focuses on CONWIP, a closed production control system where all containers traverse a circuit incorporating the entire production line. We develop estimates, for an important level of work in process inventory, for four important performance measures: the means and variances of time between departures and flowtime. We develop our estimates through the concept of a “conceptual bottleneck machine”. This concept enables us to develop an analogy between deterministic and stochastic systems. This concept also allows us to handle migrating bottlenecks, an issue generally neglected. The model is widely applicable, assuming only finite means and variances of the processing time distributions. We test our model computationally, both against existing models and on a wide range of randomly generated problems. Finally we detail insights, obtained from our analytical model, into how CONWIP production systems operate. These insights enable us to explain the sources of the values of our performance measures, thus aiding system design and modification.     

程序性知识视觉表征与用户类型对用户信息检索体验的影响

目的 高校图书馆数据库是获取信息资源的重要渠道,已有研究关注了界面系统的美学体验,忽视了视觉信息表征的逻辑性与可理解性,导致用户信息检索效率低的问题。方法 采用控制实验方法,探究程序性知识视觉表征和用户类型对高校图书馆数据库用户信息检索体验和认知负荷的影响。结果程序性知识视觉表征对信息检索体验和认知负荷均有显著影响,并且用户类型与程序性知识视觉表征对可用性和认知负荷具有交互作用。新手用户在使用下拉的视觉表征时信息检索体验最高,老手用户更适合使用组合的视觉表征方式。结论 通过提升用户在数据库中的信息检索体验,可以为检索界面设计提供依据。