系统透明度信息对人机信任和协同决策的影响

目的 以信号检测论为切入点,探究将系统可靠性信息进行透明化设计后,其在不同系统可靠性和风险情境下对操作者的信任水平和对人机协同绩效的影响。方法 采用2（透明度）×2（风险）×3（可靠性）被试内设计,通过TNO信任任务收集了32位被试的数据,记录各条件下对自动化系统的信任水平、判断标准倾向和人机协同绩效,并对数据进行重复测量方差分析。结果 在不同可靠性水平和风险情境下,透明化设计产生的效果不同。相对于低透明化设计,当自动化系统可靠性较高时,高透明化设计可显著提升操作者对系统的主观信任评分、屈从率,更接近最优判断标准,带来更高的正确率增益、更小的人机最优绩效差异。风险水平影响人机协同绩效,高风险下的人机协同绩效显著低于低风险下人机协同绩效,并且当系统可靠性高时,操作者在高风险下对系统的屈从率、正确率增益较低风险下均显著下降。透明化设计可以改变不同风险下操作者的信任行为和绩效。高风险下,高透明度可显著提升操作者对系统的屈从率,减小与人机最优绩效的差异。结论 透明化设计有利于提高操作者的人机信任适配水平、优化决策判断标准、促进人机协同绩效的提升,尤其在系统高可靠性和高风险情境下,透明化设计的优...     

Automated diagnostic aids: The effects of aid reliability on users' trust and reliance

We examined the effects that different levels of, and changes in, automation reliability have on users' trust of automated diagnostics aids. Participants were presented with a series of testing trials in which they diagnosed the validity of a system failure using only information provided to them by an automated diagnostic aid. The initial reliability of the aid was either 60, 80 or 100% reliable. However, for participants initially provided with the 60%-reliable aid, the accuracy of the aid increased to 80% half way through testing, whereas for those initially provided the 100%-reliable aid, the aid's reliability was reduced to 80%. Aid accuracy remained at 80% throughout testing for participants in the 80%-reliability group. Both subjective measures (i.e. perceived reliability of the aids and subjective confidence ratings) and objective measures of performance (concurrence with the aid's diagnosis and decision times) were examined. Results indicated that users of automated diagnostic aids were sensitive to different levels of aid reliabilities, as well as to subsequent changes in initial aid reliabilities. However, objective performance measures were related to, but not perfectly calibrated with, subjective measures of confidence and reliability estimates. These findings highlight the need to distinguish between automation trust as a psychological construct that can be assessed only through subjective measures and automation reliance that can only be defined in terms of performance data. A conceptual framework for understanding the relationship between trust and reliance is presented.     

Human performance reliability: on-line assessment using fuzzy logic

This paper presents an on-line approach to monitoring human performance in terms of conditional reliability when one is performing a task. Unlike traditional human reliability analysis, this approach develops a dynamic model that is able to cope with constantly changing conditions that affect operator performance. A fuzzy knowledge-based assessment approach is developed in order to deal with uncertainty and subjectivity associated with human performance assessment. This technology includes three main parts/functions: (i) on-line performance monitoring; (ii) real-time performance forecasting; and (iii) performance reliability assessment. The technology is demonstrated in real-time and provides timely conditioned reliability information regarding task success/failure. In general, this technology offers human reliability assessment under highly dynamic circumstances.     

How does a collision warning system shape driver's brake response time? The influence of expectancy and automation complacency on real-life emergency braking

Brake Reaction Time (BRT) is an important parameter for road safety. Previous research has shown that drivers' expectations can impact RT when facing hazardous situations, but driving with advanced driver assistance systems, can change the way BRT are considered. The interaction with a collision warning system can help faster more efficient responses, but at the same time can require a monitoring task and evaluation process that may lead to automation complacency. The aims of the present study are to test in a real-life setting whether automation compliancy can be generated by a collision warning system and what component of expectancy can impact the different tasks involved in an assisted BRT process. More specifically four component of expectancy were investigated: presence/absence of anticipatory information, previous direct experience, reliability of the device, and predictability of the hazard determined by repeated use of the warning system. Results supply indication on perception time and mental elaboration of the collision warning system alerts. In particular reliable warning quickened the decision making process, misleading warnings generated automation complacency slowing visual search for hazard detection, lack of directed experienced slowed the overall response while unexpected failure of the device lead to inattentional blindness and potential pseudo-accidents with surprise obstacle intrusion.     

Analysis of driver roles: modelling the changing role of the driver in automated driving systems using EAST

ABSTRACT

The purpose of this paper is to analyse the role of the driver within automated driving systems using the Event Analysis of Systemic Teamwork (EAST) method. We already know that as the level of automation increases within the driving task, the role of the driver shifts from that of an active operator (i.e. a driver driving) to more of a passive monitor (i.e. a driver monitoring). Task, social and information networks were constructed using the Hierarchical Task Analysis of Driving and evidence from driver verbalisations collected during a previous study to further explore the changing role of the driver using network analysis. A ‘broken links’ approach was conducted to show that momentary engagement in non-driving-related secondary tasks within an automated driving system can dramatically change the structure of driving system.     

A case for information theory-based modelling of human multitasking performance

Technological advances have led to increased prevalence of human multitasking. Traditionally, multitasking performance is measured for each individual task, utilising multiple performance metrics. Recently, the human operator informatic model (HOIM) has been developed as a quantitative model of human–machine interaction capable of measuring multitasking performance with a single metric. In this study, we investigated the performance effects of both a different number of tasks and five different task combinations while maintaining a constant level of overall difficulty. Results indicate that, at a constant difficulty level, neither number of tasks nor task combination significantly affects multitasking performance, indicating the reliability of input baud rate as a measure of system complexity. No task interaction effects were found among the different tasks. This study demonstrates the application of information theory for modelling of human multitasking performance and makes a case for the use of the HOIM in future studies of human information processing.     

水力除焦监测系统的设计和研究

结合国内延迟焦化装置水力除焦的特点,将监测技术应用于水力除焦,研制出能在国内现有装置中普遍适用的水力除焦自动监测系统.为求达到监测水力除焦目的,整个监测系统需对两方面进行检测设计:焦炭塔振动信号检测设计和除焦器高度信号检测设计,通过采集处理2种信号实现监测某位置焦炭的切割状态;同时研究实际测试过程中采集到的信号的变化情况,通过分析,确保整套系统的可靠性和安全性.该监测系统应用于上海石化的延时焦化工艺,监测水力除焦工况状态,从而实现了水力除焦的自动状态监测,避免了操作工经验判别和受环境影响大的缺点,具有良好的经济效益和现实意义     

Automation cueing modulates cerebral blood flow and vigilance in a simulated air traffic control task

Automation cueing of impending hazards or critical signals can have both beneficial and deleterious effects on the human operator, depending on automation reliability. To assess these effects, transcranial Doppler sonography (TCD), a low-cost non-invasive procedure that allows continuous monitoring of blood flow in the left and right cerebral hemispheres, was used to examine the influence of automation cues of varying reliability on vigilance performance in a 40-min simulated air traffic control task. Four levels of automation cue reliability (100, 80 and 40% reliable cueing and a no-cue control) were combined factorially with two levels of critical signal salience (high and low). For both levels of signal salience, the detection rate of critical signals was very high and remained stable over time with 100% cue-reliability, but declined over time in the remaining cue conditions, so that, by the end of the task, performance efficiency was best in the 100% condition followed in order by the 80%, 40%, and no-cue conditions. These performance effects for cueing were very closely mirrored by changes in cerebral blood flow in the right (but not the left) hemisphere in conjunction with low salience signals. The results fit well with an attentional resource model of vigilance and show that cerebral blood flow provides a metabolic index of the utilization of information-processing resources during sustained attention. The demonstration of systematic modulation of cerebral blood flow with time on task, automation cueing and signal salience also provides strong evidence for a right hemispheric brain system that is involved in the functional control of vigilance performance over time. Measurement of the activation of this system, as a reflection of operator mental workload, can, therefore, inform the design of optimal automation cueing.     

A study on the development of a task complexity measure for emergency operating procedures of nuclear power plants

In this study, a measure called task complexity (TACOM) that can quantify the complexity of tasks stipulated in emergency operating procedures of nuclear power plants is developed. The TACOM measure consists of five sub-measures that can cover remarkable complexity factors: (1) amount of information to be managed by operators, (2) logical entanglement due to the logical sequence of the required actions, (3) amount of actions to be accomplished by operators, (4) amount of system knowledge in recognizing the problem space, and (5) amount of cognitive resources in establishing an appropriate decision criterion. The appropriateness of the TACOM measure is investigated by comparing task performance time data with the associated TACOM scores. As a result, it is observed that there is a significant correlation between TACOM scores and task performance time data. Therefore, it is reasonable to expect that the TACOM measure can be used as a meaningful tool to quantify the complexity of tasks.     

Attention and complacency

The problem of complacency is analysed, and it is shown that previous research that claims to show its existence is defective, because the existence of complacency can not be proved unless optimal behaviour is specified as a benchmark. Using gedanken experiments, it is further shown that, in general, not even with optimal monitoring can all signals be detected. Complacency is concerned with attention (monitoring, sampling), not with detection, and there is little evidence for complacent behaviour. To claim that behaviour is complacent is to blame the operator for failure to detect signals. This is undesirable, since so-called complacent behaviour may rather be the fault of poor systems design.     

Reliability analysis and operator modelling

The paper considers the state of operator modelling in reliability analysis. Operator models are needed in reliability analysis because operators are needed in process control systems. HRA methods must therefore be able to account both for human performance variability and for the dynamics of the interaction. A selected set of first generation HRA approaches is briefly described in terms of the operator model they use, their classification principle, and the actual method they propose. In addition, two examples of second generation methods are also considered. It is concluded that first generation HRA methods generally have very simplistic operator models, either referring to the time-reliability relationship or to elementary information processing concepts. It is argued that second generation HRA methods must recognise that cognition is embedded in a context, and be able to account for that in the way human reliability is analysed and assessed.     

Understanding reliance on automation: effects of error type,error distribution,age and experience

An obstacle detection task supported by ‘imperfect’ automation was used with the goal of understanding the effects of automation error types and age on automation reliance. Sixty younger and sixty older adults interacted with a multi-task simulation of an agricultural vehicle (i.e. a virtual harvesting combine). The simulator included an obstacle detection task and a fully manual tracking task. A micro-level analysis provided insight into the way reliance patterns change over time. The results indicated that there are distinct patterns of reliance that develop as a function of error type. A prevalence of automation false alarms led participants to under-rely on the automation during alarm states while over-relying on it during non-alarm states. Conversely, a prevalence of automation misses led participants to over-rely on automated alarms and under-rely on the automation during non-alarm states. Older adults adjusted their behaviour according to the characteristics of the automation similar to younger adults, although it took them longer to do so. The results of this study suggest that the relationship between automation reliability and reliance depends on the prevalence of specific errors and on the state of the system. Understanding the effects of automation detection criterion settings on human–automation interaction can help designers of automated systems to make predictions about human behaviour and system performance as a function of the characteristics of the automation.     

Similarities and differences between human–human and human–automation trust: an integrative review

The trust placed in diagnostic aids by the human operator is a critical psychological factor that influences operator reliance on automation. Studies examining the nature of human interaction with automation have revealed that users have a propensity to apply norms of human–human inter-personal interaction to their interaction with ‘intelligent machines’. Nevertheless, there exist subtle differences in the manner in which humans perceive and react to automated aids compared to human team-mates. In the present paper, the concept of trust in human–automation dyads is compared and contrasted with that of human–human dyads. A theoretical framework that synthesizes and describes the process of trust development in humans vs automated aids is proposed and implications for the design of decision aids are provided. Potential implications of this research include the improved design of decision support systems by incorporating features into automated aids that elicit operator responses mirroring responses in human–human inter-personal interaction. Such interventions will likely facilitate better quantification and prediction of human responses to automation, while improving the quality of human interaction with non-human team-mates.     

基于近红外光学脑成像实验的轻型装配作业的中断效应研究

中断任务的复杂度和相似性对作业者操作绩效和脑认知负荷有不同程度影响。通过设计不同复杂度和不同相似性中断情景下的乐高模型模拟轻型装配实验，记录作业者操作绩效指标，并利用近红外光学脑成像设备监测其左右脑前额叶含氧血红蛋白浓度变化，分析中断类型对轻型装配作业者操作绩效和脑认知负荷的影响规律及其内在机理。结果表明，中断任务的复杂度不会导致作业者操作绩效和脑认知负荷出现显著差异，相似中断会导致作业者操作绩效下降，不相似中断会导致作业者脑认知负荷增加。研究发现了作业者应对中断时的注意力资源分配规律，并发展了动作技能型任务注意力焦点转换过程中的工作记忆理论，从而指导轻型装配企业合理制定标准作业时间和中断管理策略，降低作业者脑认知负荷，以实现可持续高效生产。     

Perceived production complexity – understanding more than parts of a system

To successfully manage today’s complex production systems it is essential to study operators’ perception of the system. The paper presents perceived production complexity assessed at seven manufacturing companies with the CompleXity Index (CXI) method. While other methods have measured product variants, layout, work content, tools and information items, CXI combines them into three areas. These three complexity areas (Station design, Work variance and Disturbance handling) help to create an understanding of the complex system by visualising how they contribute to complexity, and show the need to support the operators’ ability to handle the varying work.     

Optimal confidence intervals on the variation for operators and gauge in manufacturing process with repeated measurements

It is necessary to measure the attributes of the parts in any manufacturing process. It is also important to monitor measurement system in the manufacturing process because repeated measurements of the attributes include variability as well as target value. This paper considers variabilities due to repeated measurements, operators, and gauge in a measurement system. The measurement system is statistically modeled as a two-factor mixed model with one covariate and interaction. That is, this model employs J operators randomly chosen to conduct measurements on I randomly selected parts from a manufacturing process. In this experiment each operator measures each part K times. This paper aims to provide engineering practitioners with statistically optimal confidence intervals on the variation due to operators and gauge resulted from a measurement system statistically modeled. The optimal confidence intervals are based on a moderate large sample method (MLS) and a generalized p-value method (GEN). The confidence intervals proposed can be useful tools to determine whether a manufacturing process is adequate for monitoring a measurement system.     

An ergonomics study of computerized emergency operating procedures: Presentation style, task complexity, and training level

Emergency operating procedures (EOPs) are widely used in nuclear power plants (NPPs). With the development of information technology, computerized EOPs are taking the place of paper-based ones. Unlike paper-based EOPs, the industrial practice of computerized EOPs is still quite limited. Ergonomics issues of computerized EOPs have not been studied adequately. This study focuses on the effects of EOP presentation style, task complexity, and training level on the performance of the operators in the execution of computerized EOPs. One simulated computerized EOP system was developed to present two EOPs, each with different task complexity levels, by two presentation styles (Style A: one- and two-dimensional flowcharts combination; Style B: two-dimensional flowchart and success logic tree combination). Forty subjects participated in the experiment of EOP execution using the simulated system. Statistical analysis of the experimental results indicates that: (1) complexity, presentation style, and training level all can significantly influence the error rate. High-complexity tasks and lack of sufficient training may lead to a higher error rate. Style B caused a significantly higher error rate than style A did in the skilled phase. So the designers of computerized procedures should take the presentation styles of EOPs into account. (2) Task complexity and training level can significantly influence operation time. No significant difference was found in operation time between the two presentation styles. (3) Training level can also significantly influence the subjective workload of EOPs operations. This implies that adequate training is very important for the performance of computerized EOPs even if emergency responses with computerized EOPs are much more simple and easy than that with paper-based EOPs.     

Neural Operation Management: A New Avenue for Productive and Military Operations

An important effect of technological progress is the increasing replacement of manual labor by mental labor in productive and military operations. The variation of the operator’s capabilities in cognition, judgment and decision-making has drawn much attention from operation management researchers. Monitoring and evaluation of these capabilities is especially significant in conditions such as long-time operation, operation with special properties and operation under special circumstances. The military power and economic power are both the key concerns for a nation. The military power depends not only on the weapon system, but also the operators’ capabilities of manipulating the system. Similarly, the economic power is not only dependent on advanced machine system, but also the operational capability of the operators. Thus it has become a hot field of research and practice to monitor and assess the operator’s physiological and psychological states online based on neural measurement technology, and then to give real time intervention, so as to reduce the occurrence of accidents and increase the operation performance.     

Individual differences in human–machine trust: A multi-study look at the perfect automation schema

The current article reports the results of three different studies which explored the relationship between the Perfect Automation Schema (PAS) and trust in a human–machine context. It was expected that PAS would be related to higher trust and higher performance expectations for automation. Studies 1 and 2 used a human-robot interaction simulator as the context for human-machine interaction and study 3 used an applied sample of F-16 pilots and their views of the Automatic Ground Collision Avoidance System (Auto-GCAS) which is an advanced automated safety system. Results from all three studies demonstrated that the high expectations facet of PAS had a positive relationship with trust, but not the all-or-none facet. These results suggest that the PAS as measured by high expectations may be a fruitful construct for researchers in the domain of trust in automation.