Tasks error detection

During the PHEA stage, the analyst has to identify likely human errors and possible ways of error detection and recovery. The PHEA prompts the analyst to examine the main performance-influencing factors (PIFs) (see Chapter 3) which can contribute to critical errors. All the task steps at the bottom level of the HTA are analyzed in turn to identify likely error modes, their potential for recovery, their safety or quality consequences, and the main performance-influencing factors (PIFs) which can give rise to these errors. In this case study, credible errors were found for the majority of the task steps and each error had multiple causes. An analysis of two operations from the HTA is presented to illustrate the outputs of the PHEA. Figure 7.12 shows a PHEA of the two following tasks Receive instructions to pump and Reset system. 

Self-regulation strategies are strategies for deciding how to adapt to different circumstances, monitor complete or interrupted tasks, and detect errors. 

Task Analysis Describes tasks, information requirements, feedback, likely errors, detection methods and recovery procedures. 

Psychological The potential effects of triazolam on cognitive performance have been studied in a double-blind, placebo-controlled comparison of the acute effects of triazolam and alcohol in 20 social drinkers [15 ]. Triazolam was more likely to impair several objective measures of cognitive performance, such as episodic memory and divided attention, and it also slowed performance across several cognitive measures. Triazolam alone impaired other measures of metacognition, such as error detection on a choice reaction time task. 

This paper will focus on step 1 by verifying the soundness of implementation and partly the completeness of error detection of basic AN-encoded operations such as arithmetic operations (see Section 5). To make these verifications feasible we implemented an automated abstraction that reduces the muuber of test cases considerably (see Section 4). See the related work in Section 3 for our evaluation of the suitability of other methods for the described task. 

However a number of tasks impose strong constraints of availability and duration of service inteimption. This is particularly the case for telecommunications satellites. The use of computers in hot redundancy, or at least warm ", then reduces the recovery time in case of failure. Note that even in hot redundancy, a selective redundancy scheme is adopted (one unit is master and actually sends the commands). Error detection is based on the same techniques as in cold or warm redimdancy, without comparing outputs of redundant units. 

Several nontechnical factors can significantly affect the results of a nondestmctive inspection. Many of these are classified as human factors (1,2,17). Operator experience affects the probabiUty of detection of most flaws. Typically, an inexperienced operator has more false rejects, known as Type II errors, than an experienced operator. A poor operator has few false rejects but is more likely to miss a defect in the inspection, known as a Type I error. Operator fatigue, boredom, or an unfavorable environment such as lighting, cold, or rain may further affect performance. Thus it usually is a good investment for the inspection company to assure that the operator environment is most amenable to inspection, that the equipment is suitable for the task, and that the operator is alert and well rested. 

After the incident, an investigation team determined that the first operator had not added the initiator when required earlier in the process. When the relief operator added the initiator, the entire monomer mass was in the reactor and the reaction was too energetic for the cooling system to handle. Errors by both operators contributed to the runaway. Both operators were performing many tasks. The initiator should have been added much earlier in the process when much smaller quantities of monomer were present. There was also no procedure to require supervision review if residual monomers were detected. The lesson learned was that operators need thorough training and need to be made aware of significant hazardous scenarios that could develop. 

The performance standard achieved when every task is performed right first time with no errors being detected downstream. 

Process transients and equipment failures may require workers to develop a new strategy to control the process. Detection, diagnosis, and fault-compensation are tasks in which workers may have little experience and the information needs may be different from those of familiar tasks. Again, methods of task and error analyses, particularly those concerned with human cognitive functions, may be useful in deciding what information should be displayed to help workers detect process transients, diagnose their causes and develop new strategies. 

Structured root cause analysis uncovers the underlying reasons for human error and consequently provides guidance on suitable corrective actions. Humans make errors. Our task is to design systems that detect and correct an error before it leads to a serious consequence. Chapter 6 provides extensive information applicable during root cause analysis. 

However, systems are usually not made up just of computers -rather they will also involve other entities (e.g., devices and humans) which in many cases will not be able to simply forget some of their recent activity, and so simply go straight back to an exact earlier state when told that an error has been detected. Thus forward error recovery (the typical programming mechanism for which is exception handling), rather than backward recovery will have to be used. Each of these complications individually makes the task of error recovery more difficult, and together they make it much more challenging. This in fact is the topic that I and my colleagues have concentrated on these last few years. 

Some of the examples and discussion in this chapter draw on the two-class classification problem, which here is hit versus inactive . The word active refers to a validated hit, that is, a molecule that truly does exhibit some level of the desired biological response. A key point is that an assay is itself an estimator. With this in mind, definitions and a discussion of error rates are given in the context of predictive models. Borrowing from the terminology of signal detection, the sensitivity of a model refers to the fraction of observed hits that are classified as (or predicted to be) hits by the model, and specificity refers to the fraction of observed inactives classified as inactives by the model. An observed hit is not necessarily an active molecule, but simply a molecule for which the primary screening result exceeded a decision threshold. Whether such a molecule turns out to be an active is a problem that involves the sensitivity of the assay, but the task at hand is for... 

The 9-hole box apparatus has been used with different schedules to evaluate the role of the dopamine systems in tests of sustained and divided attention, in which the rats have to detect and respond to brief light stimuli in any one of five open holes (Baunez and Robbins, 1999). Although this experiment was primarily designed to consider the effects of STN lesions, which produce a general disruption of performance in the attentional task (Baunez and Robbins, 1997), 6-OHDA lesions of the dorsal striatum had no effect on accuracy but lengthened latencies of responding correctly and increased the numbers of omissions and perseverative errors (Baunez and Robbins, 1999). 

---