SHERLOCK: A Coached Practice Environment for an Electronics Troubleshooting Job

Abstract

1 "Sherlock Is a computer-based sp aortd a environment for a complex troubleshooting job in the Air Force. This chapter describes the training problem for which Sherlock was developed, the pinciples behind its development, and its ImplemaIo. The training problem is severe and representative of a common problem In our high-technology society. People who will fiN a position for a brief period (four years or lees for many in Oi Air Force Job) carry out a set of routine tasks which are well supported by technology. However, periodically, a breakdown occurs: a novel situation requiring sophisticated problem solving for which Itle support Is available. In many cases, such problems must be referred to an expert, but such expertise Is difficult to acquire. Semlautomated, routinized jobs do not afford sufficient .opportunities for complex problem-solving skills to develop, so their Incumbents lack the skill required to handle novel problems. Sherlock is an environrthent in which this missing skill can be acquired for a specific troubleshooting job.-Thu ig b. Sherlock was developed to raise the level of F-IS Manual Avionics Test Station Technicians' troUhehooting knowledge. These technicians repair electronic modules that have been removed from F-15 aircraft because of suspected malfunction. In their daily work, they follow detailed written bowlestool, procedures (part of their Technical Orders) and use a testatatio. The test station Is a age (40 iP) system of electronic equipment to which the module being diagnosed can be attachedw..,' By setting various switches on the front panels of the test station, the airman can quickly perform tests on the module. When a test In the prescribed test routine Isolates the malfunction, the Technical Orders suggest an appropriate repair for the module. A serious problem arises when the test station Itself has a malfunction. Now, Instead of following fixed procedures from a check sheet and making use of the substantial technological support of the test station, the airman i much more on hIs2 own and must engage In complex heuristic problem solving. Because they may be wordng in the field far from help, these technicians have to be self-sufficient: they have to be "4e to repair their equipment. However, because the test station breaks down only about once a month (and even then may be diagnosed and repaired by an expert to minimize downtime), a firstterm airman (someone in his first four years of duty) gets few opportunities to learn this most difficult skill on the job. The technical training airmen receive before reporting for work is oriented toward the conceplual knowledge and skill needed for the routine part of the work-which Is all there is for perhaps 95% of the lime. Yet this leaves the Air Force with a dilemma: its training supposes that test station repair will be learned on the job, but the job doesn't provide the relevant practice opportunities. Sherlock

An Interview Reflection on “Intelligent Tutoring Goes to School in the Big City”

“…So, it was definitely not the first experimental trial in schools. There were also some other intelligent tutors being trialed in real settings including middle schools (e.g., Lester et al 1997), colleges (e.g., Mitrovic 1998), medical schools (e.g., Eliot and Woolf 1996), and military training settings (e.g., Lesgold et al 1988). I think it is safe to say that, at the time, is was the largest such study and the first I know of that involved a substantial integration of an intelligent tutoring system into a full-year K12 curriculum.…”

Automated Tutoring Environment for Command (ATEC): Using an Intelligent Tutor to Model Expert Mentor Interactions

“…or where the language used is highly restricted (e.g., command and control). The major successes of the knowledge-tracing variant of intent inference in tutoring (e.g., Anderson et al, 1995;Van Lehn, 1990;Lesgold, Lajoie, Bunzo, & Eggan, 1992) have all come in such cases. On the other hand, the CHI Systems research team has substantial success with the situational approach (using the COGNET/iGEN technology) in both tutoring and performance support domains (Zachary, Ryder, Santarelli, & Weiland, 2000).…”

AI in medical education—another grand challenge for medical informatics

“…avionics troubleshooting. One example is SHERLOCK which provides a safe and efficient environment where avionics technicians can learn and practice troubleshooting procedures for dealing with problems associated with an F-15 manual avionics test station [63,66]. This ILE was developed in response to the observation that first term airmen had difficulty applying their schooled knowledge of basic electronics to the on-the-job troubleshooting of complex equipment.…”

“…Airmen were empowered to test their hypotheses and were given assistance when needed. Extensive cognitive task analyses were performed prior to the development of SHERLOCK in an attempt to represent the knowledge and problem-solving skills that learners would need to succeed [66]. A cognitive apprenticeship model [28,63] was used to design SHERLOCK, whereby novices were provided with individualized feedback that would help them overcome learning impasses in the context of problem solving.…”

“…Cognitive science offers several methodological tools that can serve to elicit the types of knowledge needed to learn in specific situations. One technique for doing so is a cognitive task analysis of the domain in question [66]. A cognitive task analysis consists of several steps where the goal is to identify the most difficult aspects of job performance in a specific domain and then study what differentiates the most from the least competent individuals in terms of the types of goal, prerequisite knowledge, actions, results, and interpretation involved in solving the task in question.…”