The skill transfer of Air Traffic Controllers is a quite important issue to maintain higher level of safety and efficiency in aviation, as ATCOs have central roles in the air traffic control system. The present research has analyzed skill of ATCOs from the aspects of strategic control of air traffic flow to prevent potential collision of aircrafts and also to reduce their cognitive workload in the dynamic and multi-task condition of air traffic control. Based on the analysis, a function which can visualize the difference of task performance has been implemented into the Air Traffic Controller Cognitive Simulation (ATC-CS) for educational purpose. Using this function, the effect of ATCO's control on air traffic flow has been successfully visualized, which helps trainee to understand the differences of the consequences of the different strategy. This result has strongly implied that ATC-CS equipped with performance visualization function can be utilized as a supporting tool for education of ATCO trainees.

It is important for Japanese industries to success the skill of veteran workers to young workers. The authors are studying the relation between the skill of plant operation and eye fixation points on the symbols of control panel of skilled operators. This study experimentally investigates the relation between safety variables and the variables corresponding to the symbols on control panel that skilled operators fix their eyes. The DURESS that is an artificial simple plant is used in this study as an example of plants. Three types of DURESS are designed by changing safety variable. The eye fixation rates of the symbols on control panel of skilled operators for each DURESS are compared in the operation from an initial condition to a specified goal condition. The results show that the eye fixation rates are high on the symbols corresponding to safety variables and the rates are even higher than those for the variables that need fine control in operation.

This study first investigates experimentally the tendency of eye fixation points of skilled operators on control panel in the operations of DURESS that is an artificial simple plant and confirms that skilled operators tend to fix the symbols corresponding to some specific plant variables and the eye fixation changes corresponding to plant conditions. Then, this study investigates the effect of the usage of the tendency in the training of DURESS operation. This study evaluates the improvement speed of skill level and the ability of managing inexperienced operation conditions depending on the method for the usage of the tendency. The results show that the development speed of operation skill increases by the usage of the tendency. The results also show that the managing ability of inexperienced operation conditions can develop by fewer training operations when the tendency is used with instructing novice operators to consider the reason of the tendency in training operations.

"Bunraku" is Japanese traditional theater that consists of three elements; music by a shamisen player, story by a narrator "Tayu" and Bunraku puppet drama. Each Bunraku puppet is manipulated by three puppeteers. Nonverbal information that one of three puppeteers "Omo-zukai" (the main puppeteer)use, make myriad and realistic movement of Bunraku puppet possible. The information is called "Zu". It is estimated that the nonverbal information "Zu" contain the factor to enable realistic cooperative manipulation by multiple persons. The aim of this research is to determine the mechanism of cooperative manipulation in Bunraku. In this paper, feature of Bunraku puppet movement was extracted by focusing attention on "Hodo" (a part of "Zu"). And relationship of Extracted feature to movement of "Hidari-zukai" (the left puppeteer) that follows movement of "Omo-zukai", was analyzed. According to the result, it has been found that movement of "Hidari-zukai" is composed of three phases.

Alarm timing for a forward collision warning system may play an important role in system effectiveness. It is necessary to determine appropriate alarm timing by considering driver's response to not only true alarms but also missed alarms in order to cope with both increased system effectiveness and decreased over-reliance on the system. By using a driving simulator, two kinds of alarm timing were compared to investigate braking behaviour toward alarms and how drivers respond to missed alarms regarding different alarm timings:(1)an alarm was given based on ordinary braking behaviour for the individual; alarm timing T,(2)an alarm was given by using an particular alarm trigger logic (Stopping Distance Algorithm)as a common timing for all drivers; alarm timing S. Alarm timing S was earlier than alarm timing Tin this study. The results showed that compared to alarm timing T, alarm timing S induced earlier braking behaviour independent on degree to which an imminent collision was critical. However, effects of a missed alarm on braking behaviour may be mitigated by alarm timing T, compared to alarm timing S. Moreover, it is possible that effect of missed alarm on driver's trust may vary according to alarm timings and the number of experience of missed alarms.

If a computer user could refer the situation of his/her own past task later, he/she would be able to recall procedures of that task. Based on these information, he/she can hit upon the useful idea for the on going task. In order to refer such situation, he/she can explore the situation in the log, sequence of situations, which consists of his/her activities on the computer. We call this log "task history". However, because he/she might not always recall the proper keywords about the target situation explicitly, he/she might waste huge effort to explore the task history. In this paper, we propose THES (Task History Explore System) for supporting to explore the task history. THES shows a user cues step by step to recall the detail about the target situation. THES shows two kinds of cues, "activity cues "that prompts to recall the memory about the activities performed in the target situation and "time cues" that prompts to recall the memory about the time of the target situation. As a result of a 5-week experiment, THES was effective for the task history exploration.

Speech recognition using electromyogram(EMG)signals measured from facial muscles have been already reported. However, it is not revealed whether speech recognition using EMG signals from neck muscles can be achieved. To verify this, we conducted experiments to recognize five Japanese vowels using EMG signals from the neck for six healthy subjects. In the experiments, we used five classifiers: k-NN, Bayes rules, neural networks, support vector machines, and hidden Markov models, and compared the recognition rate using EMG signals from the neck with the one from the face. As the results, the recognition method based on the HMM showed the highest recognition rate. When using EMG signals from the face, the recognition rate averaged across all subjects was 93.3%. In contrast, the recognition rate was 83.0% using EMG signals from the neck. When we performed channel selection, the recognition rate was 84.5%, and four of six subjects showed more than 80%. Though the recognition rate using EMG signals from the neck was lower than the one from the face, it was indicated that vowel recognition using the EMG signals from the neck muscles was feasible.