The single-switch controlled communication aid for non-vocal handicapped has been developed, using a personal computer NEC PC 6601 with a micro floppy disk, a stick/grip force sensitive switch and a color character display. The system permits a person who can activate the control switch to select any of the Japanese syllabary on the 10 rows×5 columns chart, by means of automatically row-column scanning.
Applicability of this communnication system was investigated on a severely cerebral palsy patient unable to operate a keyboard, through the examination of vocabulary, recitation and composition tasks, during a 3-month trial period. The following observations were made:
(1) The system enabled the patient to write many words and sentences as he wish via the character display.
(2) The scanning rate and the switch unit were clarified which maximize the efficiency of the operation of the system.
(3) The language skills of the patient could be grammatically and syntactically evaluated.
The results of the present study have demonstrated that the performance of the system is satisfactory expect in the case of communicating between a non-vocal handicapped and his environment.

View full abstract

Human posture is characterized by upright standing. Standing erect is one of the most popular and basic working posture. It is important the postural regulation system is analyzed. Stabilograms (Sbg), which reflects the displacement of foot pressure in a standing posture, has been analyzed using time series analysis methods. It is regarded as the realizations of linear stochastic processes. This paper describes experiments with healthy 10 males and 12 females subjects, holding and backing 5 kinds (0, 3, 6, 9, 12kg) of weight loads, stand on a force platform. Sbg is recorded 2 minutes for each load. The enclosed area, total length and the numbre of turning points of the displacement of the center of foot pressure are calculated from the Sbg. There was tendency for the area, the velocity and the length of displacement of center of foot pressure to increase with the weight of the load. But the number of turning points was in inverse proportion to the weight of loads. It was found that some parameters calculating from Sbg could evaluate control mechanism in standing posture.

View full abstract

This paper describes three experiments of reducing the normally very complicated speech wave to simple geometrical form without impairing intelligibility, and also describes the results of articulation tests of simplified speech. The simplification of the speech wave is performed by slicing the original input wave using comparator (s).
I. In this experiment the slicing level is varied using a single pair of comparators. Optimum noise reduction is achieved by a slice level that is higher than the noise amplitude. The articulation score of the sliced speech at various levels that -40.5dB are optimum for the slice level in order to keep the sufficient intelligibility and to avoid the noise.
II. In this experiment, two pairs of comparators are used to make the complex slicing speech waveform. The slice level of the first pair of comparators is set at the optimum level obtained by experiment I, and that of the second pair of comparators is varied to find the optimum level. The outputs of both pairs of comparators are combined together by an adder to give the test output, the combination ratio of each output being kept at a predetermined value. The articulation score of the complex slicing speech wave at various levels of the second pair of comparators that the range of -14.5dB to -11.0dB is optimum for complex slicing speech wave in order to keep the sufficient intelligibility.
III. In this experiment, also complex slicing speech waveform is used and the circuit and the slicing levels are the same as experiment II, but the combinations ratio of each comparator output is varied. It becomes clear that the intelligibility of this experiment is better than I over the wide range of combinations ratio, but the ratio itself is not so critical to the intelligibility.

View full abstract

The concerns of this paper are to identify the factors leading to success of robot introduction, to make clear the aspects of implementation of safety measures, and to show several cases of socio-technical system with regard to robotics, in minor enterprises.
A mail survey of 106 questions is executed for 159 companies and the answers are obtained from 30 companies, to which another actual investigation is added. The main factors obtained by a multivariate analysis are as follows; (1) productivity improvement as a purpose of robot introduction, (2) installation of robotics by a joint team of engineering staffs and line workers, and (3) a perspective of large demands of products. Most small enterprises have no special implementation of safety measures. Finally, we find four cases of the socio-technical system related to robot introduction and this system is identified as an important factor leading to success of robotics.

View full abstract

This paper deals with an automatically diagnostic models or an intelligent CAI for student's misconception of arithmetic skills. Underlying Brown and Burton's “Buggy model”, we assume that the students don't make errors in arithmetic calculation, but they perform misconceptual procedure of calculation. This research is a basically expanded version of Buggy model to multiplication and division, and implemented by LISP language based on a procedural network. The present model is able to select the appropriate arithmetic tasks for student's misconception according to the procedural diagnosis.

View full abstract

The purpose of this paper is to analyze how the beginners debug in programming. The cognitive psychological approach is applied to clarify the strategies used by the beginners to understand “procedural” information processing programmed in the artificial language (program). The general characters of the strategy are observed in the process followed by the beginners from the syntactic understanding to the semantic understanding. It is shown that this process is divided into several distinctive stages through analysis of errors, by checking to what extent the beginners have understood the program beforehand and protocol analysis. Whether they can successfully proceed from the syntactic understanding to the semantic understanding is presumed to depend mainly on how far they can understand information—“conceptually driven”—in the stage of cognition, and can image well—“data driven”—using data.

View full abstract