Abstract

A method and system for automated training of a subject in manual skills, wherein body motion is measured so as to derive a set of motion variables, the motion variables are compared to a reference motion to derive a motion error value, the motion error value is compared to a threshold error value corresponding inversely to the proficiency level of the subject, and a training signal is provided to the subject whenever the motion error value exceeds the threshold error value. One embodiment of the inventive system comprises a plurality of lamps mounted on the subject, first and second cameras for scanning the subject in accordance with first and second reference systems, respectively, a selection circuit for selecting first and second motion data derived by the first and second cameras, a processor for processing the selected first and second motion data to determine a variation between the performance of the subject and a performance measure standard, and a training signal generator for providing a training signal to the subject whenever the variation exceeds a preset threshold level. In an analog version of this embodiment, motion data relative to each respective reference system is derived in sequence in analog form; in a digital version of this embodiment, motion data is derived concurrently, and in digital form, for each reference system. In a further embodiment of the inventive system, a plurality of rotating cylinders having light sources therein direct light toward the subject, on which a plurality of photosensors are mounted. Further inventive features include a self-calibration technique.

Other References

• Edward M. Connelly et al., Computer-Aided Generation of Performance Measures for Man-Machine Systems, Oct. 1974
• E. M. Connelly et al., Continuous Performance Measurement in Flight Systems, May 1975
• Clyde N. Herrick, Color Television: Theory and Servicing, pp. 122-127