| The investigator developed an non-invasive non-contact method for detecting on-line industrial robot position errors. The method uses a low-cost sensor to detect single-axis position errors. The sensor, composed of a low-cost microwave Doppler radar detector and a low-pass filter, converts robot motion into electronic signals that are A/D converted and processed using a computer. Computer processing reduces captured signals into root-sum-of-squares error measures, with respect to a mean sensor calibration signal. Root-sum-of-squares error measures are compared to a threshold value that indicates, statistically, a 99.7% probability that an on-line position error has occurred. The threshold value can be adjusted to meet different application needs. For the... |
