Obtaining accurate measurement using redundant sensors

Abstract

Conventional wisdom suggests to accomplish accurate measurement, the sensors used must have high precision and excellent dynamic range. This generally results in sensor systems that are complex, costly, and often sensitive to environmental factors. The research conducted explores shifting the current sensor system design paradigm by developing a pose (position and orientation) measurement system using inexpensive redundant (same type and more than the minimum required) sensors, each with limited capability. Using acoustic transmitters and microphones as the sensor elements a robust, inexpensive pose measurement system for making measurements in a factory environment is proposed. A novel triangulation algorithm is developed to solve for pose, using sound time of flight measurements from 3 transmitters on a mobile stylus to fixed microphones, that is robust with respect to environmental speed of sound variations. In addition, a closed-loop (does not require external sensors) calibration algorithm is derived to determine the positions of the fixed microphones. To predict the accuracy of this system an error analysis is conducted using computer simulations. Results suggest one can construct an inexpensive redundant sensor pose measurement system with enough accuracy and range for a wide variety of industrial and commercial applications. The redundant sensor concept is general in nature and has the potential of being applied to measurement problems other than pose measurement. Its application to pose measurement is valuable because the difficulties in applying it to other measurement problems are illustrated.