Application of Acoustic Sensors to Robotic Seam Tracking

Abstract
Robotic seam tracking systems require three-dimensional information about the seam geometry. The applicability of a nontactile acoustic sensing system, which provides range information about the distance between a sensor and the workpiece, is evaluated for the three-dimensional surface characterization task. An algorithm for executing a two-dimensional surface height sampling of an unknown workpiece is developed and evaluated through digital simulation and hardware experimentation. The algorithm incorporates the limitations of the sensor's aperture size and uses only the range information provided by the sensing system. A recovery strategy is included if the sensor's aperture has been exceeded and range information is not available. The sampling procedure is found to be applicable for piecewise constant surfaces and surfaces with slight curvature. The effectiveness of the acoustic sensing system's surface characterization can be enhanced by using a priori knowledge of the workpiece. The sensing system appears to be promising for such industrial operations as part location and inspection.