Automatic path planning for high performance measurement by laser plane sensors

Abstract

The Fourth Industrial Revolution (4IR) has led to a wide development of laser sensors and their use in inspection processes to achieve high performance is becoming more and more common. High measurement performance implies high measurement quality of surfaces in a short time. Generally, the quality is evaluated on the basis of indicators such as noise and trueness. Recently, it has been found that these indicators are mostly affected by the scanning strategy defined by two main parameters: the scanning angle and distance. Indeed, the path planning of laser sensors ensuring a high level of quality is based on the definition of the scanning strategy allowing scanning well-known surfaces with reduced noise and trueness. In this paper, an original approach for the generation of automatic path planning is proposed. The originality lies in the high performance obtained by optimizing the quality and the measurement time based on the optimal path planning. First, the laser sensor is qualified in order to evaluate the quality indicators as a function of the scanning distance and angle. Configurations that provide better quality are determined. By analyzing all laser sensor orientations held by the CMM's rotating head, a set of minimum orientations allowing high quality measurement is determined based on an optimization algorithm regarding quality, accessibility, visibility and collision detection criteria. The laser sensor path is based on the determination of the path levels in each sensor orientation. An original method is proposed to generate the optimal levels guaranteeing a high quality in terms of scanning distance. The laser sensor crossing points are then calculated in each level. An improvement of the method is proposed through the connection of levels using connection lines. Finally, collisions are studied and fixed by adding translation points. The proposed approach is applied to a part defined by its CAD model. The path is generated and executed on the CMM with a laser sensor. The measurement performance is compared to previous approaches in the literature. The methodology allows an adapted planning and can be used for automated dimensional inspection.