Smart Process Observer for Crane Automation

A method of automated, noncontact analysis is presented, which scans a process crane’s work area fully extrinsically with special 3D LiDAR sensors and analyzes its motion dynamics in real time. Rule- and AI-based algorithms that interpret high-quality point cloud scans have been developed, thus making it possible to evaluate a process crane’s specific handling operations reliably. Existing CAD models of the crane assemblies are automatically fitted into the point cloud for the central process of fused data analysis. The workflow starts with the localization of the loading beam’s cables to estimate its initial orientation and position. The CAD models of the lifting beam and all other lifting system components are successively fitted into the point cloud exactly with the aid of local registration. Swivel joint design constraints are factored into the assessment. The lifting operation is displayed in a VR model automatically receiving all component orientations and positions and the load every second. The crane operator can view the current situation from defined perspectives and additionally receives information on crane component position, spacing and the load, which is needed to control the crane.