Robotic Visual Servoing for Tracking Structureless Moving Parts

The European Union has committed to achieving the 1.5-degree Celsius limit set by the Paris Climate Agreement, emphasizing the necessity for substantial energy savings across various sectors, particularly in industry. This ambitious goal requires a delicate balance between reducing energy consumption and maintaining economic strength. Concurrently, there is a noticeable shift from mass production to mass customization, driven by the need for more flexible and adaptable manufacturing processes. Cognitive robotics emerges as a pivotal technology in this context, offering enhanced capabilities for flexible production lines. One significant advantage is the ability to assemble moving parts which optimizes cycle times and therefore contributes to energy savings. This paper presents a novel approach to object tracking using NC robot-based position control, demonstrating its potential to support the industry's transition towards more sustainable and efficient manufacturing practices. The image processing algorithm, which consists of a fused methodology merging an iterative 2D image registration and stereo vision-based point cloud registration, demonstrates the ability to track an object with a mean L2-error of 3.5 mm in a dynamic environment.