Vision-based robotic system for picking and inspection of small automotive components

Abstract

Abstract-The use of vision systems for industrial robot guidance and quality control becomes much harder when the manufactured products and their components are small and possess reflective surface. To assure an effective automated visual inspection of such components, novel solutions are required, able to perform more advanced image analysis and tackle noise and uncertainty. This paper proposes a concept of multi-camera/multi-pose inspection station for star washers inspection, and presents the first results of a functional prototype implementation of it in a robotic cell. The processes of vision-guided part picking from a flexible feeder and closerange inspection in a dedicated rig are described. Solutions for the vision-based tasks of parts identification, machine learningbased classification, circular objects image analysis and star washer teeth segmentation are presented, and further directions are outlined.

Application of Machine Learning Techniques to Predict the Mechanical Properties of Polyamide 2200 (PA12) in Additive Manufacturing

“…When it comes to the prediction of mechanical properties by using traditional methods like linear regression, it can be seen from Tables 5-7 that prediction accuracy for linear regression (except for Nominal Stress and Elongation at break in Angle orientation) is low, and therefore more advanced techniques need to be used. Since machine learning techniques have also been widely used in traditional manufacturing [32][33][34][35], these methods were applied in this study to develop non-linear prediction models.…”

“…Therefore, another approach is required for obtaining robust predictions that will be less sensitive to unknown noise. Since there are examples of successful application of machine learning techniques in traditional manufacturing [32][33][34][35], these methods could be used to predict mechanical properties with limited information about the AM process itself.…”

“…Data collected using a different design of experiments can be used together (joined in one data set) for model development, and with every next use of AM machines, collected data contributes to modification of previously developed models, resulting in higher prediction accuracy and better precision for the future use of AM. Nowadays, these methods are used in different field of studies [29][30][31], and have also been widely used in traditional manufacturing [32][33][34][35].…”

Event-driven industrial robot control architecture for the Adept V+ platform

“…Problem context e functionality of the proposed system is demonstrated on a robotic application of handling a small part for detailed vision-based quality inspection. Drawing from the previous work on picking and inspection of small automotive components (Semeniuta et al, 2016), the described setup is aimed at moving the part from the pick pose ξ pick to the inspection-start pose ξ is in front of a Prosilica GC1020C camera with a 35 mm Fujinon HF35HA-1B lens, with the subsequent sequence of tool rotations while keeping the part in focus of the camera. On each rotation increment, an image of the camera is requested by the MasterControlNode.…”

Peer Review #2 of "Event-driven industrial robot control architecture for the Adept V+ platform (v0.1)"

“…Problem context e functionality of the proposed system is demonstrated on a robotic application of handling a small part for a detailed vision-based quality inspection. Drawing from the previous work on picking and inspection of small automotive components (Semeniuta et al, 2016), the described setup is aimed at moving the part from the pick pose ξ pick to the inspection-start pose ξ is in front of a Prosilica GC1020C camera with a 35 mm Fujinon HF35HA-1B lens, with the subsequent sequence of tool rotations while keeping the part in focus of the camera. On each rotation increment, an image from the camera is requested by the MasterControlNode.…”