The project aims to develop a fully functional mobile system for determining the characteristic geometric dimensions of precision forgings using modern vision and machine learning techniques.
The designed approach is based on a mobile measurement system that allows reproducible image capture of the forgings from different perspectives on a specially prepared calibration table. The project analyses different variants of camera placement, lighting, and various approaches to designing and 3D printing calibration tables. The acquired images are transferred to a specially designed database for further processing using the IIoT (Industrial Internet of Things) approach. A detailed analysis of the images is carried out, including the required depth corrections using computational geometry methods, and the accuracy is further enhanced by advanced machine-learning models.
Equipping the workstation with a robotic arm allows quick delivery of the forgings to the system’s work zone and efficient, accurate determination of their dimensions. A dedicated, user-friendly application lets the operator supervise the dimensioning process, check how many products from a given batch have failed quality control, and generate reports with technical drawings containing all key dimensions — as well as review and analyse previous data.
Thanks to the time-efficient dimensioning, it will be possible to significantly increase the volume of production being checked and the safety factor of the final products, while optimising time and labour costs.
