The goal of ever shorter cycle times in often difficult production environments requires highly efficient and robust testing of paint surfaces. The demand to achieve the highest quality and efficiency has led to the pioneering role that the automotive industry plays in the automation of industrial production. Vehicle production is a role model for many industries and drives the use of new technologies like no other. This market-driven pressure to innovate gives rise to highly efficient solutions - including in the area of quality assurance. For example, intelligent robot-guided production and inspection makes high-end automation technology for paint inspection possible.

Complete systems available today - such as CarPaintVision (CPV) from Isra Vision - combine robotics and surface inspection. Thanks to the quick and easy integration of new vehicle types and automatic offline robot path generation, they can be used flexibly. Whether on-the-fly, with continuous conveyance of the vehicle or in a cycle-based stop & go process, the entire relevant surface is inspected in-line.

Fully automatic in-line paint inspection

CPV provides an objective and consistent evaluation of paint surface quality that can be customized. It offers cost benefits compared to manual methods and supports process improvement through reliable statistical data. Isra Vision uses the Paintscan hybrid sensor for this solution, which combines two surface inspection methods. This reliably detects all topographical and non-topographical paint defects. Integrated evaluation units ensure fast surface inspection times and reduce maintenance costs. Data processing takes place in the process cycle and defect information is available immediately. The optional final marking of defects also makes reworking simple and time-saving.

Paintscan is a hybrid sensor that combines two surface inspection methods. It detects all topographical and non-topographical paint defects. © Isra Vision

The oversampling method ensures seamless paint inspection: a high scanning frequency and overlapping images in the direction of travel capture every point on the body multiple times. Redundant data with up to ten images ensures reliable inspection results with a detection rate of over 95 percent. All relevant defects - such as inclusions, craters, pigment defects, scratches, dents and dings - are detected and classified. The classification software has already been pre-trained and is the result of years of industry experience. Algorithms also enable self-learning defect detection and can therefore recognize line-specific defects. Thanks to its flexible application options, the system is particularly suitable for multi-model lines, as a collision-free robot inspection path is automatically planned on the basis of CAD models. The executing robots do not require an additional linear axis to be moved in line tracking mode.

Software and connectivity ensure future viability

The inspection data obtained enables sustainable and continuous production optimization. In conjunction with the Enterprise Production Management Intelligence solution ^EPROMI, the statistical data obtained is evaluated according to individual information requirements and displayed on standardized or individual dashboards.

Thanks to easy-to-use user software, configuration is quick and intuitive. Retrofitting into existing systems is also no problem. Compatibility with all common robot brands ensures optimum integration into the system environment, which is usually specified by the customer.

Thanks to its flexible application options, CPV is already ready for order-controlled multi-line production. The system thus supports the industry in its vision of production in which cars produced individually according to customer requirements automatically move to stations depending on the equipment line or special version - order-controlled, intelligently and efficiently.