Developing a Vision System for in-line inspection of wooden flooring
A leading provider of wooden flooring and furnishing technology turned to Data Ductus for a new innovative vision system, which has been fully integrated into their production facilities.
Quality Control staff using the company’s systems were unable to consistently identify small defects in flooring during high-speed production, and wastage was too high.
Boards are scanned by a 3D scanner on a conveyer belt. A newly developed vision platform collects, analyzes and visualizes the data, and informs machine operators about defects. All data is categorized and centrally stored for later use.
How we did it
We worked closely with the customer to develop the initial machine vision solution. Today, agile workflows are used to seamlessly incorporate additional capabilities and intelligence into the solution which has become a fully integrated part of their production.
Wastage has been greatly reduced and manual inspection stations have been minimized. The solution can be incorporated into the production practice and implemented at different manufacturing plants. Additionally, systematic production problems can now be identified and resolved.
About the client
The customer is a leading provider of wooden flooring and furnishing technology.
The company wanted a state-of-the-art production line, with minimum wastage and resource requirements. During manufacturing, multiple floor tiles are produced on a 2 x 2.4 meter hardened wood composite core, with a thinly applied veneer layer, before being cut to size. Heat and high pressure can cause blistering, bending, and other damage to multiple or single tiles during the production process.
Manual inspection of the boards is challenging due to the size and speed that the boards are produced. It’s virtually impossible for a person to consistently identify small but crucial defects with just a few seconds to inspect them, especially as a defect can be smaller than one square centimeter and protrude less than 0.5 millimeters from the wooden surface. To consistently identify such defects, an effective machine vision system is required to rapidly and accurately inspect the boards.
Inspecting the details with 3D scanning
Since the flooring has extremely low tolerances, the choice of hardware for data collection was critical. Minute height differences – as small as 0.03 millimeters on the veneer edges – had to be recorded and analyzed. The boards also had to be inspected while traveling on a conveyor belt. Therefore, a 3D scanning solution was deemed most appropriate.
A new vision solution
A new vision solution was developed to collect, analyze and store the data generated by the 3D board scanner. This provides a range of benefits. For instance, each board is marked with a unique identifier, to which the stored data is coupled. With this data, single tiles within a core panel can be automatically rejected if they have a defect, at a later stage in production. Previously, the whole panel was rejected, now, single tiles are discarded once the panel is cut into tiles. Additionally, machine operators are informed of any defects and can make an instant decision on how to proceed.
All measurement data is stored in a central database and is available for post analysis. Together with other process data, it can be used to detect and resolve systematic problems in production.
A collaborative partnership
What began as a vision system project, has now evolved into a long-term partnership. New measurements and features have been added and the solution has been fully integrated into the production process. Collaborative agile workflows enable the scanner to be rapidly adapted for parallel projects while small on-the-fly improvements can be quickly implemented.
A future-proof vision system
The new vision system provides objective inspection of products, with high traceability in the form of quantification of the measurements stored in an ordered and accessible format. Automated sorting of defect tiles means wastage is minimized. Furthermore, the need for manual inspection has been greatly reduced, freeing up resources in manufacturing. Finally, the ability to identify issues in production, provides an innovative tool for identifying production improvements and alerting staff about potential upcoming service requirements.