Subscribe Button 1

New Era in Weld Seam Inspection Launched

Mapvision has launched a new multi-camera inspection application for weld seam inspection. When it comes to welded assemblies, the quality of welding is arguably as important as the dimensions of the final part. According to Mapvision, until now, no comprehensive solution existed for checking the existence and quality of the weld seams. 

Weld Seam Inspection Challenges

All weld seam inspection methods in use today have obvious downsides. 

Manual checking is expensive and unreliable. Even the best inspection specialist will get tired after repeating the same task for hours on end, and the subjectiveness of the classification cannot be avoided. 

Scanner systems are limited by their speed. Full inspection of a part, such as a subframe, requires several robots using both serial and parallel configurations. Scanning systems are also slow to learn. Needing, typically, six months to get a scanner-based weld seam inspection system into operational status is no exception. The upkeep of the systems is tedious, too, and the systems are often simply bypassed when they cannot keep up with the changes in production. 

Weld seam inspection systems based on voltage/current monitoring do not offer the required level of reliability as they do not inspect the final seam, only the signal produced when creating it. Such a system will not be able to inspect the seam after it has been completed. 

Unified Inspection System for Dimensional Measurements and Weld Seam Inspection 

Mapvision Multi-camera technology in one coordinate system enables a visual memory of production

Instead of adding a new inspection station to the production line, Mapvision decided to approach the weld seam inspection dilemma differently: wouldn’t it be great if weld seam inspection could be done reliably and with full coverage in one step at the same time with dimensional inspection? 

The Mapvision in-line quality inspection solutions are based on a unique multi-camera technology which performs dimensional checks by taking measurements based on the generated images. Being image-based gives Mapvision endless opportunities to upgrade the system with software. The software uses artificial intelligence, machine learning, deep learning and neural networks to learn and monitor weld quality.

The inspection station is already there doing the dimensional checking with a low-footprint cabin. With the same system, Mapvision can now also take all the weld seam images needed, and then allow the neural networks do their thing for weld seam inspection.  

If you order your Mapvision solution for dimensional measurements with a hardware setup that supports Weld Seam Inspection, the weld seam inspection functionality can be added at any time as a software upgrade saving a substantial amount of floor space. 

Variations in Applying Weld Seam Inspection 

Weld seam inspection in the industry today can mean many things. Companies approach weld seam quality control in diverse ways and there is a lot of variation in their requirements. Some customers, for example, allow sampling inspection while others already require 100% inspection. In some cases, controlling every aspect of the weld seam is on the wish list. In others, only a few criteria of the seams are to be verified. Full automation is the goal for some companies, but many still rely on the human eye for verification. 

While the trend clearly is towards more weld seam quality control and 100% weld seam inspection, the wide spectrum of requirements means that flexibility of the weld seam inspection system is important. 

Given that the Mapvision Weld Seam Inspection solution is fully digital and a software only solution, allowing Mapvision to exactly offer the necessary flexibility.  

The new weld seam inspection application of the Mapvision multi-camera application has been in development over a period of time and has been in beta testing with selected customers for over a year. The initial launch version will provide what is called Level 1 Weld Seam Inspection for Anomaly Detection.  

For more information: