![](http://9vy.6up85.com/__data/assets/image/0023/112928/varieties/big-tile.jpg)
博客
The launch of a landmark EPC project in Texas
Precision and traceability requirements in automotive manufacturing often require that each individual part be assigned a part unique number (PUN). The marking / labelling process can be challenging, time consuming, and in some cases not practical. In addition to the process challenges the marking equipment is costly.
Wood’s Machine Vision team out of Michigan is collaborating with a leading global automotive manufacturer, and software technology provider Alitheon, on an in-plant pilot for a cutting-edge AI-enabled vision system that will respond to the traceability challenge in automotive powertrain assembly.
Using a combination of powerful AI and machine vision methods, an object is registered by creating a FeaturePrint™ model and storing it in the cloud or local database. Registered objects can then be identified by creating another FeaturePrint™ model and comparing it to the database of registered objects. If there is a match, the object can be identified in seconds as the original and its associated metadata accessed. One of the most unique features of this technology is that an object can be identified even if 90% of its original surface has been removed.
The team’s innovation and ingenuity has resulted in a solution that eliminates the previous requirement of marking or labeling each individual part has true potential to be a game-changer in the manufacturing industry. By optimizing this process, substantial cost-savings can be realized for our clients – saving valuable time through the reduction in steps required and removing the need for expensive marking equipment.