5 Keys to Automated Quality Inspection

During my time in Germany for Hannover Fair 2015, I had the opportunity to visit BMW’s motorcycle production facility in Berlin as a guest of Siemens. All BMW motorcycles sold worldwide are assembled in this plant, which also happens to be BMW’s oldest manufacturing facility. Motorcycle components were first built there in 1949, with full motorcycle assembly starting in 1969. Today, the facility can produce up to 650 motorcycles daily.

While this BMW facility is impressive for a variety of reasons—from its clean and ordered production operations to the outright cool factor that one can't help but sense while surrounded by such an impressive array of beautifully designed motorcycles (whether you ride them or not)—the segments of the tour that struck me most were the automated quality inspection stations.

Though quality inspection may sound dull when talking about a production facility that creates a product like the R1200GS (see photo inset), when you watch a robot operating autonomously to conduct quality inspections of connecting/piston rods that drive the crankshafts of BMW’s motorcycle engines, it’s difficult not to stare. Inside the cage, this one armed robot (sorry, no pictures were allowed during the tour) picks a newly finished connecting/piston rod from a bin, brings the rod over to an inspection instrument—which performs quality measurements on the rods to a .5 micrometer level of precision—and then deposits the rod (if it passed inspection) into a slot for use in production or into another slot that sends the part back through the system for scrap or rework.

All components either produced in the BMW facility or received there from other BMW production facilities and third party suppliers are quality control tested to ensure their ability to fulfill BMW’s requirements. These requirements combine scientific measurement with “feel”, as determined by BMW motorcycle experts who, during the test run of every motorcycle produced at the facility, are said to be able to detect by feel if something is not quite right about the overall quality of the finished motorcycle.

Stressing the importance of quality to BMW’s production process, our tour guide pointed out why even a 99 percent quality level is not acceptable. A 99 percent quality acceptance level equates to some 80 parts in a BMW car potentially being problematic, or 40 parts in a motorcycle.

Having seen the leading edge of automated quality inspection at the BMW facility, I started wondering what the most critical factors are when it comes to automating a quality inspection process. This led to a discussion with Wayne Cantrell, Siemens Factory Automation systems engineer in Johnson City, Tenn., and Stefan Werner, marketing manager for Siemens Factory Automation in Norcross, Ga.

Out of that discussion came five key points of automated quality inspection:

• The need for accurate boundary limits and reference data for every part being inspected. “Any discussion about automating quality inspections with your system integrator or automation supplier needs to start with accurate data about the part and/or product specifications that need to be met,” said Cantrell.
• Tracking and tracing components and their pass/fail of quality parameters, along with the ability to retrieve such data on demand, is increasingly important for industries of all kinds—but especially for the pharmaceutical, food and beverage and transportation industries. Cantrell and Werner both pointed to the increasing use of MES (manufacturing execution systems) in QC operations to approve/certify quality processes and provide dynamic, real-time overviews of production quality.
• Proper instrumentation. Cantrell said that when working with customers to review existing quality inspection operations, he often finds that there is not enough of the type of instrumentation needed to conduct quality inspection at the level sought by the manufacturer—which can result in subpar components passing through the inspection process.
• Controlling material movement. This relates to the ability to have interlocks enacted by the automated quality system to prevent movement of material until a part passes inspection or has specific actions performed on it first. “This enforces proper production flow for a product or component,” said Werner. Cantrell added that he has seen automated QC systems make use of an MES to send instructions down to an industrial PC positioned near an operator to ensure that the operators manually input data if that data is not able to be automatically collected due to the production process. “This is key in traditional manual operations where a part can go through several processes before being caught at QC,” Cantrell said. “Plus you can put this into place at several points along the production process to check quality data and provide immediate status of a part, component or product.”
• Communication with and interface to the quality control (QC) control system. The interface between the QC automation system and the controller containing the setpoints, as well as the network over which the system components communicate, are the most critical points of the five listed here, according to Cantrell. The interface and communication aspects are not just about part identification, but operating parameters—such as different torque specifications, he said. “To communicate those setpoint commands and receive verification, there needs to be two-way protocols in place. This can be done through serial or Ethernet communications, as determined by the hardware interface and protocol the equipment supports.”