YORK, PA. (July 18, 1 p.m. ET) — MKS Instruments Inc. and the
Dan Hazen, MKS product manager, explained the technology, called multivariate closed-loop control, during a presentation at Engel Medical Days, held June 15-16 in York, Pa. MKS offers it on its SenseLink quality control system.
The technology is called “multivariate” statistical process control, a form of closed-loop control. That's different from common fault-detection methods that are uni-variant, which means they only can control a few variables, by looking at one variable at a time, measured against pre-set upper and lower process variable limits, according to Dan Hazen, MKS product manager. The limits are adjusted as the process changes. That can result in false alarms. Also, he said, when you optimize one variable, another variable can go out of whack.
Multivariate SPC, on the other hand, does not look at each process variable individually, and is able to determine alarm windows around process variable correlations. Instead, it summarizes multiple signals, and the relationship of the data is just as important as the individual values of the data. For example, the process looks at the relationship of two variables that are coordinated, such as temperature and pressure.
Hazen described a case study MKS ran with Engel, molding a two-part medical hose connector on a 55-ton Engel e-motion press. The multivariate controller was allowed to change 10 different machine settings. The result, he said, was a cycle time reduction of 6 percent. The control was optimized based on changes in material viscosity.
A study at UMass Lowell showed the cycle time could be cut by 9 percent, and dimensional variation was reduced by 33 percent.
Hazen is based in Austin, Texas.