Blow molder plunges into CIM.
How did Winn justify such advanced computer integration of a collection of workhorse Battenfeld and Bekum shuttle blow molders, many bought used ranging in age from two to 10 years? Winn installed CIM for two reasons--to speed just-in-time delivery of stock and custom bottles and to tighten statistical process and quality control (SPC/SQC) for part qualification in an increasingly competitive cosmetic-bottle market. Winn, which makes bottles from 2 oz to 1 liter for beauty parlor and other private-label toiletries, expects CIM to help cut a week or more off its current four-week delivery times by reducing scrap and increasing productivity.
Winn's production and process monitoring were installed and running by last March. Winn has also greatly upgraded electronic quality-measuring devices and will soon have these hooked into the CIM network as well. Ultimately, Winn's goal is to project orders and manage inventory, automatically, updating for months ahead. For now, its CIM program updates current jobs, but not the next job in line. Future jobs are still entered in a separate computer file.
The "brain" of Winn's CIM network is an IBM RS6000 RISC-type workstation, running Focus 2000 software. The RS6000 acts as a server for nine standard IBM-compatible PCs that can access Focus 2000 screens in various company departments. Focus 2000 runs on a Novell Network and is being upgraded to faster, more powerful Windows 3.1 software. That software will cut the computation time required to graph trend charts from 1.5 min to 3 sec. Users also won't have to pause as displays are redrawn while a ripple of new machine data is absorbed.
Monitoring data are collected from six PlantStar M-CIM (manufacturing cell interface module) data-acquisition terminals mounted on pillars in the plant's center aisle between the double row of blow molding machines. The M-CIM terminals gather data from two or three blow molders apiece. Machine packer/operators can enter scrap and downtime reasons on the touchscreens for their machines. On any terminal, a user can call up touchscreen graphic displays of real-time data for all the machines.
M-CIMs take their data direct from sensors on the blow molding machines, not via the machines' process controls. This avoids difficulties with the differing levels of sophistication among machine controls of different ages. M-CIMs monitor six data points at Winn: cycle time, trimming time, barrel temperature, mold cooling-water temperature, blowing-air pressure, and screw rpm. These sensor points are hardwired to the M-CIMs.
Do M-CIM sensor data and the machines' built-in process controls always agree? No, and that's an added safety feature. "It's a way of double checking the accuracy of your sensors," says Winn quality control manager Michael Ramos. Several times already, he says, Winn has detected sensor problems because of such a discrepancy.
To protect stored data in case of power failure, the CIM system is backed up with battery packs at each M-CIM and at the network computer (RS6000) level. Batteries can support the server for 20 min, with automatic shutdown after 15 min. The M-CIM database is maintained up to 30 days under battery power.
CIM is bringing some efficiencies already. Production manager Wayne Whitaker says he sees as much as 15% higher productivity, though some of that improvement may be more apparent than real. As operator/packers learn to use CIM, it makes them more aware of coding a reason for rejecting parts. "They used to reject parts just to catch up in packing," he says. "They're more aware now, and that doesn't happen any more."
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Winn Packaging Co.; computer-integrated manufacturing|
|Author:||Schut, Jan H.|
|Date:||Nov 1, 1993|
|Previous Article:||Cadmium pigments no hazard in molding.|
|Next Article:||Infrared sensor takes thermal portraits of blown-film bubbles.|