What is gauging For SPC

To measure up to the task at hand, gauges for SPC need to have certain features. These include: 

• The ability to produce variables data. SPC requires resolution to detect minute changes in a process. The higher the resolution of a gauge, the better, as long as other criteria are met. In general, GO/NO-GO gauges do not provide the kind of measurement data needed for SPC. 
• Repeatability and reproducibility (R&R). SPC requires precision. If you use a gauge to measure the same part repeatedly, you should obtain the same reading reliably. Similarly, a different person should be able to reproduce those results. Error caused by lack of gauge R&R reduces confidence in the data and the ability to make good decisions regarding parts. All human engineering or “ergonomics” put into a gauge such as method of holding it, positioning a part, and training needed must evidence a concern for good R&R. 
• Efficiency. SPC requires large quantities of data. This calls for a gauge that produces data fast, and can be used over and over again reliably. 
• Ability to support statistical analysis. SPC acts on groups of data, but a gauge is essentially a serial device that produces one piece of data at a time. A gauge must also support data collection by providing data in a uniform format, so they can be recorded and grouped.

For maximum efficiency, the gauges used should have electronic data output. A gauge that does not have electronic data output still must be read, no matter if it has LCD digits, columns of lights, mechanical pointers, a large display, or a small one. If the gauge must be read by someone, it becomes subject to human error. That’s because any gauge can be read incorrectly, the reading can be interpreted incorrectly, or the reading can be recorded incorrectly. This potential for introduction of human error is solved by direct electronic output to a data collector.

A gauge with electronic output speeds both measurement and data processing. The person taking measurements can concentrate on getting good readings, and the data collector can convert the serial stream of numbers into groups of data for analysis.

Most gauges now have electronic output, and can be connected to data collectors that record and analyze data. Several standard gauge outputs have been developed, such as serial BCD, serial ASCII, serial binary, and RS-232C. Some column gauges and metrology displays have analog outputs at standard voltage levels.

From gauge to Data Collector

The SPC process can be looked upon as a set of subsystems, each of which has a specialized purpose:

• gauging, the first subsystem, has minimal intelligence, but provides data output and a digital readout. 
• The data collector, the second subsystem, is used at the source to record data in computer-readable form, and to provide feedback to the operator. In some unattended data-collection operations, this feedback may go directly to a process controller such as a PLC or SLC. 
• SPC analysis software is a third subsystem, and is used to perform analysis, generate graphs on a CRT, and produce hardcopy reports.

Separating the data collection system into subsystems allows you to better identify which type of specialization belongs where. For example, if a data collector with built-in statistical analysis software is used at a station with multiple gauges, buying gauges with built-in statistical software would be redundant. In this case, all you need are gauges with data output.

Most electronic data collectors concentrate some intelligence — that is, SPC software designed especially for machine and station operators — in the data collectors themselves. This provides several advantages: 

• The data collector gives the operator needed feedback at the process, more feedback than any single gauge could provide. This feedback includes prompting an operator for which gauge to use and which dimensions to measure. 
• The data collector compares each gauge reading against reasonable limits and part specifications, and compares subgroups against control limits. The collector then alerts the operator if something is wrong. 
• A data collector performs math operations on several gauge inputs at one time, and plots charts on derived variables such as total indicator readings (TIR), minimums, and maximums.

Where a process is largely unattended, data are recorded at the process, and analysis takes place off-site. Typically this is on a desktop computer in a quality control department. As an option, a handheld data collector with SPC software could be employed at the process site.

From Data Collector to Computer

A data collector is a vital link to a third type of subsystem, computer software for SPC. Though most of the actual SPC work occurs at the process, most of the decisions affecting that process occur at the plant level.

Good decisions require good data. An SPC data collection system that feeds information to management serves to focus resources on the big problems (the Pareto principle). Moreover, SPC software on a computer provides a common language for discussing problems that occur at each process, and this software is a good tool for estimating capability, scrap, rework, and cost of quality.

Software for PCs is now part of most data collection systems. Some, such as those offered by DataMyte, maintain an open architecture. Many commercial software vendors have developed interfaces to these open-architecture systems. Use of these interfaces allows SPC users to choose from among many valuable features for data analysis and reporting, and to maintain SPC databases on a variety of different computers. In addition, most SPC software programs now provide their own open-architecture databases.