Printer Friendly

Data collection and SPC.

Data collection devices and statistical process control (SPC) are a winning team that dramatically reduces variations in manufacturing processes and contributes to US competitiveness. For manufacturers serious about their own prosperity and survival, data collection and SPC are more than just a good idea. They are an integral part of an overall quality strategy that fosters continual improvement.

Still, data collection and SPC are really nothing new. They are simply tools that collect data on our processes, notifying us if our process is out of control. That's half the equation. The other half is what we do with that information. Proper SPC follow-up by trained personnel who understand the process will find the cause of the variation, eliminate it, and bring the process back in control.

Conversely, without SPC metrics in place, personnel may not react when they should, or may react when they shouldn't. Such tampering causes more variation in the process. Fortunately, SPC differentiates variation due to common causes from variation due to special causes. A common cause is a source of random process variations that affects all the individual values of the process. In contrast, a special cause is an intermittent, unstable, or unpredictable source of variation.

By using control charts, operators can track and eliminate variation due to special causes that improve process capability. Problem solving teams can then work to reduce common cause variation by making changes to the process. It's important to note that management action is required here. Management action may consist of changing suppliers, changing the method of operation, or approving the purchase of new machinery. Whatever their course of action, the ultimate purpose is to improve the process. Such actions directed at special and common causes will reduce work, reduce scrap, reduce costs, reduce customer complaints, and improve overall quality.

Computer-aided SPC can provide benefits in accuracy, efficiency, and timely information. SPC software has three primary functions: charting, alarming, and data collection. Many types of charts are available for use with SPC software. They include Pareto charts, histograms, cause and effect diagrams, and scatter diagrams. The most common chart is the control chart, which differentiates common causes from special causes. It allows a plant to determine if the process is changing, and it measures the effect of a change made to the process. X-bar and range charts are examples of control charts.

Alarming involves the identification of out-of-control conditions. This identification is usually based on the Shewhart test for points outside the control limits, trends, and other non-random patterns of points within the control limits. SPC software can be designed to alarm for these conditions by something as simple as highlighting a point on a chart or as sophisticated as flashing lights, sounding bells, or activating beepers. Alarming sends a message to appropriate personnel: "Your process is out of control." Designated personnel must react by investigating the process, discovering the cause, and correcting it.

It follows then that a key aspect of SPC is speed. It is not enough to analyze data. Data must be analyzed quickly enough to react effectively to out-of-control conditions as they occur. Integration of SPC software with data-collection devices is used to provide data to SPC software. The devices range from gages to optical mark scanners to pen-based computers.

A wide variety of gages can be used to measure critical dimensions or features on manufactured pieces. There are calipers, micrometers, and indicator gages. Plus, there are gages to measure force, weight, surface, bore, and hardness. Typically, the operator takes measurements using the gage and then uploads them to a computer, which performs SPC analysis and gives immediate feedback on a control chart. For example, at an automotive components plant, an operator lowers a probe over a substrate, pushes a foot pedal, and substrate thickness is automatically measured.

Another type of gage used for larger parts is a coordinate measuring machine (CMM). Unfortunately, in many applications, CMMs are housed in clean rooms located off-line, sometimes far from the manufacturing line. Thus, turnaround time in moving product samples from the production floor to the clean room and obtaining results is too long.

Portable collectors

At the manufacturing plant locations where it is prohibitive to have SPC stations, a common form of SPC data collection is the hand-held data collector. Portable data collectors permit the collection, display, and analysis of both variable and attribute data in real time. Data collectors can be used with a variety of gages and measurement tools, such as torque wrenches, gap and flushness gages, and calipers. Data collectors use bar-code wands for attribute data and part label data such as lot number and operator. Data collectors with SPC capabilities only require a push of a button to bring up control charts, histograms, and Pareto charts on a high contrast LCD screen as data is collected. Some units even have multilingual reporting capabilities. Another valuable feature that some data collectors have is a comment field, which enables inspectors to indicate the special causes of out-of-control conditions.

If the inspector is using a data collector without SPC capabilities, the data collector is taken to a personal computer where data is uploaded for statistical analysis. Whatever else is required for the application can be uploaded--typically, part number, date, and comments. This permits building a history of special causes. If the cause of variation is unknown, or if the operator is new, the computer software can be used to help find the cause by examining the history database. No one has to reinvent the wheel to solve that problem. At a metal stamping facility, for instance, the cause might be a disturbance in the process such as a loss of air pressure or a temperature increase.

There are a number of data-collection technologies that provide data for SPC analysis. They range from optical mark recognition (OMR) to voice. OMR is a method of data collection used to collect attribute data, which is gathered in the form of non-conforming units or "nonconformities." OMR is simple but highly efficient when applied properly. OMR requires an inspector to fill in with pencil or pen a bubble on a paper ticket, similar to how students take the SAT test. The ticket is a pre-formatted sheet that lists all possible product nonconformities. In addition, the nonconformity can be described on the ticket itself.

In the case of an auto assembly line, for example, the ticket might list subsystems: cigarette lighters, door locks, power windows, stereo speakers, etc. The inspector might put a check next to "Cigarette Lighter," and then note that the cause was a loose wire. Once marked, the ticket is scanned to provide data to SPC software that notifies personnel of out-of-control conditions.

Touch of a pen

Hand-held, pen-based computers are a relatively new application of technology to data collection. This technology uses a pen or stylus in conjunction with a high-resolution graphics screen. In practice, an operator walks around the product being inspected--a tractor, for instance--with the pen-based computer that has a graphic representation of the tractor. Different graphics are available for different models. Instead of itemizing paint nonconformities on clipboard form as is typically done, the operator touches the pen to the area of the tractor graphic where the nonconformity occurred--the hood, for instance. Then a list of possible problems appears on the screen. The operator identifies the problem by touching the pen to the problem: a scratch, dent, ding, low spot, etc. That information is later uploaded to an SPC package, or an SPC package may be embedded in the pen-based computer which constantly monitors for out-of-control conditions.

In another scenario, the pen-based computer may be linked to a host computer via radio frequency that is performing SPC monitoring and routing functions in real time. Whatever the configuration, pen-based computers provide the advantage of portability, paper savings, and instantaneous results when embedded with an SPC package.

In addition to gages and data collecting equipment, programmable logic controllers (PLCs) are an important data-collection element source for SPC. PLCs are often set up to read process information such as paint booth temperatures and humidities. PLCs are often part of a network, such as a plant monitoring network. Information is tied to an SPC module which analyzes the condition of the process. This information can then be dispersed to specific groups like maintenance, engineering, etc.

Other data-collection technologies include bar-code equipment, voice technology, and vision equipment. While they provide data for SPC analysis, their use is usually limited to niche applications.
COPYRIGHT 1993 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:statistical process control
Author:Paul, Linda; Miller, Walt
Publication:Tooling & Production
Date:Jan 1, 1993
Previous Article:Whiskers toughen ceramics for hardest cuts.
Next Article:Cutting carbide specials in days, designs in hours.

Related Articles
SPC software: ready when you are.
SPC leads to the right mix for investment casters.
SPC for small to medium foundry's sand system: Part 1.
Quality statistical process control at Cherry Textron.
SPC takes off for aircraft parts producer.
New SPC technique excels for multicavity molds.
Closing the loop on data collection.
MES needs real-time SPC.
SBC Enhances ABC.
Implementing statistical process control.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters