noviembre 19, 2003
Selecting SPC Software With Confidence
Embarking on a journey to select statistical process control software (SPC) software does not need to be fearful or uncertain. Reviewing and confidently selecting an SPC software product that's right for your company will be an investment of time and money. However, the right choice can ultimately be made if your thoughts are organized and you know what to look for prior to the purchase. Taking the time to evaluate your choices will prevent years of unnecessary frustration.
First of all, you need to know what type of SPC software your company needs because not all SPC products are vying for the same market. Knowing your needs from the start will save tremendous amounts of research time. The purpose of this article is to provide a list of topics that go beyond the obvious and the obvious is SPC software that creates control charts, histograms, statistics, capability indices and support some form of data entry.
Post Analysis or Real-Time
There are two main categories of software that perform SPC functions: Post Analysis and Real-Time. Phrases like high-end analysis, spreadsheets, data files, and ODBC links are used to describe post-analysis software. These products are typically stand-alone spreadsheet-type applications or linked out to external databases like LIMS and ERP systems. They are popular in academics, laboratories and research. Post-analysis statistical software is designed to "pull gold nuggets" from pre-existing data sets. Typically, data gets collected and someone else performs the analysis. Popular products that fall into this category include Statistica, Minitab, Statgraphics, SPSS, SAS, JMP, Genstat, Excel and various Excel add-ins.
In contrast, Real-Time SPC products are designed to be accessible across a corporate network and specifically deployed on factory floors where machine settings, product runs, measurement devices and the people are always changing. In the real-time world, each new data point offers an opportunity to: 1) make a process change or 2) do nothing. A good real-time SPC software should be able to provide effective alerts when process changes are necessary.
From this point on we will focus on selecting a real-time SPC software product.
Flat Files or Database
Data must be stored somewhere. The choice is either into data files or a database. Products that use data files typically translate into a separate file for each part. That is, 50 files will be required to store data from 50 parts. If your company makes thousands of parts.well, you get the picture. Setting up a single file is usually very simple. But managing data files from thousands of parts soon becomes a big task. However, companies that have limited SPC requirements get along quite well with a file-based system. If all your company needs are basic printouts of control charts and histograms you might want to consider a wonderful little $199 Excel add-in called SPC XL. There is no point in spending thousands of dollars when an inexpensive solution will do the job.
In contrast, databased products are best suited in corporate-wide SPC deployments where the same data needs to be accessed from multiple locations and workstations. Since all the data resides in a database, these products are excellent for comparative analysis across multiple parts, processes and test characteristics. Comparative analysis, however, requires a relational database organization. That is, all the raw data resides in one table and the items tagged to the data, (such as the part, process, test name, employee, specification limits, and so forth), are referenced from other tables. A true relational database design employs "referential constraints" where a database item, such as a part number, is defined only once in the database but can be referenced to give the effect of having that part number defined several times in the database. This type of design is very efficient and provides unlimited data analysis, sorting and comparison functionality. As a warning, just because the data is stored to a relational database, it does not necessarily mean that the relational features of the database are utilized in the application. Lastly, the database tables should be accessible by third-party applications like Crystal Reports.
Compliance and Validation
The software you select must comply with any regulatory requirements your company operates under. The most common is FDA's 21CFT Part 11: Electronic Signatures and Records. This regulation is specific to the pharmaceutical and medical device community, but will some day be common practice across industries producing products that enter the human body. As a rule, applications that store quality records to data files cannot become compliant. Any application that stores data to a local PC and then forwards the data to a master database is managing a system that has two copies of the same quality record. This approach also violates 21CFR Part 11.
Also, ask for the validation document showing that the calculations are accurate. Ask about the precision of the calculations. That is, how many significant digits are reliable.
Shop Floor Acceptance
If the software is to be deployed on the factory floor, representation from the floor must be included in the decision process. The best way to test the acceptability on the shop is to run the software on the floor in a pilot mode for a couple of weeks. This takes a lot of time, so this step should be done when you've narrowed down the selection to no more than three products.
Features to look for are easy and intuitive navigation with very little required button clicking. The interface should be modern, clean and simple. All unnecessary buttons and menus should be removed and not accessible unless the user has proper security privileges. Charts should be full-screen and/ or locked in place, and no individual windows should move or close. Big buttons, big message boxes, and use of color that easily distinguishes different screens and results are also important. The data collection interface needs to be simple and automatically flow through the collection steps in a logical manner.
Put the company's support to the test. Call them during your evaluation period. Make sure they can effectively answer your questions. Ideally, the support person will have the ability to access your desktop using a safe collaboration web tool such as WebEx. Also important, ask to speak to their in-house statistician. They should be able to help you with recommended statistical tools for your given situations. How the company responds to these statistical types of requests will help determine the usefulness of their support when the issue goes beyond making the software work.
Evaluating a company's support is especially important these days because many SPC software companies have changed hands or fallen on bad times. You must always ask yourself, "Will my selected company be around to support me for years to come?"
Configurable and Flexible
Regardless of the software, set up involves at least two required phases—configuring a data collection plan and creating charts. Within the data collection plan, one specifies the part, the process, the test characteristics, sample size, additional tag fields, data entry method and collection sequence. Once this is set up and data has been collected, many software products don't allow for any changes to be made to the collection plan. But in reality, collection plans will probably need periodic modifications. Someone wants to add a test characteristic, change the sample size, or change the collection sequence. To avoid future headaches, make sure any and all options are changeable on pre-existing data collection plans. Charts are the same way: they need to be configurable and flexible.
There are two issues which are very important for companies that intend to make business decisions with their SPC data. First of all, the control limits need to be unique for a given part number, process and test characteristic. If the storage of the control limits is based only on a unique part and test characteristic (as found in most SPC products), the user needs to be careful not to mix processes within the same control limits. Tagging the data with the process and then filtering all but one process is not a good solution. Correct implementation with these products requires a different file or set up for every process that runs the same part. Software products that define control limits using unique part, process and test characteristic combinations are statistically correct and represent the most efficient control limit management approach. Also, make sure the control limits are time-based. That is, updated control limits support effective dates. Control limits from old data are not changed when updated limits are established on newer data.
The other issue involves the standard deviation calculations. Specifically, the software needs to differentiate between the sample (long-term) and estimated (short-term) standard deviations. The long-term sigma (s as calculated using the n-1 formula) uses all the data values, independent of subgrouping. Long-term sigma is used in the Pp and Ppk calculations. Short-term sigma is calculated based off the range chart. Short-term sigma is used in the Cp and Cpk calculations. Make sure these differences are understood and correctly applied within the software you chose.
Make sure the company has industrial statisticians on staff. This will help to ensure the statistical validity of the software.
Integration with other Systems
It would be extremely valuable if the SPC package you choose can be integrated with other products such as document control and corrective action. Launching a rejection form when an event is triggered in the SPC will prove very useful. For example, with such functionality, you are able to automatically tell the operator that the gauge being used will need to be recalibrated in three days, or that the gauge, based on the most recent GR&R study, will consume 26% of the tolerance of the measured characteristic. These scenarios, and many others, are usually on a company's wish list. There are some SPC products that make these types of integration very simple to implement. If you need integration, make sure you ask the SPC company how they can help you accomplish it.
Get The Most Value for Your Money
There are many subtle differences among the different products. Look at the items below to make sure you are getting your money's worth.
- Don't pay more than 15% for standard annual support fee. This should include unlimited telephone, fax, email and Internet support. This should also include all future updates.
- Ask for copies of their software update logs. Look at the update record. Make sure the updates include substantial enhancements and new features, not just bug fixes.
- If the software supports e-mail, make sure it is not an imbedded proprietary e-mail system that requires a software license for anyone needing to receive the e-mail. The e-mail system should be SMTP compliant allowing email to be sent across the Internet.
- The fairest licensing scheme is concurrent-use. That is, the software can be installed on an unlimited number of workstations, but access to the software is controlled based on the number of purchased concurrent licenses.
- You don't need to pay extra for web publishing. Some companies have built-in utilities, at no additional cost, to allow automatic publishing of the SPC charts and data to web pages.
- Beware of the company that wants to sell you their data collection and control charting software and then try to sell you a "high-end" statistical product for the analysis. This not only increases the cost of the system, but complicates the deployment. Instead, make sure the software supports simple data extraction into any number of the popular post-analysis statistical products.
- You should not be charged for a company's validation documents.
- You should be able to use the software for at least 30 days at no cost and if you want to conduct a pilot, expect to be given 90 days.
Remember: Don't pay for expensive software if Excel Add-Ins will satisfy your needs.
There is no better way to have candid discussions about your thoughts than to talk to those companies that have already made an investment in the software you're considering. Ask for three types of references:
- A company that has made a recent purchase - this company will be able to talk about their selection criteria, initial implementation issues, and any buyers' remorse.
- A company that has been using the software for several years - this company will be able to tell you about how they've been treated after the sell. They will also be a wealth of knowledge of lessons learned.
- A company in your industry - this company will be able to talk your language. There are bound to be common problems that all companies in your industry share. Ask questions about how they've used the software to solve these problems.
Most importantly, take the time to call the references.
There are several providers of excellent SPC software, the key is to first determine your needs before you go shopping. Using this approach should limit your final selection to a couple of products. At this point, start your pilot program. Implement both products in a limited fashion. Let all interested parties play with the software and express their opinions. Within a couple of weeks your choice should be obvious. By following these steps, you will have confidence that you are making the right decision for your company.
Written by Steve Wise, a degreed Industrial Statistician and Director of Statistical Methods at InfinityQS International. Steve was one of the coauthors of the original Boeing D1-9000 specification, "Advanced Quality System for Boeing Suppliers." He is also coauthor of the book "Innovative Control Charting," published by ASQ Quality Press. Over the last 15+ years he has worked with companies on a daily basis implementing real-time production floor SPC and advanced quality management tools