Many companies have very sophisticated, state-of-the-art data collection systems that are supposed to provide all the information they need. Unfortunately, too often this information is inaccurate, fragmented, outdated or just too cumbersome to retrieve. If businesses are run based on timely and accurate data, then it is critical they identify needed information and ensure they can get it when they need it.
In some companies the act of data collection and compilation can be all-encompassing. Hours of non-value-added time can be spent feeding data into one, or in some cases, multiple systems. Online data collection systems are easier to work with, but they can be both a blessing and a curse. With such systems, employees soon get comfortable punching in (or preferably wanding in, if bar-code technology is in place) information that magically appears in a database of some sort—this is the “blessing” of the process. The “curse” is that once the information is entered, it may not be used for any meaningful purpose. In fact, companies may be missing (or ignoring) important results. Even worse, they could be using only a portion of the information and making bad decisions because of it.
Recently, I worked with a company that had a very sophisticated data collection system tied to all of its machines. The company’s employees did an admirable job of entering various types of data into terminals located throughout the plant. This data was then used to generate reams of summary documents. Unfortunately, management personnel were not able to do much with this information. I found this out when I asked some of the managers what I thought was a simple question: “On average, how many hours a day is a machine down for job change-over?” After almost two hours of searching for, reviewing and analyzing data, someone came back with the classic answer: “It depends; some days it is never down for change-over, and other days it is down for just about the whole day.” Now that was a good example of a system that required a lot of care and feeding but was not able to provide important information in a timely manner.
To get the most out of their “systems” companies must determine the type of information they need to effectively make decisions. Certainly, some of this information is obvious, such as output, equipment downtime (planned or unplanned) and some measure of the rate of quality of the parts being produced. Within each of these major categories, more detail is probably needed, such as reasons for downtime (for example, setups and breakdowns) and some of the more common causes of machines running slower than normal (such as stoppages to change tools, load material or address fixturing problems). This additional information may be useful in determining how effectively equipment is being used.
However, because data can be so difficult and costly to obtain, we have to choose what we want to capture wisely. If we focus on a limited number of reasons for production delays, we may be able to identify other reasons indirectly and not have to spend time recording them as well. For example, if we are effectively capturing scrap and rework totals by segregating and accounting for all defective parts produced (which we should be doing), we can attribute a portion of a specific production overrun to the time spent producing these bad parts. Additionally, if we have a good handle on the time taken for setting up a machine, and if we record the amount of time a machine is stopped for mechanical breakdowns, then we have further narrowed the causes for jobs taking longer than planned. If more details are needed, they can be obtained through conversations with the machine operator or others involved with the production process.
Ask yourself what information you need to make good decisions for your operation and how much of it you can reasonably expect to capture from your current system. Keep in mind that the more difficult it is to record information and the more burdensome it is to manage, the less likely it will be accurate (or recorded at all) and put to good use. If what you are seeking is just too hard to obtain, you may have to think seriously about a different system, or adjust your expectations and focus only on the information that is truly critical to the business.