The Last Word: Daunted By Data
Displaying measurement data can be dangerous. A recent experience in another area of life gave me some insight on how overwhelming uninterpreted data can be and how too much data can crush someone with an untrained eye.
My wife and I spent several weeks at the hospital after our son was born early. He was put in the neonatal intensive care unit (NICU) for a time, which proved a bit daunting. Being surrounded by all the equipment in a typical NICU added to our confusion of our son’s condition. Equipment in a NICU measures such things as a baby’s heart rate, respiratory rate, body temperature, blood pressure, blood oxygenation and more. Real-time numerical values and data trend lines are displayed for each of these measurements. If that weren’t enough, each display is capable of monitoring multiple patients, so we had to determine which display was specific to our son. To add further to the situation, whenever one of these measured body rates falls out of a preset tolerance range, alarms sound-and they also sound if another data set falls out of tolerance. Guess what? Premature children often have measurements that stray “out of tolerance.”
The NICU nurses have a great combination of skill and compassion to help Mom and Dad understand, but not get fixated on, these data collectors. They stress the importance of focusing on your child rather than the numbers. When an alarm sounds, they know whether it is something to be concerned about. We, on the other hand, became mesmerized by the numbers to the point of data overload.
After a few days we learned much of what the nurses knew about the data collectors. We learned to view the data as a whole instead of just its individual parts. We learned whether alarms should be heeded or whether they were anomalies.
The data collectors you have on your production floors can have the same effect on your operators. In full production mode, your measurement, test and inspection equipment sends a continuous stream of data to monitors, displays and software packages. Those on your floor who are novices about interpreting data are likely to be just as mesmerized and overwhelmed as my wife and I were at the NICU. Are parts being made in or out of tolerance? Why does that alarm keep sounding and should I be worried about it? What is this data telling me about the overall health of the process?
This issue of Quality is filled with useful software offerings that help answer some of those questions and go a long way toward controlling which data is reported. Software is becoming more advanced each day, so it doesn’t pay to scrimp on your spending there.
Despite the power of software, however, interpreting the data is still largely in the realm of human expertise. The people in your plant responsible for quality should be the ones who understand the big picture when it comes to data. They are the ones to provide reassurance to those less familiar with data analysis. They are your “NICU nurses.”
However, those “nurses” need to explain the data collection to new “Moms” and “Dads” and teach them to interpret it as well. Those who are inexperienced need to learn which alarms are to be heeded. Understanding the ebb and flow of numbers, alarms and graphs will go a long way toward having more people understand the “health” of the process and be able to contribute to its care.
With such skills in hand, your operators will soon more fully understand the process at hand and see the data collectors simply as a tool that provides some information for their broader knowledge of what it takes to manufacture good parts. They will understand the need to focus on that which is most important, the “baby,” rather than the data itself.
Share your thoughts with other members of the Quality community at the Quality Magazine LinkedIn Group page , the the Quality Facebook page and on Twitter .