Uptime: Beware Of Those Simplified Metrics

By Bob Williamson, Contributing Editor

In our culture, there are certain images and symbols that need little explanation: They’re intuitive. This includes some used in the workplace. Have we become too dependent on them?

Performance charts and graphs in today’s workplace often use RED and GREEN indicators for unacceptable and acceptable performance. Many bar charts, Pareto charts, line graphs and other performance or productivity reports incorporate the visual communication of RED as unacceptable (needing improvement) and GREEN as acceptable. These colors are universally intuitive in our culture. My advice: Beware of these simplified metrics!

Percentage indicators are also a common progress- and performance-communication tool: Percent of downtime caused by “X,” overall equipment effectiveness percent, percent reliability or production efficiency percent of goal. Again: Beware of these simplified metrics!

In last month’s Uptime column, I discussed “Problem-Solving and Culture Change.” Understanding the change process and leading changes in the workplace were explored, along with the importance of answering the question “why change?” Unfortunately, RED-GREEN and PERCENTAGE indicators can lead to sluggish improvements and mask the path to problem-solving as we get to the fundamental question of “changing what?” on the plant floor.

It’s not granular
Let’s start with those RED-GREEN indicators. It’s not difficult to understand that any RED indicator is not a good sign. But key questions are not answered: Specifically, what to improve, why and how?

RED on a bar chart or line graph basically says “you need to get better at doing whatever you are doing.” It can motivate a group of people to work faster and harder to meet the goals while causing other problems like defects or injuries—unintended consequences. If, for some reason, the indicators turn GREEN, the work group may be unaware of what specifically happened to improve their performance and think “we just got better.”

RED and GREEN are what I call “macro indicators.” They’re not granular enough to point to the causes of poor performance or specific actions taken to improve performance. As “macro indicators,” RED-GREEN charts and numbers are great for front-office and plant-floor management. Alas, they’re pretty much useless for work groups that are responsible for making changes to improve the performance of their work processes or equipment.

When the numbers or graphs and charts turn GREEN, the work group sees its performance as “acceptable.” But is “acceptable” really what the group and its equipment are capable of producing? Being GREEN does not necessarily lead to the team spirit or the efforts associated with continuous improvement. Acceptable performance, in the GREEN, is mostly interpreted as meeting the minimum requirements.

An example: Visual production tracking charts were recently implemented on the plant floor. The day-shift production group struggled to meet its hour-by-hour goals, Monday through Thursday. Its charts were mostly in the RED. Maintenance responded to trouble calls daily. But on Friday, for some unknown reason, the group exceeded all of its hour-by-hour goals and the daily goal by 2:00 p.m. The group finally got in the GREEN and was praised for it. This trend went on for weeks.

What, in this example, changed on Fridays? Did the work group do anything specific to improve its performance? Did maintenance fix a problem with the equipment? The answer: The work group worked faster as a team so people could finish their shifts and receive the week’s pay early. The maintenance calls during the week gave the production workers numerous additional breaks.

It’s relative
PERCENTAGES are not real: They are relative indicators. “Percent” means “per hundred.” Eighty-five percent could be written as 85%, or 85-percent, or 0.85 (85/100). But what does the percent really mean? Consider the following: 

• Eighty-five percent (85%) defect-free production also indicates that 15% is defective. How many defective units were produced?
• Eighty-five percent (85%) of the defects were caused by equipment problems. The remaining 15% were caused by human error. What were the “equipment problems?” What were the “errors” (i.e., mistakes, sabotage)?
• Eighty-five percent (85%) overall equipment effectiveness (OEE) is the goal. 85% of what? What if we could get to 98%?
• An 85% performance bonus was awarded to employees this quarter. What does that mean in terms of dollars and cents?
• Bobby received an 85% on a final exam. What does that mean: pass/fail, an A, B or C grade? What questions were missed?

We rarely use percentages in our daily lives without interpreting what the actual results mean. So why do we continue to use percentage indicators in ways that require translation on the plant floor rather than actual performance indicators to make improvements?

Making improvements
Plant-floor work groups typically include front-line supervisors/leaders and their production and/or maintenance crews. When making sustainable improvements, these groups should focus on actual units of measure—ie., time (seconds, minutes, hours), production parts, finished assemblies or volumes (CFM, GPM)—and the actual causes of poor performance. As noted last month, the nagging question among plant-floor work groups is “why change?” Let’s expand on that question with another example.

Why change? The packaging line in this example was comprised of eight machines all inter-connected with conveyors. The RED-GREEN production downtime reports showed this line performance as all RED (55% OEE) for the period being measured. The message was “improve the packaging line performance up to the goal of 68% OEE (overall equipment effectiveness)!” 

The posted “Downtime Loss” bar charts for the packaging line showed the following:  

What did the percentage-based “downtime loss” information tell us? Changeover is the biggest reason for downtime, and Machine A is number two. So what?

Focusing on “changeover” improvement is admirable. But once the line is changed over, it still doesn’t run any better. Because the remaining downtime losses aren’t addressed, improving changeover times can actually lead to more scrap product per shift until the downtime/defect-causing machine problems are improved. Truly an unintended consequence.

Let’s look at the “Downtime Loss” report from a real-world plant-floor perspective. Reflecting the same data in the previous chart, it’s expressed in units everybody can understand (production-time lost). It just wasn’t posted on the plant floor:  

Since Machine A is the most problematic (16,550 minutes of downtime loss), it’s the perfect place to begin improvement. Actual production loss at a line rate of 100 per minute equals 1,655,000 units NOT produced because of Machine A problems. When the data is drilled another layer deeper (by section), we can see the reasons for such high downtime. Again, this information was not posted on the plant floor:

With this information the work groups can now see that Machine A Sections 7, 3 and 10 are the most problematic and to what degree. Given the new information, the group knows exactly where to begin focusing its problem-solving efforts—and how to measure the results of its improvement actions.  

Leading culture change
Given detailed downtime loss information, plant leadership can focus its plant-floor work groups on an actual “business case for change.” Together they can all focus on fast and sustainable packaging-line improvements using the following:

• Report actual production rates and requirements of the Line (not RED-GREEN, percent).
• Report the units NOT produced due to documented equipment problems (1,134,900).
• Report lost production impacts on customer shipments, overtime and cost per unit.
• Provide resources and empower the groups to identify and eliminate problems.
• Learn from the problem-solving efforts to address additional downtime losses.

When plant-floor metrics and improvements are visible and linked to realistic indicators, they require little or no translation. RED-GREEN and PERCENTAGES, though, are open to interpretation—and require translations for real-world use. 

Don’t forget: What gets measured gets done. What gets rewarded gets done. Be sure to measure, report and reward the right stuff. MT 

Robert Williamson, CMRP, CPMM and member of the Institute of Asset Management, is in his fourth decade of focusing on the “people side” of world-class maintenance and reliability in plants and facilities across North America. Email: RobertMW2@cs.com.