In our last two columns, we?ve discussed how to select the measures you?ll be using for your lean initiative and how best to communicate them to employees. This month we?re going to talk about how you and your colleagues, as company leaders, are going to actually use the metrics.
We can sum this up in just a few words: You?re going to discuss your progress on the measures each month without fail. You don?t need six sigma, you don?t need statistical quality control (though those tools certainly can be helpful). You just need to be willing to look at each metric as a leadership team and start a discussion that seeks to answer two questions:
(1) Did we get better, stay the same or get worse?
Let?s back up just a moment and talk about the idea of regular discussions of the metrics. The leadership team needs to meet at least twice a month to review the overall progress of the lean initiative. One of these semi-monthly meetings should be devoted to a close look at the metrics.
?Can?t we meet once a month, or less often, and discuss the metrics then??
No. One meeting a month that covers both metrics and reviews lean activities isn?t enough. Meeting less often than monthly is totally unacceptable, no matter how mature your lean project is. It shows that leadership simply isn?t committed to the initiative.
?Do we really have to devote an entire meeting to just looking at the measures??
Yes. Less time doesn?t allow for enough discussion. You aren?t simply going to check the measures and move on. You?re going to dig into each one. Why did scrap go up after a three-month decline? Is inventory going down while customer service gets better because our lean initiative is working, or is it just good luck? Changeover times got better for awhile after the kaizens we held, but now they?re trending back up. Why? Why did tooling repair turn-around times get much better last month?
?So that?s it? We just start a discussion on the measures??
That?s right. To help you along, here?s an agenda with a bit more structure:
(1) Have the updated graph for each metric ready before the meeting. Don?t try to start the discussion without the graphs. Unless team members can see what?s going on, they won?t be as motivated to weigh in on the discussion. Make sure each leadership team member has a copy of each graph or project the graphs so that everyone can see them clearly.
(2) If it?s early in the lean initiative, start the meeting with the metrics that are still being developed or that team members don?t fully understand. One company I worked with spent several months developing a clear definition for ?equipment availability,? but this turned out to be the company?s most useful metric. Other companies have spent several meetings figuring out which of their several versions of scrap measures would be used as the lean initiative metric.
(3) Move on to asking the fundamental questions for each metric: Are we getting better, staying the same, or getting worse? Why?
(4) In answering the fundamental questions, dig beneath the surface a bit with questions like:
(a) Is the change in the metric part of a trend or aberration?
(b) Is the change significant or just a random blip? Do we see a pattern in the behavior of the metric?
(c) If we?ve improved, do we seem to be reaching a new level of performance or are we just having a few good months?
(d) If we haven?t improved (or even gotten worse), can we figure out why? Is there a particular machine, operation, or product that?s causing most of the problem?
(e) If the measure hasn?t changed in spite of lots of effort, do we know why?
(f) Is it the right metric? Is the data behind the metric correct?
(5) Develop action items. Every discussion of metrics, especially during the early stages of the initiative, should generate action items of some sort. Perhaps the metric itself needs further study or development. Maybe the way the data is collected or collated needs to be reviewed. Or an improvement idea needs to be pursued. Every discussion of metrics should end with an action list.
Let?s look at a quick example nearby in Graph 1, which shows the number of unplanned tearouts and resets each week on a set of presses and the average time from beginning of tearout of the die to a successful reset. This chart can and should generate a number of questions:
(1) What causes the numbers of tearouts/resets to spike occasionally?
(2) Can we sustain the good tearout/reset numbers at the present levels? Is the reduced number because of our improvement efforts or have we just been getting lucky?
(3) Why is the average tearout/reset time trending up so sharply?
One more example will give us an illustration of how a graph might lead us to drill further into the data. Graph 2 nearby shows the average running efficiency for several presses. Essentially, we need to know what?s causing the monthly cycle of efficiencies. The leadership team needs to dig further to find out what?s changing at the start of each up and down trend. It might be product mix. It might be that the company was hiring new operators. It might be that one or two presses were acting up in this cyclical fashion. It might even have something to do with vacation scheduling (notice that the graph covers the spring and summer months).
These discussions could lead to further study of the conditions or to solutions and new improvement initiatives.
– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Rick Bohan is the principal of Chagrin River Consulting, LLC, Chagrin Falls, OH, www.chagrinriverconsulting.com. For questions or comments on this column, contact Rick at 216-409-9056 or firstname.lastname@example.org.
Survey from ABB Reveals Re-industrialization at Risk from Education Gap in Automation
One in four education establishments use robots in teaching despite planned investment in robotics and automation by 70% of U.S. and European businesses.
Beckwood Press Co. to Launch Linear Servo Press™ Tech at FABTECH 2022
The patents-pending Linear Servo Press line uses fewer components to create force resulting in cost savings, increased throughput, and easy setup.