Organizational Opportunities from the Frontline Story 7: Creating Process Stability | Operational Excellence Quick Hits

Max Krug: (00:07)
In today’s session, we’re going to continue on the series of Organizational Opportunities, Stories and Lessons Learned from the Frontlines. Today’s story comes to us from a company that was a contract packaging facility. Then they package products on an automated packing line. So the company was experiencing increased demand and seeing a decline in productivity on the line due to frequent minor stoppages and breakdowns. So how could productivity be improved in this situation? I’m looking to make improvement to productivity immediately, not taking months to improve productivity.

Max Krug: (00:44)
So let’s talk about this situation. So what they had was when I analyzed their system, their pack line, they had a balanced line. So what does that mean? That means that each operation on the line, they had the speed set to the same rate. So in this case, I’ll use an example of 10,000 units per hour. So each operation was set to 10,000 units per hour, and there was very minimal work in process between operations. So we had huge dependencies on this line. So in this case, if I looked at the OEE, the first operation had a 95% OEE. So at 10,000 an hour, the output was 9,500.

Max Krug: (01:23)
If I took the OEE from each operation, I take the 9,500 times 90%, then it produces an output of 8,550. If I continued down the line with the different OEEs at the different operations, you could see that the output kept declining and declining, and at the end, they’re only getting 2,878 out of the line with an expectation of 10,000. So what was happening, they would have a minor disruption in one operation. That would stop the line. Then they’d get back up and running, and then operation six would go down. Then they’d get that back up and running. Then operation five would go down. So we never knew where the minor stoppage was going to happen or how long it was going to be down for.

Max Krug: (02:07)
Of, course then operation two goes down, so these stoppages were happening all over the place, and then operation seven. So it was contributing to the low productivity of the line. So the overall effectiveness of the line was 29%. So it’s like how do we improve this dramatically in a short period of time without doing a ton of effort? So when we look at a balanced line, the characteristics of a balanced line, in most cases, delays are transferred to the downstream processes and gains are usually never transferred. So why is that?

Max Krug: (02:42)
So in this case, what happens is we have variation and we have dependencies. So dependencies are the operations are tied together physically with very little inventory between. Of course, the variation is the minor stoppages that we were seeing. What happens is the greater the variation or the greater the number of dependencies, the lower the expected results. So if I were to add a couple more operations to the end of that line that had even 90% OEE, the output would be less than the 2,900. So we got to figure out how can we improve this system. So due to that variation of the dependencies, the minor stoppages was causing huge line productivity issues.

Max Krug: (03:28)
So what’s the solution? So what I did is I started look at how can we unbalance the line. So when we unbalance the line, what we’re doing is we’re reducing the dependencies between the operations. We’re not even looking at the variation right now. We’re looking at just reducing the dependencies. So the front of the line, we are still running at 10,000 units per hour, and with same OEE, so 95%, 90%, 88, and then what we did is we actually slowed the fourth operation down to 7,000 units per hour.

Max Krug: (03:59)
So people were like, “What are you doing?” But the other thing that we did is we put some inventory in front of operation four and after operation four. Now, this is a buffer. We’re using inventory as a buffer here to decouple operation three from four, and also inventory as a buffer to decouple our operation four to five. So why is that important? So just by putting this buffer in place, now I’m preventing delays from upstream stopping operation four. So even though I slowed it down, and by doing that, by limiting those disruptions, the OEE and that operation goes up because I don’t have the dependencies coming downstream and the variation coming downstream to stop operation four.

Max Krug: (04:44)
So in that case, if we’re running 7,000 an hour at 90% OEE, the output goes to 6,300. Then because we have the buffer of inventory between four and five, delays from operation four aren’t causing operation five to stop. So now at that rate of OEE of 90% on 6,300, their output goes to 5,670, and then 85% on operation six goes to 4,820, 85 on operation seven goes to 4,097. So in this case, what we saw is the overall effectiveness of the line increased from 29% to 41%, which represents a 42% increase in the line productivity, and that’s within a few days you can see that effect, not taking months to get there.

Max Krug: (05:33)
So by decoupling these operations, by using inventory in this case, we can reduce the dependencies and increase the productivity of the line. Again, my general rule of thumb is we want a minimum of 20% protective capacity in all other operations above the pace-setting operations. So in this case we had 30% and maybe a little over 20%. But that’s important so that delays from the upstream don’t cause operation for to stop and that the downstream operations can pull the work away at a faster rate.

Max Krug: (06:12)
So what are the characteristics of an unbalanced line? So the characteristics are that we gain improved productivity by decoupling. This prevents delays from being transferred downstream and gives protective capacity to most of the operations on the line. So why is that important? So this allows most operations to sprint, which means we can run at a faster rate to catch up when we have minor stoppages. Upstream operations can build that inventory buffer back up before it stops the pace-setter operation.

Max Krug: (06:44)
So we need that sprint capacity to run faster to build the buffer, and then the downstream operations can pull product away from the pace-setter, preventing downstream delays from stopping the line. So that’s the importance of unbalancing the line. Again, we can get immediate impact. Now, this isn’t where we stop. The next step is to focus. So where do we focus? So first, let’s eliminate the minor stoppages on the pace-setter operations. So what’s causing those stoppages? We did a failure modes and effects analysis to figure out what was causing the stoppages, and we started focusing on eliminating those failure modes.

Max Krug: (07:23)
Then once we get the pace-setter operation running more effective, then let’s find out which operation is causing the most what we call holes in the buffer, which is what’s preventing the buffer from being rebuilt before the line stops. So we take statistics on which operations are causing the biggest holes in inventory buffer, and that’s where we focus next to focus on improving the productivity of those operations. Then we systematically do that through the whole line to improve the overall effectiveness of the whole line.

Max Krug: (07:56)
So it’s not uncommon to see the output of this line go up significantly. So I know this company had a record day, and the record day was typically about three times what they were doing previously when they had the balanced line. So the output went up 3X. So that’s pretty exciting. So that’s our session for today. Again, you can connect with me on LinkedIn, visit our website, and also subscribe to our YouTube channel.