Five years ago, MS Aerospace identified an opportunity to react to market demands for shorter lead times if the heading process to make its bolts could be improved. Harry Smith, Lean Manufacturing Coordinator, tells Roberto Priolo why systems thinking was the only way to achieve lasting results.
Roberto Priolo: How did your lean journey start?
Harry Smith: Reading Deming’s work and going on benchmarking visits had already helped us to develop an understanding of the philosophical approach of always trying to see things as a system. Visiting with Dr Bill Bellows of the In2:InThinking Network and his coming to MS Aerospace was important as well. Understanding the scope and interaction of those systems was the real epiphany. It wasn’t long before a relationship between lean manufacturing concepts and the potential for business improvement revealed itself.
When we were asked by senior management to look at the performance of the heading department, which had been experiencing a big backlog of orders, we realised that the only way to solve the problem would be to look at that process as one part of a larger system. Once changed, what would that process do to the rest of the organisation? Whatever we did would have to benefit the whole system, not just one section.
RP: How did the journey unfold and what benefits were you able to see along the way?
HS The MSA journey began as a couple of slow starts and restarts until a better and more focused understanding of how to apply the lean principles was understood. Typical stories of unmaintained 5S programmes, “let’s put that on hold for now” approaches, and all the other anecdotal references that can be attributed to a failing lean implementation.
However, after a couple of difficult starts, perseverance won the day. In concert with a couple of our customers, we began having a few organised events that started to bear fruit. Today, 6S programmes, setup reduction events and value stream mapping serve as the core of the programme. There are all classic lean tools, but they are applied with the understanding that they serve interdependent parts of a system, not stand alone processes. There is still a long way to go but we are at a stage of understanding that allows us to select and implement the most appropriate tools that yield the greatest returns.
One thing is certain, you should expect nothing less than a 50% improvement in a process after a lean event or series of events.
Looking at a specific project, we started at the beginning of the value stream with order receipt through work order creation. When we began, this process, which spans several department boundaries (it truly is a system), was in terrible health: it routinely carried a backlog of over 400 jobs with a throughput time of six weeks. Using classical VSM techniques and by understanding how the interaction of each individual step affected the entire system we made changes that reduced the backlog to 50 jobs with a throughput time of one week. More importantly, we have sustained these gains and as a result we can now offer lower quoted lead times to our customers. Reclaiming five weeks of throughput time and passing that onto the customer was a massive achievement.
We then moved to the first manufacturing step, heading. When we started, the department was hopelessly behind schedule, frequently with orders going late to the customer before they were even started. After numerous improvements, all grounded in systems thinking the backlog of late orders dropped to zero. The labour content of this operation was reduced by 25%, allowing us to cross-train one operator into a different bottlenecked department.
It’s important to note that these improvements complimented each other and that the entire system benefitted from them.
RP: How did you communicate the necessity of adopting a system-wide approach and how did you get around resistance?
HS: Many of the things we now do differently require extra time and effort from people who often don’t see a direct benefit to themselves. When you try to change one step of the process, those involved in it will tell you that the new way requires more time and energy and are inclined not to follow it. That’s when education and training come into the picture: we had to explain to people that, yes, that one step may take longer to complete, but that the benefit of that change will be felt throughout the company. It will help their colleagues, and the parts the organisation produces will be better.
At that point, the argument we heard was: “I am not measured by how helpful I am to the overall system! I have clear production goals.” Systems thinking training was very helpful in overcoming this barrier, and we also had to rethink what we measured, to ensure it wouldn’t only focus on each individual’s output.
HS: Our aim was to understand what issues our people were experiencing. We identified tooling as being one of the challenges they were facing.
We were talking about something that preceded heading, but in the department there was a perception that the world started right before and ended right after their process. People thought they lived on an island where they did the best they could given what they had. We tried hard to stress that was not the case.
One of the issues was how well we could polish the inside of our heading dies. The technology we had back then did not allow us to polish them as well as we needed to. We went upstream to visit the people who make and polish our dies. We understood the impact that they have on the heading process and started wondering what we could do to improve things. It turned out we needed to invest in new machineries, even though that would add 30 to 45 minutes to the die-making and polishing process. However, the end result would be a better product. The question really boiled down to our willingness to invest in the making of a die, understanding that a better die benefited the entire process. In understanding that the goal wasn’t to make dies as fast as we could, but to make a die that served a customer and enhanced the entire system. Viewed this way the decision to spend capital dollars and to slow a process down was easy. By the way, this was a theme that was repeated throughout our improvement efforts.
If the changes you make increase the value for someone who uses your output or for your customer, then you know you are making a difference.
RP: It’s not the easiest concept to accept in our busy every-day lives, is it?
HS: It was tempting to look at the heading process from a traditional SIPOC (supplier-input-process-output-customer) view. It has but a few suppliers, only a few inputs and the processing is easy to see. The output is measurable, but it’s the customer concept that needs to be understood.
It’s at this point in the improvement process that it was necessary to step back and understand the role of this one operation in an overall system. It is here that the temptation to simply fine tune all of the inputs and maximise the output of one process had to be resisted. More ‘pieces in the pan’ was not the answer; better pieces in the pan was.
It’s true that ‘more pieces in the pan’ would have reduced a bulging backlog, and reduced the throughput time for this step – both traditional goals in an improvement process. But had we viewed the operation in isolation, whatever gains made in this step would be more than offset by losses downstream.
RP: From a practical standpoint, what other measures did you take to try and improve your operations?
HS: Several improvements were made that resulted in a better part: tighter size controls with parts closer to finish size or at finish size and with additional features added at the heading stage; brand new equipment purchased that introduced new steps to better prepare the raw material; new equipment to better polish heading dies. All of these added time to the heading step.
Had the goal simply been to reduce setup times and increase parts per hour, none of these changes would have been considered. Downstream steps would have soon encountered parts with excess material, unfinished features and far greater variation in critical diameters. And more of them too.
RP: Did you leverage results to gain traction down the line?
HS: Doing something that is a bit more difficult but benefits others down the value chain encourages you to think that way upon developing the next idea. Nothing proves more effective than results: the first steps we took were successful, which was very important, but we had to change the department internally if we were keep up the good work.
When you are sent in to increase production, the first thing you are tempted to do is to run the easiest parts that you have, and to run more than you need. The temptation is real, but you have to ask yourself what that is going to do to the rest of the system. At MSA we decided to change our scheduling methodology to make it more difficult for the heading department to fall into the trap. Changeovers, for example, happened more frequently.
The purpose of all of this is to make people understand their impact on others. You need to appeal to that part of human nature that wants to be part of a team, and extend the definition of a team from the three people in your department to the three hundred people in the organisation.
One’s department cannot be one’s world, and it’s important that our people start seeing our customers and material vendors as part of the value chain, too. At one point, operators in the heading department thought one of the issues was raw materials. We spent a year trying to analyse the problem and speaking with our vendors’ engineers, and it turned out that the problem we were experiencing did not generate from the vendor. While it is quite easy to blame the guy who doesn’t work in the building, pursuing this potential issue was the right thing to do.
RP: Why is that?
HS: Showing operators that we listened to their suggestions was critical, and it helped strengthening the idea that the system doesn’t start in our department, but the day they mined the nickel out of the ground and doesn’t end until the bolt is in the jet engine. This approach allowed us to show our department that we recognised that material vendors are part of the process too. It’s all combined.
RP: The question “What’s in it for me?” must have popped up quite a lot…
HS: It used to at first. It’s very tempting to introduce a change that will make your work easier but somebody else’s more difficult. In heading, for example, you could change the dimensions of the die you produce, to make the head form easier and to boost production. However, making the part bigger is not so great for the guy down the line who will have to grind that material away. This is what Taguchi was talking about when he described the cost to the system when moving away from a target, something we learned in the time we spent with the In2:InThinking Network.
We were adamant: the heading department would make the part exactly as it needs to be. To what extent are we willing to ask someone else to suffer to make our own job easier? The answer is simple. We are not. What we do, instead, is trying to find a way to make a part that is higher quality and requires less work downstream. We are now willing to spend more time and energy in the heading department to make things easier for others, even though this requires an extra effort from our part.
It’s important to note what we didn’t stress the important of simply reducing variation. Having a statistically stable process is fundamental, but I find that only pursuing variation reduction narrows your vision to a single process. You always have to ask yourself how you can do things better, how you can make a better part. That’s different than reducing variation. It’s when you view the whole system that you understand the benefit of making a better part.
RP: Looking back, what were the benefits of adopting a more ‘horizontal’ approach?
HS: The direct benefit is that we can make a part that requires fewer subsequent steps and we are able to form a head that is very close to or exactly on size. We have eliminated machining steps, and increased velocity as a result. An even bigger benefit is a measurable improvement in on-time delivery.
Our goal all along was to deliver to the customer on time: by spending a bit more time, energy and effort on the first operation, we completely eliminate the need for two or three more steps in the process. Going from perpetually late to always on-time and at the same time becoming more efficient has also allowed us to move into new markets as machine time and capabilities increased.
RP: What input did you receive from leadership? Was there support?
HS: Five years ago customers were calling to ask for a follow up on their orders and we often found out that we hadn’t even started working on them yet. We needed to fix this. The market has moved towards shorter lead times, too. ‘We don’t know how you are going to do it, but we need to make parts faster,’ leadership told us. We soon realised that making a better part actually resulted in making a part faster.
Support from leaders was complete and unwavering, also in financial terms. They had the faith and they were willing to spend money on new technology.
RP: In general, why do you think many lean implementations start by focusing only on one part of a process? Is it lack of resources or time? Is it because starting in one department is perceived as easier?
HS: For us it was important to start where real opportunities existed and to avoid the temptation to go where it was easy. We wanted to tackle real problems and achieve lasting change. To achieve that, we needed to stay away from low hanging fruit and get ready to jump into the deep end of the pool to make a real difference.
RP: But don’t you need low hanging fruit to gain traction and support?
HS: You need to deliver a slow, steady, and consistent message that addresses a real problem. When you are done, you should have something real in your hands. In our case, a better bolt. You get traction when you have something that is tangible, not when you get low hanging fruit. Such quick gains are only good in brochures.