Lean fighting machine
E&T looks at how lean is helping Boeing in their production of Apache helicopters.
At the Boeing facility in Mesa, Arizona, close on 5000 employees work on manufacturing and servicing the Boeing Apache programme. The Apache Longbow, or more correctly the AH-64D, is a multi-role combat helicopter that features fully integrated avionics and weapons, plus state-of-the-art digital communications capabilities that enable real-time, secure transfer of battlefield information to air and ground forces. Apaches are in service with the US Army and 10 international defence forces worldwide.
You would have thought that, after building over 900 Apache helicopters, Boeing would have all the wrinkles ironed out of its manufacturing processes, but nothing is further from the truth. The Apache programme began over 30 years ago under the eye of Hughes Helicopters, then, after acquisitions, first McDonald Douglas and, in 1997, Boeing.
“As we built each aircraft, every one took longer in terms of cycle time, in terms of hours – and our quality numbers were heading south,” Jim Luby, quality and lean enterprise site manager at Boeing Mesa, says. “And we had a need for additional facilities because the contract that we were awarded said that we could ramp up to 8,10,12 a month but we were only building one or two a month and we had a factory full of helicopters.
“The reason we got where we are today is because we steal shamelessly; nothing proprietary obviously, but stealing best practices. And really that is a key. It’s really about doing things that others have proven and taking them to the next level in your own businesses.
“When the company merged with Boeing it gave another impetus to the lean initiative as on the commercial side Boeing was already heavily into lean. Prior to Boeing, we learned a little bit about lean. We would run events, run five-day events and then go back to work on Monday. When we became Boeing, they said no, a little twist, you have still got to run a one-week event; you're calling it an accelerated improvement workshop (AIW). The only thing is that, on Friday, you are going to implement 90% of what you do.”
When Luby recalls the state of the manufacturing process back in 1998 he recalls that there were 22 aircraft in line for a two aircraft per month delivery. Looking at the line some people were impressed seeing so many helicopters, but what Luby saw was a huge amount of dollars tied up in inventory; big time dollars.
“The other thing that struck me at the time was the lack of mechanics working on the helicopters,” Luby says. “They don't build themselves, but mechanics spent most of their time off the line collecting parts, tools or drawings. That doesn’t mean this is a prison and that you have got to be at the aircraft every second of the day, but the point is, if you are going to be efficient you have got to be the best cost producer at the highest body product in order to continue in business.”
There were several problems – a rigid pulse line that took an hour to move the entire line position to position, weak processes, crippling overtime and high support ratios. “Weak processes are really key; before you can get into things called standard work you have to go and stabilise your processes,” Luby says. “Or you are trying to standardise something that isn't stable and you are all over the map.” That was an early lesson that Luby learnt the hard way. As was the high cost of overtime that was well over 20% in the early days; now it is between 3% and 5% and that only because of the demands of a mixed model line.
Support ratios were also very high with almost three support personnel for each mechanic, a problem made worse by am ‘us versus them’ attitude. “Guess what, the people who build the product rule and, until you start thinking about that, you are not going to get to the next level,” Luby says. “We look at the mechanic, the operator, the technician, whatever you call it at your individual manufacturing sites, as the surgeon. Put yourself in an operating room; would you like the surgeon to be running around all over the place looking for sutures and scalpels and everything else?
“So we employed that same kind of methodology and we measure that. The surgeon in the middle is the operator. Does he/she have everything they need within their work area to conduct business, at the starting bell and at the closing bell? That means tools, parts, paper, everything.”
The early vision was that, with no extra bricks and mortar, no extra people, they would achieve a 50% reduction in cycle time and a 50% reduction in total billed hours. Like any good movie production it all began with models. The Apache was green painted wood and all the other jobs on the line laid out using anything from Styrofoam to waste paper. “We worked on what we wanted this line to look like; then we grabbed that vision and worked on it to perfection,” Luby says. “Now when I say perfection, that doesn't mean 100%. Perfection means to execution and there may be an 80% rule. Quick and crude is better than slow and elegant. Once you get the 80% rule, it’s time to march forward and implement everything that you have done and everything that you have talked about.”
With the vision created, the next step was communicating this to every employee, and according to Luby you cannot do enough communicating. “You launch the vision, you watch it start to execute and guess what you are going to find out. Maybe one thing didn't work or it worked great and you go and modify your approach. So make sure that you communicate with everyone, at every stage.”
A change for the better
The difference when you walk around the factory today is impressive. The first impression is the number of people actually working on the aircraft, and on the newly created feeder cells. These are shops that used to be called back shops where they build up some assemblies that would become basic modular units to be installed on the aircraft; they were always in the same facility, but in most cases were a long way from where they were actually used.
“We had one guy back there and he built up nine or ten engines, it was great,” Luby explains. “No one knew what to do with ten engines at a million dollars each; we can only install two at a time. But he kept building, but why? His schedule said ‘build’ so he kept building and he was rewarded for that kind of behaviour because he met 100% of his targets all the time. In the meantime we had jet engines sitting here and when there was something wrong with the engine. Guess what? Rework meant cost that could not be charged to the customer on a fixed price contract. So we built feeder calls, because you want to think of building stuff in modular units as close to the aircraft at the location where it will be installed.”
Another early move, and probably one of the biggest cultural shocks, was to locate all of the support cells right alongside the aircraft. That was not only manufacturing supervisors, but also all of the engineering, supply, production and control. “There is nothing worse than having an operator or mechanic call up needing a part and they are in another building and you might see them that afternoon,” Luby continues. “When you start to put out eight aircraft a month, you cannot afford that.
“Now comes the cultural shock. First you have an engineer away from the engineering building, which upset many of them. The second thing that was even more shocking was that they took their day-to-day direction from a manufacturing supervisor. The manufacturing supervisor doesn't tell the engineer how to do his or her business, but indicates when and where they are needed, to support immediately any issue. We got over that rather quickly, but it really was a cultural shock. This is the key thing that we talk about. Lean is taking people where they cannot take themselves. It is the ability to translate vision into reality, from Styrofoam models into where you want to be when you grow up.”
A key improvement tool was the accelerated improvement workshops (AIW). Luby recalls one particular event that yielded huge rewards. “I knew it was fate,” he recalls. “A process was so screwed up; my target was 75% reduction. It was really bad. It almost failed because people didn't even want to even sign up to a 75% reduction. I didn't make it. I made 72% reduction. I could have set a standard 50%, but I knew that this process had so much non-value added in it, we should go for 75% because if we get 60% that’s great, because we are only shooting for 50%. We got 72%.”
To add to the complexity at Mesa they do not just build for the US government; they have a mixed model line to serve several other countries, as well as a remanufacturing programme with all the 937 A models slated to come back. They are completely stripped down, a third of the parts are thrown away, a third sent back out to the field to support the remaining A models out in the field and the final third are inspected, repaired and put back on the aircraft. For all the other countries, with the exception of Israel and Egypt, new build means that they got a brand new fuselage.
Take the Israeli version, for example. Because of the different options required by Israel the aircraft requires over 1000 hours more work for the same helicopter than on the US government, but they all go down the same line together.”
The tact time, or, as they refer to it, the attack time, is 2.75 days, so the line moves in five minutes every 2.75 days, with all these different models on the line. “So let’s take it back to our motives,” Luby says. “That’s why Toyota is putting different models down the same line. They all have different packages, yet they all move at 2.75 days. The reason we are successful in doing that is it takes a lot of up-front planning. We don't just get an order in from a country and have to start building tomorrow. There is some lead-time associated with the uniqueness of those helicopters. But what we do is planning up front; so that we know when that aircraft drops in we know what is going to happen. The key here is the agility and flexibility of a lean production line allowed us to go and do that without any disruption to the customer deliveries.”
Every 2.75 days, that line goes to the next position. Why 2.75 days? The time, in true lean fashion, is calculated from customer demand. “We have the positions in line in order to complete the work package in 2.75 days and it’s ready to move to the next position, which is a new set of mechanics and a new set of work to progressively build the line. When it leaves that building the aircraft is ready to be painted and undergo flight tests and go fly. So that’s what we call a pulse line. It’s not a continuous moving line like it would be on a chain, like at Toyota. It pulses every 2.75 days.”
The results are not just stunning visually but make equally compelling reading, the most startling an 85% reduction in hours per aircraft in final assembly. “Some of that was easy, the proverbial low hanging fruit, we were so inefficient that it was pretty sad,” Luby concedes. “We cannot get better until you are ready to go and take the medicine; that means roll up your sleeves and give me that shot. I use that saying, the medicine is there, the fix is there, its called lean and all its tools, but until you are ready to go and take that shot, you cannot get that much better.
One of the key lessons learnt was to focus on the process and not the product. “It’s a process and that is one of the things that we learnt early on,” he says “We did get some training by going to Japan; I was offered it through the Boeing Company. We went to see the Hitachi Air Conditioning facility.
“My boss, who is the site manager here, he looked at that, he saw these big air conditioning units and he said, ‘Jim what do you see about those?’ And I said, ‘well they are big air conditioning units with parts and they go down an assembly line’. He said, ‘exactly. They have metal, they have wiring, and they have tubes and so on. What does our helicopter have? They have metal, wiring, and tubes. Why can't we go and do that?’ That’s what helped my boss to come up with the vision. This is what we are going to go and do, and we can do it. Don't accept ‘can't’ – I have said that enough – but I will say it a thousand times.”
Other key elements of success Luby cites are strong leadership and involvement throughout the enterprise. “Set the tone in urgency and sell the way, coach and demand,” Luby says. “Coach first and, once the expectations are understood and they are executing, great. If they are not executing then you demand – that comes in the other edge of the double bladed sword.”
There is a culture at Mesa of what they call high-performance work teams, something often called self-directed work teams. Over 98% of the operations is on high performance work teams, which at Mesa is a four-stage maturity model – form, storm, norm and perform. “We are truly the stage four mature team, which runs its own little individual teams as a business,” Luby explains. “They are responsible for working to their budgets, scheduling their vacations, dealing with any issues without supervision. The supervisors in the area are there really for knocking down and getting rid of issues that prevent that team from being successful. Not doing the day in day out activities that they used to do.”
Increased quality is an important by-product of a lean organisation, and at Mesa quality was not only improved, but it was achieved with less manpower by a process they call manufacturing self-examination. They used to have an additional 87 inspectors on the line who are now working as mechanics. There is still a team of 16 monitors who audit the process. “What is significant about that is that when you stabilise your processing, you measure it, you now exactly what it takes to go on with the entire process,” Luby says. “In terms of paperwork and so on. I don't have a reduction in my requirements for the government, but we streamlined how we report it and the metrics that we capture.
“There are two other levels. Some critical processes get a second set of eyes, and another, MSE. Some that the customer deems as flight safety still have a traditional type of inspection. One of the auditors of the 16 auditors puts his inspection hat on and inspects that product.”
*Please note: This article originally featured in the June 2006 edition of Manufacturing Engineer magazine