Eliminating the barriers between inspection and production with FARO’s measuring devices speeds up manufacturing response times and increases car reliability at Vodafone McLaren Mercedes F1.

By Brendan Coyne, QMT, April 2011

“Speed of what we do is what it’s all about,” says Simon Roberts, operations director for McLarenRacing. Simon is responsible for all the operations, from design of the F1 race cars, manufacturing, supply chain logistics, quality and service of the parts, getting everything out to the track. “To put it into context, if we take all the design changes, all the updates we do, and averaged it throughout for the year, it’s a design change every twenty minutes on a 24/7 basis. Basically, the car is never the same one week to the next. There’s always something new, something different. It may just be something small or it could be something major such as a complete change of bodywork – such as new front wings or new rear wings. This is where measurement technology provides accuracy and good data for the engineers to make decisions.

One of the very important things, particularly with the FaroArm and software, is that we can do things really quickly. There is no point in us having elaborate inspection processes that take 3 or 4 days of time. We are not going to wait for that. We would end up skimping and just measuring a few key points. With the non contact scanning devices on the Faro portable arms, we can actually see the surface. Without a laser scanner, its much more difficult to see the surface and interpret that data. One of the things we’ve been talking about at Vodafone McLaren Mercedes is moving the point of inspection to the job rather than inspecting in a separate department. There will always be a place for a CMM for small parts, racking them out – it’s more productive. But when you are measuring a chassis or patterns or moulds, if you can measure it in its fixture – then you can probably do something about it immediately.You’re not re-jigging it. You can get the engineer there. That’s how I see inspection evolving.

Workshop inspection
As mentioned, we’re at 24/7. All the composites, all the manufacturing, are four shift systems, day and night and over the week end. We just haven’t stopped. So it’s all its about time compression. Every hour a job’s delayed is an hour out of lead time.

About 2 years ago, composites (inspection) used to be in a room on its own, as is mechanical parts inspection currently. We’ve deliberately moved composite inspection out of its room and put it in the shop where the parts and the rest of the department are because we got fed up of having those conversations in production meetings: “Where are the parts? “ “Oh, they’re in inspection.” If you put inspection in the workshop, then the parts aren’t in inspection – they’re in the workshop. You then expect the production manager to know because it’s just another production process – whether he’s trimming a part or inspecting a part, its all under his remit. This works really well.

Having the Faro arms removes the barrier between production and inspection. That’s a huge step. Take a wall down and it makes a huge difference to hearts and minds. Some of our production machines are vast – 4 metre cube volume. With a portable measuring arm, operatives can walk into them and check a chassis whilst it’s on the machine without moving the chassis. You can verify it before you remove it, just in case there is a rework issue. This provides an independent check from the machine system. If there is something wrong with the production machining, which utilises a measuring head to check the part, you are simply going to repeat the error. Whereas, with the portable arm as an independent measuring system, you turn up, find the datum and measure to find the error. Because you can go round the part with the arm and keep re-datuming, it’s perfect for our complex shapes. This inspection advantage applies to both finished part and interoperational
checking.

Data driven inspection
We manufacture the chassis, transmission, everything except the engine which is supplied by our partner, Mercedes-Benz HighPerformanceEngines. For parts such as chassis and wings, it is essential that we know, and are confident about the shape and form. Without data, you cannot do this. Mclaren are a very data driven organisation. When the car is running on the track, we record magabytes worth of data. And that logic is also true when we are making a part. We don’t want to take off-the-cuff decisions about what type of tool to use. We want to know what the right tool is for a wing assembly, or the right tool technology for a chassis, and we only know that if we understand the limitations of what we do.

The right tool is defined at the design for manufacture stage. So this data, this knowledge has to feed back into design for manufacture.” Quality control manager Keith Holland explains, “Traditionally, metrology has been a big safety net set under a manufacturing area, catching all the bad apples as they go through. What we try to do here is to build quality into the start of the process.” Simon Roberts continues, “Bear in mind, we make very low quantities. So for us, sixoff could be the full quantity of a component. For example, we make only four chassis or six wishbones and then the design will change. So you haven’t got time to do traditional statistical process analysis. We do stats analysis, but you have to do it with small sample sizes and you have to understand the limitations of that. Our approach tends to be more generic. What we are trying to do is to understand the yield of a process in its totality. We want to know what the yield of wishbone manufacturing is. When we identify specific issues, then we’ll change things – process, tooling or techniques – to eliminate them. If you are only going to make six components and you make 3 scrap ones, then you’re having a pretty bad day. If you make 20,000, three scrap are in the noise. There is variability – we make mistakes, machines don’t exactly follow CAD, material is not always what you thought it was, things go wrong.“

Keith Holland comments, “A lot of parts have a life cycle, so we need to build a certain amount in the knowledge that we have a certain amount of kilometres to fill through the year.” Simon Roberts continues, “Like aircraft, all the structural parts are serial numbered and they are all “lifed”. They all go through various proof testing and NDT operations as part of their manufacturing process. Part of their servicing is to bring them back into the factory and have them re-inspected from an NDT point of view and reproofed.

Proactive measurement
The drive for data has evolved over quite a long period of time. Certainly, in the last 8 years, inspection has moved from being a safety net to being a proactive force with the organisation. The resource restriction, which came into effect last year, has also caused us to look long and hard at how we spend our money. In the past, we have sometimes been able to buy our way out of trouble: if the process wasn’t very robust or if scrap parts had been made, you just bought more. We can’t afford to do that now. We are regulated to how many people we employ and how much cash we spend. So spending money on bad parts we’ve spent the last two years raising our game in terms of our own understanding, and the engineers understanding, of what yield is all about and what the benefits are of having robust processes that give you exactly what you want when you want. It’s about making parts right first time – zero defect manufacturing is the goal.

To process FMEAs, which is a great help. It gets you thinking in a different direction. We have a very good mechanism for recording faults. For faults seen on the race car – for example, if something doesn’t work the way you expected or fails before you expected – we have a very sophisticated mechanism for reviewing that. It’s a real-time close loop system which leads right back to design through manufacturing. We also have an internal non conformance system for parts we make and for parts we buy in, which is very proactive. It looks to eliminate waste in its broadest sense – in time, material or cash. Together, all this forms part of what you would call our quality system.

Hit it hard
Typically, every year, we have some two or three really difficult quality problems that are not just as simple as redesigning a part, or changing the source, or changing the manufacturing process. You might not even understand why a part doesn’t perform the way you expect. Over the years, our experience has taught us that if you look back in hindsight, you can often see early signs of those things. As a result, we call them Achilles Heels. They are things that you thought were OK but, in hindsight, weren’t. They normally end up being very long running and expensive. So our approach now, when we think we’ve got something like that – when we get that uneasy feeling when we see all the data – is hit it hard, hit it with really good people who have got all the skills, tools and techniques and don’t let it go till it’s fixed. A good example of that was when we lost the World Championship in Brazil 2007, where we lost gears for 30 seconds on Lewis Hamilton’s car in the last race of the year. Eventually we found it through design of experiment and sticking with it, looking at data and trawling through everything you know and everything you think you know.

Six sigma?
Six sigma was born out of the electronics industry, with companies such as Motorola. Today, these companies operate at 8 and 9 sigma levels. McLaren aren’t a six sigma company and there’s a reason for that. We do have six sigma black belts and green belts. You could very easily convince yourself that, if we were six sigma, all this stuff would go away. But there is a problem with that and it’s to do with complexity. When you multiply sixsigma quality by our levels of complexity, you very quickly get to a condition where there will be a determinate number of failures. We have studied that in depth. When we looked at our processes for the assemblies we make, we actually run around four sigma, but individual parts are generally in excess of six sigma. It’s that fact that you multiply up – if you build a F1 gearbox with six sigma parts, you don’t then get a six sigma gearbox assembly. So we can’t sit here and say our race car is a six sigma product. For us, a defect is anything that means a car doesn’t perform to the ultimate. Reliability is half the battle, you’ve also got to create a fast car and have the skills (strategy and drivers) to race well.”

About Vodafone McLaren Mercedes
Since its foundation in 1966, the McLaren Formula 1 team has become one of the most successful and groundbreaking teams in grand prix motor racing. McLaren Racing is the nerve-centre of the racing operation, housing the extensive design office, 60-per-cent-scale windtunnel, manufacturing plant, fabrication facilities and build stations that dovetail together to create some of the most iconic Formula 1 cars in the sport’s history.

The McLaren Formula 1 team has won four consecutive Drivers’ and Constructors’ Championships (1988-1991) and secured eight Constructors’ World Championship titles, the first of which was in 1974. Today, the team has a record 12 Drivers’ World Championship titles.