Get a manufacturing six pack

Millions of dollars have been saved by GE, Motorola, Dow Chemicals, Allied Signal and others thanks to their adoption of the ‘Six Sigma’ quality philosophy, which famously aims to reduce rework and waste – by rooting out the ‘hidden factory’ in almost any manufacturing concern. Inward facing it is, but the rewards are undeniable and big for all that. This all too often misunderstood methodology, which takes its name (but thankfully, little of its practice) from the statistical ‘six sigma’ acceptable variation envelope (effectively, 3.4 defects per million opportunities (dpmo)), is one of the few non-regulatory quality-based initiatives that deserves much more attention than it’s getting in the UK. Author of ‘Six Sigma: SPC and TQM in Manufacturing and Services’ and ‘Design for Six Sigma’, Geoff Tennant says: “25% of everything – floor space, labour, electricity and so on – is being spent on inspection, control, rework and process improvement.” That figure invariably comes as a surprise to far more in management than it should. And so does the next: although no-one is pretending you can expect to remove all of it, the plain fact is that in companies that properly adopt Six Sigma methods, it certainly can fall to around 5%. There is the potential here for ROI (return on investment) of a full 20% of total annual expenditure. Accepting that this is bound to vary widely, even so that scale of cost reduction is certainly not to be sneezed at. So is it applicable to you? Six Sigma originated at Motorola 20 years ago, when the electronics giant was facing stiff competition from its Far Eastern competitors, and it’s true that to date most success stories have been associated with large multi-nationals across a range of industries. But don’t let that mislead you. There’s something in this for many, much smaller manufacturers, even if for most of us it’s only used as a mechanism for focusing anew and better on factory issues and processes using its structured and measurable change approaches. “Everyone has known since the 1920s how to measure processes,” says Tennant, “but now you need to be intelligent about it, to fit it into corporate strategy and let everyone understand what you’re doing.” And that’s where the Six Sigma methodology scores, and why it can work at several levels, crossing functional and business size boundaries. “Six Sigma will stretch from basic strategy in the boardroom, and give people something to do on a day-to-day basis that means they are working towards that strategy,” he says. This is so pragmatic Arguably, it is the practical aspects of Six Sigma – rather like a TOC (Theory of Constraints) initiative – that set it apart. They certainly seem to sit comfortably with those at the sharp end that have done it. But there’s also the supporting software available today that helps to cement it in and keep it going. As Harry Levy, continuous improvement manager at Leyland Trucks, notes: “What makes Six Sigma effective is the way tools have been pulled together into a single project management system, with a logical problem-solving roadmap, and the availability of PC-based statistical software, which makes the stats easier to use.” The software also makes it is easier to understand, work with and disseminate, which is vital to the success of any initiative that’s going to involve a lot of people in an ongoing improvement project potentially spanning several departments and locations. Pity really that Six Sigma gets confused with its (included admittedly) six sigma statistical namesake. As Ralph Rio, director of analyst ARC Advisory Group, observes: “Six Sigma is a very specific, rigorous business practice, but because it involves statistics some people have built a lot of complexity around it.” And ultimately that is anything but helpful. Reassuringly, Tennant’s overview of the methodology is anything but difficult to grasp: “Find out who your customers are and what they want, identify your processes and make them better,” he says. We would all vote for that and indeed, once explained, the methodology is, at the top level at least, quite simple, consisting of just five steps – Define, Measure, Analyse, Improve and Control (DMAIC). Define means seeking out process defects from a customer’s viewpoint. Measuring it is self evident, involving the usual mix of data acquisition aimed at identifying, quantifying and verifying the causes. Then you analyse the data which needn’t be too complicated. As Rio says: “A lot of the data analysis can be done with an Excel spreadsheet and simple Pareto analysis; rarely do you need to get into a lot of the complex statistics that are taught in some Six Sigma programmes.” Then, improving the process to reduce the defects at step four speaks for itself, while control – conventionally the last step in Six Sigma – consists of monitoring the improvement made to the process so you know it’s sustainable. Part of this, obviously, is sharing the lessons learned to spread and maximise the financial gains. And put like that, the mystique vanishes, doesn’t it. Because there is no formal certification, Six Sigma is generally interpreted and adapted to suit an organisation’s specific needs. That being the case it’s worth looking at some examples. At defence manufacturer Raytheon Systems, the Six Sigma programme is based on five tenets, according to Neil Roberts, Six Sigma champion. “They are: understand value in the eyes of the customers; understand the process by which you deliver that value and eliminate waste and variation; make value flow at the pull of the customer; involve, align and empower employees; and continuously improve in the pursuit of perfection.” Note the subtle differences from basic DMAIC and the focus on customers and employees. 5% yield improvement Already, Raytheon has notched up substantial achievements. “In semiconductor manufacturing at our Glenrothes plant, for example, we reduced process defects from 6,000 dpmo to 200 dpmo,” says Roberts. “That’s a 5% device yield improvement, and gives savings measured in hundreds of thousands of dollars.” In fact Raytheon has extended the remit of its Six Sigma programme to include cutting lead times. “The time-value map is one of the most illuminating methods of data collection,” says Roberts. “90% of the time a part spends in a plant is likely to be ‘non-value adding’. If you want to shrink lead times, that’s the time to cut.” It’s a common enough observation, but you can see how the focus and the methodology help organisations en masse to get away from what ‘they’ve always done’ and move towards what they should be doing. Meanwhile, for Leyland Trucks, the strength of Six Sigma is in its emphasis on data-driven problem solving. “Six Sigma makes you follow the actual data and can lead to some surprising insights into processes that we thought we understood,” says Levy. The firm’s inventory saving of over £2 million makes the point. “Prior to Six Sigma, we sought to limit the disruptive sequence queue changes we were making 10 times a day by using buffer stocks,” explains Levy. “But during the course of the Six Sigma investigation, the green belt [one of the levels of Six Sigma training] demonstrated statistically that the number of changes didn’t correlate to buffer stock at all – so we could reduce the total size by over 90 vehicle sets.” Another manufacturer – a customer of Corel, which sells the iGrafx process management and modelling software – unearthed a similar surprise, according to Tim Nicholls, product marketing manager. “The perception was that severe bottlenecks were due to not getting raw materials in on time, and that the answer was some sort of supply chain management solution. By applying Six Sigma and our modelling software, in just one afternoon they found that the true bottleneck was in the packaging process.” Decent data? Digging around, these are not exceptional stories, but results like them can only be achieved reliably where you have accurate information. As Levy says, “Before you go too deeply into a project, make sure that you’ve got decent data.” In fact, data collection and analysis don’t begin in earnest at Leyland Trucks until after another phase – measurement systems analysis. Jim Stephenson, senior business manager at software developer Minitab, puts it like this: “The variances you measure may consist not just of [those] in the end product itself, but in the people and tools that are taking those measurements.” Rubbish in, rubbish out. Still sceptical? Dow Chemicals is another big one, racking up savings in the billion-dollar category thanks to Six Sigma, and the key point here is that financial scrutiny of all projects at the company is legendary. “On every project team there is a financial representative, and all financial aspects have to be signed off,” says Martin Hodgson, Six Sigma black belt at the firm. Dow is already more than half-way towards the target it set itself in 1999 – of $1.5bn earnings by the end of 2003. No doubt part of the reason for its extraordinary success is that Six Sigma itself recognises financial measurement as the key goal. But before we go too far down this route, remember: savings don’t always come from cutting measurable costs. There can be significant intangible benefits. As Mike Burkett, research director at analyst AMR Research, observes: “Another way of looking at [improvements] is savings tied to strategic initiatives. For example, improving on-time delivery and speeding up new product introductions may be difficult to tie to a dollar value.” And potential for savings isn’t restricted to manufacturing processes either. One process that everyone undertakes, but is extremely prone to error for build-to-order manufacturers, is leads-to-order (LTO) – needs analysis, recommendation and refinement, pricing, quotation and order. “We had one customer who had a 98% error rate on orders,” says Klaus Besier, CEO of software developer Firepond. Fixing these errors is an expensive business. “Another customer spent on average $10,000 per order on engineering review to correct the errors – that’s $20m every year,” says Besier. Although product configurators partly solve this problem, they do not address the LTO cycle in its entirety. Six Sigma global transport manufacturer Bombardier acquired Firepond’s SalesPerformer suite – a package that addresses LTO process problems – and now gets 100% perfect orders. The firm has also turned to more sophisticated software to address processes at the opposite end of the manufacturing cycle – design and design for manufacturing. Tecnomatix eM-TolMate simulates how different tolerances amongst the components of design will impact successful manufacture. Thomas Edwards, director of advanced engineering at Bombardier Transportation, says: “By earlier validation of our design tolerances using eM-TolMate, we aim to get 100% of parts to fit right-first-time every time. This will result in significant savings in assembly times and therefore costs.” Relevance to Six Sigma? Eric Gautier, European marketing director at Tecnomatix, explains: “We simulate assembly of parts with different dimensions, and then determine those that we need to pay attention to in order to increase the percentage of successful assemblies. Out of 200 with corresponding tolerances perhaps only 10% may be really important for its manufacturability.” So there’s another area for attack. As he says, “The tendency is to specify relatively tight tolerances for components to ensure quality products. But having high tolerances is costly – so it’s important to know which dimensions are relevant and which aren’t.” Companies not in Bombardier’s league are unlikely to consider six figure software licence fees, but that doesn’t mean wide ranging aspects of Six Sigma aren’t accessible to them. The Dow Chemicals plant where Hodgson works employs just 60 people, and although he does benefit from group support, he is confident that the methodology and its approach to solving plant and business problems can be applied to smaller companies. There is, however, one problem. Ann Esain, senior research fellow at the respected Cardiff University Lean Enterprise Research Centre (LERC) doesn’t believe that smaller firms could afford all of Six Sigma. “We have to think differently about how we deliver this to the general industrial population,” she says. One approach is that of the Welsh Development Agency. Six Sigma ‘black belts’ are currently undergoing training, and the agency is contemplating providing them for SMEs. Our advice: take a phased approach, perhaps training green belts and yellow belts initially and focusing on specific objectives before moving on. Whatever the challenges, nobody can afford to ignore this. Asking whether or not you need Six Sigma is akin to asking whether or not you need quality, which is uncomfotably close to asking whether or not you want to be in business.