Lean thinking: smarter ways to better business
By Brian Tinham

Manufacturing Computer Solutions - January 2003

Lean thinking: was there ever a more inclusive term? Most of us understand, or think we do, lean systems and methodologies on the factory floor, aimed at streamlining production, cutting out waste, doing things better and more efficiently with less. But lean' as a concept, a philosophy almost, by which to guide the whole manufacturing business? What's that about? How much is hype, how much reality? For that matter, since it's currently high on the business agenda, where does 'Six Sigma' fit in? Also TOC (Theory of Constraints) and the rest?

Fact is few now question the power of lean methodologies to help, in a structured way, to identify, challenge and then improve factory floor operations. It doesn't have a fancy name and there are plenty of reference sites where resulting reductions in key measures like materials, inventories. WIP (work in progress) and product lead times, as well as improvements in flexibility, responsiveness and productivity, arc beyond question. And much the same is true of the other main factory-based initiatives which have broadly similar goals and some overlap - indeed some TOC implementations are quite breathtaking.

Business results
In short, lean on the factory floor, although not quick to gain its financial credentials, is proving itself time and again through good solid business results. Hence, in times of economic difficulty, the appeal of extending its thinking into other areas of the business, like procurement, engineering design, financial management and even sales management, and getting comparable or even greater returns. Cutting costs there, and on the way improving agility by changing processes, roles and culture while also simplifying where appropriate using lean's classic 'value stream mapping' to establish priorities, would get most business leaders' votes.

That's why lean is one of the kings among strategic business issues today - especially as we march warily into 2003, keenly aware of the global commercial realities. However, those that go the lean route warn that while it need not cost a fortune, nor be that difficult, experience proves its success hinges on willingness, ideally enthusiasm, to change - and that ultimately, without those and a guiding hand, it might not be trivial to sustain improvement. They also say that users' manufacturing and business management IT has a key role. Which is also why development of our IT strategies needs to be conducted with clear understanding of the dictates of the ways of lean. Plainly, there will be some different demands on it, both as supporter and enabler of the new business processes.

It all needs thinking through. What are the requirements? What are the cost/benefits? In which ways is your IT likely to have to change? What are the pitfalls to avoid? What are the people and business culture implications? Who do you turn to for consultancy and implementation? Late last year we gathered 25 senior manufacturing business managers (managing directors, operations directors, production managers, IT directors and lean champions from across industry), some software vendors and a heavyweight bunch of lean experts round a table for our eighth MCS forum, and put these questions to them.

We charged all present with the task of coming to some conclusions to help everyone minded to go lean to do so most productively, aware of the implications both in terms of expectations and preconditions at every level. We wanted advice, observations and opinion to be freely shared for the benefit of all. And it was.

First though, definitions and background. Ralph Woodhead of consultancy Open Logistics and a "reformed MRP person", summarised lean as "a philosophy based around responsive customer focus, and about eliminating waste and pulling work through the supply chain, through the factory, most efficiently to meet customer requirements."

In doing so he also made the point that 'lean' hasn't been developed in isolation. It owes much to Japanese Just In Time (JIT) thinking and even to the whole MRP production planning and control improvement experience, as well as to TOC, the 'Five S' housekeeping improvement principles and the rest. And it's evolved alongside endeavours like Six Sigma, originally out of Motorola and with its heritage in quality initiatives like TOM (total quality management), reducing variability. All of which has resulted in some 90-odd complementary techniques using different treatments and with applicability according to your current manufacturing and business environment, industry, systems and people/skills, as well as the precise problems and objectives.

So, for example, TOC is nowadays focused on achieving quick, pragmatic improvements by starling from the position that factories and supply chain have what John Tripp. Goldratt affiliate, called "statistical events and dependent events". TOC, he said, is "trying to create predictability from an inherently unstable environment. Our [TOC] approach is to focus on the thing that's going to get the maximum return on investment (ROI), find the problem, build the solution, do the next thing, and so on. That's how we get rapid fixes."

And depending on how holistic you want to be, lean can behave similarly: even some of the terminology and goals can be happily mapped. TOC's 'bottlenecks' and 'constraints' are lean's 'pacemaker' processes; TOC's 'drum' in DBR (drum buffer rope) is equivalent to lean's 'takt time'; the logic of appropriate VVTP stock buffer locations is comparable. What Forum made clear was that there are few magic wands; there's a lot of common sense and ways to get to it. You need to think about what you want lo achieve and why, and then match the techniques and the thinking to that.

Right first time?
Sounds simple, but it matters, because there are choices. And while that's OK, getting those right is key to progress - and that applies to supply chain, product development and procurement improvements every bit as much as to the factory. There's no need to be dispirited: finding your way through the maze is actually about what operations directors and production engineers are good at - pragmatism. But it does need attention and there is an argument for some 'hand holding'.

Above all, Forum agreed that it's critical neither to be dazzled by any one set of techniques, nor to allow dogma to get in the way of sensible choice. Chris McKellen, of lean consultancy Manufacturing Awareness, although admitting his bias, insisted that Allied Signal's experience of Six Sigma, with which he was involved and which apparently saved it millions of dollars, could have been bigger, better and faster achieved had the firm selected much easier lean techniques.

His point: Six Sigma has got a compelling organisational structure - black belts, green belts and the rest - and it's centred on business metrics, and lean doesn't enjoy the visibility that confers. One sure way to get 'buy in' from the board is to talk in terms of improvements, growth and profitability, rather than 'removal of waste'. Talk of better, more efficient processes, rather than stripping out waste and dead wood.

Woodhead. "We did it very simply down at the factory level with a pilot scheme that showed the benefits and then got the attention of the guys at the top. That turned into a strategy for the organisation." At the time he wasn't a lean expert: he was working in what amounted to an SME division and he started by picking up the Lean Thinking book by Professor Dan Jones.

Getting started
All that said, how and where should you get started to make maximum difference? Can that be generalised? Forum thought not: we can generalise the kinds of issues and opportunities, and the likely fixes, but prioritising is very individual. It also rapidly became apparent that for some, solving one problem highlights another. Also, lean really is about continuous improvement; it is absolutely not a one-off project. As Jen Tyler, plant director at office furniture maker Steelcase in Margate, said: "the more waste you remove, the more waste you see."

And that's both a strength and a weakness. You can use the thinking and the methodologies for specific problems, ideally (but not necessarily) prioritised and built around a strategy to give incremental, relatively quick gains, which is good. But you've got to keep on going and not see that last lean project as the cause of the next problem. Yet at the same time, don't be blind to the limitations of the techniques you're employing and the problems you're trying to solve. As Tyler said: "I have a Steelcase Production System leader - a full time position - who reports direct to me, not through a production manager." The message: to get the best out of this long term you have to take it seriously.

Some around the table, for example, found apparent mutual dependencies as a stumbling block that in one case meant that what started out as a factory floor project rapidly mushroomed into something much larger. Graham Leake, business systems manager with Britax Aircraft: "We partnered with Boeing ... on the shop floor, but we soon found that the progress we were making on workshops, doing all the lean and waste initiatives, the 'Five Ss', was really hampered by our ability to get the supply chain working. So we then looked at the supply chain - and our ability to succeed there depended on getting the design criteria correct..."

One thing led to another, and Leake found that identifying and justifying resources to tackle the problems being exposed became an issue in itself. "We really had to stand back and start linking the objectives of the business to the actions." The message: think about the value you expect to be able to deliver from different activities, and their practicality, and prioritise accordingly - and be prepared for issues to crop up.

It's also fair to say that his experience was always going to be problematic because Britax's business is about low volume, intensive engineer-to-order streamlining production in that environment is going to be tricky. Lean is usually associated with high batch automotive style work. As McKellen said: "The issue is knowing where to start it off and then sustaining it." And that in turn should indicate the kind of approach and appropriate tools. If you're involved with design cycles closely linked to short run manufacturing, you might want to consider software tools geared to making the design cycles faster and more integrated.

Getting it wrong?
There is another side to this. Tripp said that finding the wrong starting point needn't be a disaster because, "the minute you find the wrong one, you know, and you'll find the real one much more quickly." And he adds: "The real problem is people spend weeks analysing it and not doing it." Unless, of course, you can't afford mistakes - most shuddered at the thought of jumping from project to potential project, particularly SMEs.

And Hugh Williams of the Hughenden consultancy added the inevitable political angle: "In any organisation you will find individuals who know something about one particular technique who may choose to pursue that - and it's better to let them do so than to do nothing." On the other hand, the purists would prefer a documented strategy because it provides an agreed structure and common understanding - and Tyler attributed much of her company's success in the US with lean to its strategy which, she said, was founded exactly on the acclaimed Toyota Production System.

David Brunt, continuous improvement manager at Porsche UK, felt a lean strategy is essential. "You've got lots of people with lots of agendas that may want to go in a whole series of different directions, and unless you can define exactly where your company wants to go then you've not got a very good chance of implementing any of this stuff." And as for who should 'own' that strategy: "Everybody should own it - which is a difficult thing, and very few companies have achieved it."

Tripp's recommendation was the Goldratt TOC 'Four by Four' in which heads of the various business functions go off site and, using TOC teaching in operations, finance, distribution, production and so on, are exposed to 'false assumptions'. "Over a four day workshop, their issues are analysed ... so they can build a solution together. It's a structured way to make people take strategy to tactics, to implementation."

Is that daunting. As David Lloyd, group systems director at Stadco, said: "You sit there as a management team and you think, 'we'll do this top-down' and you size it and the task looks huge. You also say, 'we are going to do it from the bottom up at the same time' and that kind of looks huge as well. And by the way, we have to get the bits out to the customer 24 hours a day, seven days a week, and that's kind of huge anyway!" Just so. "It's a great idea, but it's not so much where do we start - we can start anywhere. It's look at the size of the task."

Horses for courses?
There's got to be a balance. As Steve Baker, process improvement practitioner with The Manufacturing Institute, said: "What [SMEs] don't want is a 12 month plan to give them some gains. They want a very short term plan to give them immediate gains that then allows them to move on." Hence the different approaches. His recommendation: "Typically, we use 'value stream mapping'; it's very, very capable of identifying a problem within an organisation." And while he accepts that does sound a bit fancy, he says it can be as simple as drawing process maps "on the backs of very big fag packets". It can also be done graphically with business process modelling tools, but the result is the same - projects prioritised according to the value to your customer.

He wasn't alone. Brunt was emphatic: "That's where you've got to start. Unless you do that you can have all the shop floor improvements you want, go into your supply chain, go into new product development - but if your customers are wanting something different, all you are doing is building yourself a better dinosaur. So you've got to start with value [and] specify it through the eyes of the customer. That's going to prioritise things - not picking on one thing because you think it's a good idea or somebody's got an agenda for it."

Meanwhile, the big question from the FDs, eloquently put by Francis Gaffney, general manager at Goodwin International: "Will going lean lead to more sales?" And the answer, after some equivocation, was yes - and in spades. Mean improvements in the first year specifically after TOC factory implementations show around 60% improvement in sales, said John Tripp, citing an independent study by Professor ,. Vicky Madden in New Zealand. 'And you get inventory reduction of 50—70% and operating expenses and excess overtime down," ' because suddenly you're able to harness what he called your "hidden capacity". Beyond that, Lloyd cautioned all not to underestimate the intangibles. "Even if you don't see things on the bottom line, you may perceive improvement in the type of people working for you." And his point: you get a virtuous circle of improvement, as well as the flexibility and agility that comes naturally from a 'can do' culture.

IT in lean thinking
One of the fallacies of lean is that it's at odds with IT. Forum agreed that's nonsense. What is the case, however, is that some of it ought to change to match the support required for the new initiatives. So, as Woodhead said, you probably need MRP as originally conceived "to avoid all the stock you'd have if you had re-order levels throughout the whole factory," but not as it's so often turned out. Lean is about 'pull through' production, not 'push'; and in far too many factories the limitations of older MRP systems have led to the usual safety stocks and ultimately a mix of push and chaos.

"If the MRP reflects the fact that you are making smaller batches, that you've reduced lead times etc, then it can be a pull system," said Woodhead. "Lean thinking will try to encourage you to reduce lead times, but if for whatever reason ... you've got long lead times, then you need MRP for the material planning anyway." And since most today have an ERP system, or similar, for long term planning, financials, sales order processing and the rest - accepting the above caveat, for the most part, those aspects needn't change.

It's the short term scheduling, materials replenishment, shop floor data collection (SFDC) and so forth that needs attention. As Stan Stuart, lean manufacturing business champion at ERP software developer SSA Global Technologies, said, SFDC implemented solely to get round problems of poorly implemented or misused MRP (to see where things are) will come under fire.

But, if you want electronic recording for costing or control, that's different. Likewise, finite capacity scheduling (FCS) and advanced planning and scheduling (APS). And if you need electronic capture of customer demand, or want to operate with suppliers electronically, or run analyses beyond ERP reports, then you also need additional IT support. It's all about visibility - accurate shared real time information.

The IT for lean
Starting in production, Brunt made the point, that "the better the plan, the better the flow through the facility." And the same applies to factory floor information. For Porsche, he spoke of "writing an Access database for hand-helds, so we can click on all of the things required [for a car service] and sink it into the database ... and sort out the process up front... We've made big strides in terms of our fulfilment because we get the right information."

As Rick Sharp, sales manager at Soft Brands (Fourth Shift), said: "You've got to react to what comes in through the door, and have a support infrastructure that allows you to get that operation done as quickly as possible. To me it's that philosophy which is still missing in shop floor layouts. And at the top people need to get away from [too much] forward planning and become reactive."

Translating that to other industries, on the one hand, the closer we can get data collection to real demand the better, as long as it's then factored into a flexible and reactive finite schedule. And on the other, the closer we can reflect production status throughout the business, with analytics, the better will be the management - although there has to be a debate about how much SFDC you need, depending on throughput, Nevertheless, as Ron Bligh, senior applications consultant at ERP developer Mapics (now also the proud owner of Frontstep, formerly Symix), said: "The value of SFDC is you get a very fast route to seeing quality problems before the guy's finished making 1OO of them!"

Production monitoring
This latter is the argument for MESs (manufacturing execution systems), or more accurately production monitoring. Steve Carr, business development director for MES firm Mestec, likened electronic production management to having GPS in your car - you can see you're going in the right direction and react, or be proactive, around events. "It could be monitoring the fact that product has been consumed at this point, so it's ready to work on something else, or it needs replenishment... You need to know what's happening with people, materials and machines, not just quality. Without that information you can't make decisions."

As far as lean is concerned, the only debate here is whether that information should be presented visually or provided electronically. And Angela Coombes, lean enterprise executive at aerospace manufacturer UECD, was clear that we're likely to need both. "Visual for the shop floor where it's all happening, but. especially in aerospace where we've got to maintain material trace-ability, you don't want someone writing out batch numbers so barcoding is the way." And for her. SFDC is an every-operation activity, although clearly it's entirely industry and throughput dependent.

But Carr's point is a good one, and when you consider manufacturing with yield dependencies, or skills issues that MES can reveal, you see what he means. As he says: "If you've got a potential variable it's worth measuring if the measurement isn't going to significantly impact the overall process time."

Williams observed that you might want to be guided by the principles of TOC in determining what's worth monitoring and what not. Just as significant, Tripp urged us to collect production information, not. just for tracking and status, but so we can diagnose problems. "What we need is information about what is stopping production flow. I don't want to know that an assembly isn't where it should be: I want to know where it is, why and how long it will take to get here. That would help drive improvement processes." Yes.

And Clyde Bennett, support manager for Microsoft Business Solutions, urged users to remember that even information with little operational significance today, if available, can shape future production strategy. "If you're moulding plastic parts and you have a jig that tells you if the parts are good, you may find yourself five years later with the same technology in a whole different market where understanding the characteristics of those products is of value to the designers." There's more than one reason for data recording and storage.

Simulation systems
Another point. At a recent AME (Association of Manufacturing Excellence) conference on practical lean learning with Ralph Woodhead's 'Can do kanbans'. I was struck not only by the power of his technique to demonstrate production improvement thinking, but the difficulty for would-be users in a real factory. It's very easy to move pieces of plastic round on a table and redesign a process or a factory floor, but in the factory machines are where they are and moving them isn't cheap. So you might want to consider simulation IT.

Nigel Prue. managing director of Dassault-owned simulation and production modelling firm Delmia, found it easy to agree with that point. His systems, and those from rivals like Tecnomatix and also Simul8 and Lanner. let you build a graphical functioning model of existing or new plant and put 'production' through it to see what works best - before you move anything.

"You can move a machine from here to there at the click of a mouse; if it made it worse put it back and try something else; implement a pull system rather than a push system - all in a virtual environment. If you don't use simulation ... you really are taking a big chance." He reckoned Daimler Chrysler has made significant savings using Delmia for line balancing for example, and claimed typical savings of around 5% in workforce. Again, look at the cost/benefits.

While on the subject, he also argued the case for lean thinking to smooth progress between engineering and production by getting manufacturing engineers in the loop with information and IT. For him the requirement is a manufacturing database populated from engineering with a product structure so that process and assembly planning can be synchronised with new product development, change control and procurement. "Part of that might be defining operating procedures which are then available on the shop floor. Then at the end of product development you've got all your manufacturing processes defined and centralised - and many people can work on the process at the same time."

Steve Nevey, CAE manager at Jaguar Racing, agreed, saying that product structures and BoMs ought to be available to engineering and manufacturing, "so that production staff can start to identify what's coming through, how it's assembled and start making comments." Again, getting it right faster is all about early feedback and that in turn is about visibility of information in appropriate 'languages' and formats.

Good points, and Jon Jarrett, IT director at Pace Micro Technology, added: "New product development absolutely has to involve the ERP. Some of our components have 12 or 16 week lead times [so] if you find issues it affects the supply chain." In short, make sure that information is managed between systems, so that users in different departments don't make mistakes and lose confidence. And if you've outsourced manufacturing, ensure consistent supplier approvals too.

At the other end of the spectrum, Stan Stuart reminded Forum of the value of rules-based product configuration management software in terms of rationalising products into families with variants. "That lends itself to cellular manufacturing because while the processes and BoMs are similar, the tool - based on the generic product - creates new variants which can be linked to your CAD and PDM system. Eventually, you can move from a position where you are engineering to order, to simply configuring products directly." Another good point in the spirit of effort - and indeed now waste - removal.

Business processes and ERP
Moving on to lean transactional business processes, principal issues involve, for example, setting up relationships with key suppliers so that you're no longer sending purchase orders, raising invoices and the rest, but running electronic 'kanbans', so that admin is reduced along with WIP and inventory - for all parties. Exception management and event management IT then become valuable supporting technologies - and that can be as simple, or sophisticated as necessary, with the web being the undisputed big enabler.

At its simplest, Rob Evans, lean projects manager at Druck, now part of GE, pointed out that this can be a 'web cam'. "We had a 'supermarket' at the end of a production line and when it dropped below a limit the supplier looked on his web cam and delivered the next day." At the other extreme, there are web portals and exchanges or hubs - some operated by third parties like Wesupply - that handle complexity, lan Broughton. Wesupply marketing manager, pointed to their value in providing for collaborative forecasting and replenishment, while also alerting on events and exceptions, like quality issues at the supplier, inspection failures at the receiving end, or capacity or delivery problems.

They also help to reduce what lean experts refer to as demand amplification over a supply chain as ''individuals manipulate the data they don't trust," as Brunt put it, for safety margins. "By getting clear information down those supply chains (using web-based supply chain systems) a lot of complexity can be taken out."

Similar technology can be used to make tooling lean. Bob Shanks, site services manager for tool maker Sandvik Coromant, said his company will shortly be providing tooling to all Ford factories, assigned to operators' cabinets. The IT alerts Ford management and Sandvik, which in turn packs and despatches tooling for replenishment. "You can shift information around a lot quicker, and it does change the way you look, not only at supply chains, but at the rest of manufacturing itself." Goodwill's Francis Gaffney confirmed that going this route "cut our tooling by 30% in two months".

Very good. It also has to be said that a great deal can be achieved by 'doing a bit of lean' on the ERP itself. McKellen cited an example from his past where looking in detail at ERP data massively reduced lead times and scrap by matching numbers of axle cups and cones. And Ken Smith, site director at Dawn Foods, claimed stock turns up from about nine to 14, with inventory reduced from £3 million to £750,000 through similar work.

Finally, people and the big picture are just as important. Not least because failures here lead to the whole nightmare of work-arounds and departmental databases and spreadsheets operating alongside so many ERP systems. As David Lloyd quipped: "If when you put your card into the ATM machine it connected back to a spreadsheet that Kevin the gerbil maintains, how confident would you be?"

But it happens all the time. Why? Because, said Forum, users feel they need these to get the information they need - like Unite scheduling, statistical quality data, maintenance, standard operating procedures, customer complaints, raw material specs. Lean or not, this has to be addressed - and Lloyd's solution leans heavily towards web-based ERP add-ons.

top of page