Focusing Energy to Banish Muda
Firms which never start down the path because of a lack of vision obviously fail. Sadly, we’ve watched other firms set off full of vision, energy, and high hopes, but make very little progress because they went tearing off after perfection in a thousand directions and never had the resources to get very far along any path. What’s needed instead is to form a vision, select the two or three most important steps to get you there, and defer the other steps until later. It’s not that these will never be tackled, only that the general principle of doing one thing at a time and working on it continuously until completion applies to improvement activities with the same force as it applies to design, order-taking, and production activities.
What’s critically needed is the last lean technique of policy deployment. The idea is for top management to agree on a few simple goals for transitioning from mass to lean, to select a few projects to achieve these goals, to designate the people and resources for getting the projects done, and, finally, to establish numerical improvement targets to be achieved by a given point in time.
For example, a firm might adopt the goal of converting the entire organization to continuous flow with all internal order management by means of a pull system. The projects required to do this might consist of: (1) reorganizing by product families, with product teams taking on many of the jobs of the traditional functions, (2) creating a “lean function” to assemble the expertise to assist the product teams in the conversion, and (3) commencing a systematic set of improvement activities to convert batches and rework into continuous flow. The targets would set numerical improvement goals and time frames for the projects—for example: Convert to dedicated product teams within six months, conduct improvement activities on six major activities each month and at least once on every activity within the first year, reduce the total amount of inventories on hand by 25 percent in the first year, reduce the number of defects escaping to customers by 50 percent in the first year, and reduce the amount of effort required to produce a given amount of each product by 20 percent in the first year.
Most organizations trying to do this find it easiest to construct an annual policy deployment matrix, as shown in Figure 5.3 , which summarizes the goals, the projects for that year, and the targets for these projects so everyone in the entire organization can see them. In doing this, it’s essential to openly discuss the amount of resources available in relation to the targets so that everyone agrees as the process begins that it is actually doable.
F IGURE 5.3: L EAN P OLICY D EPLOYMENT M ATRIX
It’s also important to note that the process is top-down in the first step of setting goals but top-down/bottom-up in subsequent steps. For example, once the specific projects are agreed on, it’s essential to consult with the project teams about the amount of resources and time available to ensure that the projects are realistic. The teams are collectively responsible for getting the job done and must have both the authority and resources from the outset.
As the concept of making a dramatic transition begins to take hold, we often observe that everyone in an organization wants to get involved and that the number of projects tends to multiply. This is exhilarating but is actually the danger signal that too much is being taken on. The most successful firms we’ve found have learned how to “deselect” projects, 2 despite the enthusiasm of parts of the organization, in order to bring the number of projects into line with the available resources. This is the critical final step before launching the lean crusade.
Smashing Inertia to Get Started
We’ve now reviewed the basic lean principles, the five powerful ideas in the lean tool kit needed to convert firms and value streams from a meandering morass of muda to fast-flowing value, defined and then pulled by the customer. However, there’s a final and very serious paradox inherent in introducing thinking in real organizations to pursue perfection.
The techniques themselves and the philosophy are inherently egalitarian and open. Transparency in everything is a key principle. Policy deployment operates as an open process to align people and resources with improvement tasks. And massive and continuing amounts of problem solving are conducted by teams of employees who historically have not even talked to each other, much less treated each other as equals.
Yet the catalytic force moving firms and value streams out of the world of inward-looking batch-and-queue is generally applied by an outsider who breaks all the traditional rules, often in a moment of profound crisis. We call this individual the change agent.
In fact, there is no way to reconcile this paradox, no way to square the circle. The change agent is typically something of a tyrant—what one of our most thoughtful research subjects calls a “Conan the Barbarian”—hell-bent on imposing a profoundly egalitarian system in profoundly inegalitarian organizations.
Yet there are tyrants and there are tyrants. Those who succeed in creating lean systems over the long term are clearly understood by the participants in the firm and along the value stream to be promoting a set of ideas which have enormous potential for benefiting everyone. Those who fail (like many of the failed leaders of reengineering campaigns) are either identified as narrow technocrats with no concern for the very real human issues inherent in the transition, or they are dismissed by the organization as self-promoters who are simply seeking to advance their own position by riding the wave of the next “program.” Both quickly fall victim to organizational lassitude, if not to active sabotage.
Because lean systems can only flourish if everyone along the value stream believes the new system being created treats everyone fairly and goes the extra mile to deal with human dilemmas, only beneficent despots can succeed. We hope that many readers of this book will take up the mantle of the change agent. And we are equally hopeful that self-promoters and cold-blooded technocrats will look elsewhere.
For those of you with the right spirit and a willingness to invest five years in gaining the full benefits, the examples in Part II are designed to show you how to succeed.
PART II
FROM THINKING TO ACTION: THE LEAN LEAP
Even once you begin to see the importance of the five lean principles, it’s often hard to imagine how to install them in your own organization without a clear example of successful practice to follow, a template for action. This needs to be specific enough to show the real nuts and bolts, but broad enough to keep the big picture in view. What’s more, the example needs to share enough of the characteristics of your situation that extrapolation is possible with confidence about the results.
We’ve therefore provided a series of examples selected from two dimensions—size and complexity, and nationality. We will begin with three American examples which progress from a small, family-owned firm with a simple product range and only a limited past to overcome, to a massive, publicly traded organization with highly complex product and process technologies, a complex supply and distribution chain, a culturally diverse, unionized workforce, and a long history to overcome of conflictual relations with its employees, customers, and suppliers.
Then we switch our focus to the three great national industrial systems by comparing the installation of lean principles in a leading German firm and in two Japanese firms of broadly varying degrees of complexity.
Your own organization is probably different from any of these in some important ways and some customization will be required. However, the examples are sufficiently broad and the results so startling that no manager can any longer claim that lean principles cannot be applied to their situation.
CHAPTER 6
The Simple Case
Pat Lancaster of Louisville, Kentucky, is a heroic American type, the stand-alone inventor-industrialist often found at the heart of capitalist lore. He grew up tinkering in the family workshop, convinced from an early age that he could be an inventor. After college, he tried the family business of selling packaging materials to industrial firms and then life in the product development group of a large ch
emical company. “But it just wasn’t satisfying. From my earliest memories I wanted to be an independent inventor, manufacturer, and entrepreneur.” When he was twenty-nine (in 1972), he had his big idea, a new way for manufacturers to wrap their products for shipment. He and his brother invested $300 in simple metalworking tools to build their first machine, rented a small warehouse, and went to work under the corporate name of Lantech, a contraction of Lancaster Technologies.
Lancaster’s big idea was for a device to “stretch-wrap” pallets of goods (for example, the cases of cola we examined in Chapter 2 ) with plastic film so they could be shipped easily from plant to plant within a manufacturing system and then onward, as finished products, to the wholesaler and retailer. Traditional “shrink-wrapping” was then in wide use by manufacturers and distributors who laid plastic bags loosely around large pallet loads of goods that were then run through an oven to shrink the plastic and give a tight fit.
Stretch-wrapping, by contrast, pulled the plastic wrap tightly around the pallet load as it rotated on a turntable. As the plastic was stretched taut, it rebounded slightly to give a snug fit while eliminating the energy, equipment, effort, and time required for heat treating. In addition, stretching the wrap practically halved the amount of plastic required to secure a pallet load for shipment.
Lancaster’s next invention was the key complement to his fundamental insight that the plastic should be stretched rather than shrunk. He discovered that a complex set of precision rollers (collectively termed the roll carriage) could exert a smooth force on the plastic to stretch it dramatically before it was wound around the pallet. Eventually, he found ways to decrease the amount of plastic needed to hold a pallet load together by a factor of 7.5 compared with shrink-wrapping.
When Lancaster obtained patents for his concepts at the beginning of the 1970s, they were so general and broad that he could easily fend off competitors for years. All he needed was a market. This was supplied by the world energy crisis of 1973, which unfolded just as he completed his first, hand-made stretch-wrapping machine. As energy prices zoomed, the amount of process energy and plastic (made from natural gas) which his new technique could save created an overwhelming advantage for stretch-wrapping in the contest with traditional shrink-wrapping.
Suddenly he had a real business and needed to think about how to make his product in volume. He had created his initial design and his first machine in a continuous flow of activities, so Lantech, like most start-up businesses, was born lean. However, it didn’t seem plausible to run an established business this way.
In reconstructing his thought process during the transition from start-up to established firm, Lancaster recollects that “I had no production experience—remember, I was an inventor—so I decided I should get myself an experienced operations manager. What’s more, I knew I would need to engineer a variety of configurations of my basic concept for different wrapping tasks, so I got an engineering manager. Finally, I had a complex product which needed explanation to the customer, so I got a sales manager. I knew instinctively about the division of labor and returns to scale, so it seemed natural that my operations, sales, and engineering managers should organize my rapidly growing firm into a series of departments, each with a specialized task, and each operating in batch mode.”
The operations manager created a series of departments in the manufacturing plant, one for each of the basic steps in building a Lantech stretch-wrapper. The Sawing Department used metal saws to fashion frame members from steel beams. The Machining Department drilled and punched holes in the steel to create attachment points for component systems. The Welding Department welded the frame members together to form the completed frame for the machine. The Painting Department applied a corrosion-inhibiting base coat and a cosmetic finish coat to the completed frame. Component systems—notably the roll carriage, the turn-table, and the control module—were assembled in the Sub-Assembly Department from parts purchased from suppliers. These were attached to the frame in the Final Assembly Department.
Final Assembly was not the end of the line for products making their way from department to department and storage area to storage area. Because it was thought to be efficient, Lantech built its four basic types of machines in batches. Ten or fifteen machines of a type would be fabricated and assembled at a go. The nature of the product, however, meant that individual customers usually bought only one. Therefore, it was necessary to store many machines in a finished goods area for some time before they could be matched up with customers.
When it was time for shipment, it was often necessary to remove grime and to paint over nicks caused by moving machines from department to department. This meant a journey to a Touch-Up Department. Often the machine had to be sent back to Final Assembly as well to change its mix of optional features in order to accommodate changing customer desires. Finally, the machine was sent to the Crating Department for actual shipment.
The progression of a stretch-wrapper through Lantech is shown in Figure 6.1 , often called a “Spaghetti Chart” by firms who have mastered lean thinking.
F IGURE 6.1: P HYSICAL P RODUCTION AT L ANTECH
Physical production of the machine was not the only process to manage. The real complexity in volume production began to emerge as Lantech tried to move the orders gathered by the sales force (a group of about fifty independent firms distributing industrial machinery) through the office and into the plant.
Because the machines were often customized and cost from $10,000 to $50,000 apiece, it was decided that a standard price list would not work. Instead the sales force contacted Lantech for authorization before quoting a price on any machine with special features. The proposal was sent to the Engineering Applications Department within Sales for cost analysis. After analysis, the “right number” was sent back to the sales force. Then, once the offer was accepted (with the distributor negotiating a final price with the customer which included the distributor’s margin), the order was sent back to Lantech for production scheduling.
Upon arriving back at Lantech, the order proceeded from the Order Entry Department to the Credit Checking Department to the Engineering Applications Department (for its second visit). There, a Bill of Materials (BOM) was generated for the order. This was the precise list of every part which would be needed to manufacture a specific machine. Because every department had a waiting list of orders, there were usually delays. Typically, an order took twelve to fourteen working days to travel from the Entry Department to the Scheduling Department, while the actual processing time—what we will call “continuous flow time”—was less than two days.
The order with the BOM was then taken to the Scheduling Department inside Production Operations to work into the master schedule. Because it became apparent immediately that the flow of production through the plant would be very erratic, a separate Order Management Department was created in Sales to maintain liaison between the independent sales force and the plant on just where the machine was in the production process and to initiate expediting (using a technique we will examine in a moment) if the customer was getting restless. Information progressed through the system as shown in Figure 6.2 .
F IGURE 6.2: L ANTECH O RDER F LOW
The master schedule itself resided in the Scheduling Department inside Production Operations in the form of a computerized Material Requirements Planning system. The MRP melded a long-term forecast for orders with actual orders as they were received to create a daily production schedule assigning tasks to each department in the plant. Each morning, workers in each department—sawing, machining, welding, paint, sub-assemblies, final assembly, touch-up, and crating—would pick up a printout with their tasks for the day. At the end of each day, each department would report its progress back to the computerized Scheduling Department.
This system was fine in plan, but always a mess in practice because of the conflict between changing customer desires and the logic driving the production system. In order to gain scale economies, Pat Lancast
er and his operations manager decided from the beginning that each department should do its work in batches: ten frames welded for the E model, then twenty frames welded for the T, then twenty-five welded for the V. This minimized the time Lantech’s machinery was idle during the changeover to a new part. In addition, running long batches was thought to improve quality by minimizing opportunities to misset machines and by keeping operators focused on the operation itself rather than changeovers.
Separate departments for each production step, batches of parts run through the departments, and waiting time at the entrance to each department inherently meant long lead times. Typically, it took sixteen weeks to turn the incoming steel for the frame into a completed machine on the shipping dock. Most of this time was spent waiting as batches of parts were built in each department and then sent to storage to await the next fabrication step in the next department. The actual amount of time needed to complete the physical transformation of raw materials into a stretch-wrapper—the “continuous flow time”—was only three days.
Long lead times meant in turn that the sales force distributing Lantech’s machines to the end user tried to figure out how to beat the system. A favorite approach was to order machines on speculation and then, as a real customer was found, to alter the options requested (or even the base model) very late in the production process. This tactic created the need either to rework the machine initially ordered or to slip the delivery date and build a properly configured machine from scratch.
Soon the factory was being pulled in opposite directions by two conflicting planning systems—the master schedule worked out by the Scheduling Department based mostly on sales forecasts, and the ever-changing demands from the Sales Group intent on pleasing actual customers.
Lean Thinking Page 12