Many of the forward leaps that have shaped today's world-class production systems have been spurred by new organizational structures--and new attitudes about the role of the workforce.
Increasingly, by fashioning leaner, flatter, and less-hierarchical organizations, companies have pushed decision-making responsibility down to front-line workers, giving them greater ownership of the production process. Today, empowered and self-managed teams not only seek ways to improve quality and productivity, but also they've assumed numerous tasks once reserved for supervisory management.
"It is hard to generalize across all industries, but a combination of organizational changes have had the effect of breaking down a lot of the old rigidities and barriers that existed in traditional manufacturing organizations," observes MIT's Lester. In The Productive Edge, he recalls that a 1989 report by the MIT Commission on Industrial Productivity pointed to a need for American industry to "accelerate its historic shift away from the system of mass production of low-cost, standard products." Replacing traditional mass-production methods, the report asserted, would be new systems capable of producing smaller volumes of products "tailored to different segments of an increasingly demanding, sophisticated, and global market."
Among the characteristics of the new mass-customization systems are "more flexible and less bureaucratic organizations," Lester points out, along with cross-functional teams, closer links to suppliers and customers, and "intelligent use" of information technology. "Most important," he notes in his book, "success in this new system of production [hinges] on the knowledge that people at each level of the organization [are] acquiring and bringing to bear in the course of their work."
Although he stops short of declaring that the age of mass customization has arrived, Lester does suggest that elements of it parallel what he sees as a trend toward "service-enhanced" manufacturing. "While 'manufacturing' has traditionally meant the production of tangible objects," he says, "for today's customers the value of manufactured products increasingly hinges on intangible attributes -- speedy delivery, convenience of use, style, brand identity, reliability, that -- were they not embodied in the product -- would be thought of as services."
In at least one respect, he observes, the predictions of a transition away from mass-production seem on target: "In many industries, information technologies are promoting a shift away from 'selling what you can make' toward the demand-driven principle of 'making what you can sell.'" Indeed, many leading manufacturing firms have been converting their production planning and scheduling from forecast-driven "push" systems to demand-based "pull" systems where products aren't made until a customer order is received -- or where the factory gears production to replenishment of finished goods based on real-time feedback about actual sales.
Christopher Gopal, Dallas-based national director of supply-chain and operations consulting for Ernst & Young LLP, contends that mass customization "leads you to a build-to-order system, rather than a complete pull system throughout the supply chain." To avoid prohibitively long leadtimes, components and semifinished products will have to be stocked at various points in the production channel, he says. "You can't pull product through the entire supply chain if your customer expects response within 24 hours. That would be an unrealistic expectation."
Mass customization, he believes, will become a major trend -- but it will require a restructuring of operations as well as changes in product design. Companies pursuing such a strategy will need to think in terms of "design for mass customization," with an emphasis on commonality of parts and postponement of final configuration. "With postponement, at least the final operations -- final assembly, testing, and packaging -- are done very near to the customer," Gopal points out. In fact, it is more a case of final-assembly-to-order than build-to-order, he notes. "As far as possible, the manufacturing up to that point -- as well as the design -- should be as vanilla as possible, with as many common parts as possible."
Some manufacturers have adopted mass-customization methods to serve at least niches in their markets. At Taylor Made Golf's new plant, for example, two of the 14 assembly cells are dedicated to making custom-fit clubs, which may have grips, shaft configurations, or loft angles that are different from its standard lines of clubs. "We'll do a few additional operations to make the club unique to the individual," notes Bill Dettenborn, general services manager. "We'll build it to the specifications we get from the pro shop or from a golf instructor who places the order. We'll tweak the club to make it fit a particular golfer's body or swing."
In addition, the Carlsbad facility has a "PGA Department" -- a special work center that creates clubs for touring professionals who may have special requests. "They will come in and tell Willie King, who runs that shop, just what they want done to make a club feel a little better," Dettenborn notes. "They may want a more flexible shaft; or if they don't like the way the club looks, they might want some grinding done. These guys are real fussy. But it's their livelihood. They want a club to feel the way they want it to feel."
The Time Factor
Since the late 1970s, when U.S. and European manufacturers began to appreciate how just-in-time (JIT) manufacturing practices had given Japanese companies a competitive advantage, a common focus for firms striving for world-class performance has been on time compression. Clearly, that emphasis -- in manufacturing operations, front-office order processing, and customer leadtime -- has been accelerating. Some firms, in fact, now use time compression as the "glue" that aligns all major improvement efforts.
In discrete-manufacturing industries, the oft-stated ideal is continuous-flow systems that minimize work-in-process (WIP) inventories, support small-lot production, and enable quick response to shifts in marketplace demand. And some firms have leveraged advanced information technology to enable high-speed production in a high-mix, small-quantity environment.
EFTC Corp., a Greeley, Colo., contract electronics manufacturer, developed an innovative methodology that it terms "asynchronous process manufacturing" (APM) -- which involves a combination of high-speed manufacturing equipment, sophisticated information systems, and standardized process teams "to produce small quantities of products more flexibly and more quickly through the factory." Prior to the adoption of APM, the company notes, high-mix manufacturing was "incompatible" with high-velocity techniques.
A finalist in IW's 1998 search for "America's Best Plants," EFTC describes APM as a "hybrid" of continuous-flow and batch production techniques. The implementation of APM has decreased setup and cycle times and increased productivity. Proprietary software -- called Assembly Execution System -- was a key enabler and helped to drive manufacturing cycle time from 21 days to just five days.
Many other Best Plants winners and finalists have reported manufacturing cycle-time reductions of 50% or more -- often permitting comparable reductions in order-to-shipment leadtimes. However, the IW Census of Manufacturers found that only 7.8% of nearly 2,600 plants providing data had collapsed cycle times by as much as 50%.
Using process-mapping techniques to identify activities that don't add value, plants that are moving aggressively to lean production typically seek to eliminate unnecessary steps or motion, thus shortening the cycle and improving productivity in the process. And more often than not, they've restructured factory operations from large-batch production to continuous-flow systems relying on manufacturing cells.
In product-focused cells, multiskilled teams have control over all of the equipment and support resources needed to produce a product or family of products, and travel distances, often through a U-shaped arrangement. In contrast, traditional batch production systems grouped machinery and equipment by type. As a result, work in process had to travel long distances from one department to another, and large amounts of WIP tended to build up in between operations.
In cellular/continuous-flow systems, lot-size reductions -- supported by quick-changeover methods to reduce setup times -- have enabled dramatic leadtime reductions. The IW Census of Manufacturers detected a strong correlation between quick-changeover methods -- such as SMED (single-minute exchange of dies) -- and shortened cycle times. Whereas the median cycle time for firms that have widely adopted quick changeover was 16 hours, across all industry sectors, the median for nonadopters was 48 hours.
The basic model that many companies have emulated in their time-reduction strategies is the Toyota Production System (TPS), which evolved into the prototype JIT/lean-manufacturing system. Auto-industry analysts began to realize that Japanese firms that had adopted lean production systems "required far fewer resources than their American counterparts -- fewer people, less space, smaller inventories, less capital investment, and so on," notes Lester in The Productive Edge. "But the lean production system in fact offered much more than just a reduction in waste. It also delivered higher-quality vehicles, an organizational capacity for continuous improvement, and fast, flexible response to changes in the market. . . .
"By relentlessly driving inventory down, Toyota was deliberately imposing increasingly high levels of stress on its system. But where others saw only the prospect of chaos, Toyota perceived a crucial benefit. Removing the safety net provided by inventory would quickly reveal the weak points in the process. . . . And once identified, these problems could be quickly solved. Indeed, they would have to be -- otherwise the entire factory would grind to a halt."
Meanwhile, in plants operated by Detroit's Big Three automakers, "the existence of all those buffer stocks practically guaranteed a lack of knowledge of what was really happening around the system at any given time." However, in the last 15 years, U.S. carmakers -- and other manufacturing industries -- have adopted lean-manufacturing practices, both within the walls of specific plants and throughout the extended supply chain. But many companies, including prominent multibillion-dollar firms, are still struggling with the implementation process.
"Lean manufacturing is not a one-time event," notes an executive at Lockheed Martin Corp.'s 11,000-employee Ft. Worth plant that builds military fighter aircraft. "It is a systematic and continual refinement of processes over an extended period of time." Moreover, extending lean-production concepts to reduce inventory throughout the supply chain calls for long-term agreements with key suppliers and "looking at ways to reduce waste in a supplier's process."
Just as lean production increases a plant's flexibility, leanness throughout the supply chain is a key to efficiently reacting to changes in demand. "We have to be very flexible in responding to upside demand and flexible in controlling [products] during the product life cycle," asserts Ralph Russo, executive vice president of operations at Bay Networks Inc., a Santa Clara, Calif.-based maker of internetworking products. "Our competitors will . . . change the game rather rapidly, and the last thing we want is to have millions of dollars worth of units build up in the market in anticipation of sales."
QRM vs JIT
Despite the current emphasis on JIT/continuous-flow systems, the classic "lean" manufacturing model isn't entirely suited for all manufacturing scenarios, contends Rajan Suri, a professor of industrial engineering at the University of Wisconsin-Madison and director of the Center for Quick Response Manufacturing, a 40-company consortium. In Quick Response Manufacturing: A Companywide Approach to Reducing Leadtimes (1998, Productivity Press), Suri advocates an approach that builds on JIT/flow concepts -- but with variations. His quick-response-manufacturing (QRM) methodology, for example, makes time reduction the sole driving force, and it isn't confined to high-volume linear-flow production systems that rely on kanban "pull" signals to control the movement of materials.
"There are several limitations to JIT," Suri stresses in his book. "First, most firms that adopt JIT or flow techniques need somewhat repetitive manufacturing and somewhat stable demand. In contrast, companies having a wide variety of products with demand that varies considerably can apply QRM methods -- even one-of-a-kind manufacturers. . . . QRM allows you to focus more and more on individual, customized production, while still maintaining low inventory and fast response."
Whereas JIT/lean-production emphasizes the elimination of non-value-added "waste," QRM prescribes a singular focus on reduction of leadtime in all aspects of a manufacturer's operation -- "from receipt of order to delivery to the customer," Suri says. Through "relentless pursuit" of leadtime reduction -- which has implications for organizational structure, purchasing policies, capacity planning, and other activities -- practitioners not only get time compression but also eliminate waste, improve quality, and reduce costs, he contends. A study of companies that had reduced leadtimes, he points out, found a 2:1 ratio between reductions in leadtime and cost. "In other words, a 50% reduction in leadtime resulted, on average, in a 25% reduction in overall product cost."
Implementing QRM often requires rethinking policies related to lot sizes and capacity utilization, Suri says. Striving for 100% equipment utilization, for example, can create "dysfunctional interactions" that cause work queues to grow and leadtimes to stretch out. He recommends that manufacturers plan to operate at 70% to 80% of capacity on critical resources, contending that "idle capacity actually serves as a strategic investment that will pay for itself many times over in increased sales, higher quality, and lower costs."
Like Ernst & Young's Gopal, Suri argues that pure pull systems are impractical for high-mix production -- or where products must be custom-engineered. "Pull starts with the sale of a product already in stock and then works its way back upstream through replenishment of inventories. This scenario simply cannot exist in a custom-engineered environment," he stresses. "For the QRM company that makes tens of thousands of different items, this implies that there is inventory of each of these items, or it's partly manufactured stock, at each stage of the supply chain," he observes. And that, of course, would not be feasible.
Product-oriented manufacturing cells are an important element of QRM, but -- unlike JIT cells -- they often need to combine push and pull techniques, Suri says. In a high-mix or custom manufacturing environment, "Orders for different customers need not have the same routing through the cell," he emphasizes. "QRM cells are designed to be more flexible and do not need linear flow." Flexibility is achieved by allowing for variation in the sequence of operations performed within the cell -- which rules out typical pull signals.
