What’s Next After Lean Manufacturing?

Genetic algorithms, virtual-engineered-composite cells and the Internet help increase speed and product variety.

Reading Time: 5 min 
Buy
Like what you’re reading?
Join our community
Member
Free

5 free articles per month, $6.95/article thereafter, free newsletter.

Subscribe
$89 $45/Year

Unlimited digital content, quarterly magazine, free newsletter, entire archive.

Sign me up

Technological complexity and burgeoning product variety are placing more demands on manufacturers than they can handle, even organizations that claim to have adopted lean-manufacturing principles. That is why companies such as Auburn Consolidated Industries, Deere & Co., Maytag Corp., Fleetwood Enterprises and VEC (Virtual Engineered Composites) Technology, formerly Pyramid Composites, are investigating new technologies and Internet-based processes that offer potential solutions. These pioneers are hoping that techniques and technologies, such as genetic-algorithm software, VEC cells and manufacturing control via the Internet, will enable them to cut costs, decrease cycle times and deliver personalized products with more features faster than ever before. We may even be seeing a new stage in the evolution of manufacturing.

Distance and distribution complexity have long created problems for manufacturers, but today companies have new tools with which to tackle the challenges. The problem of shipping parts to where they are needed has always gotten in the way of faster cycle times. Distance creates particular challenges for companies making products from materials that are molded to make Jacuzzis, Jet Skis, automotive parts and the like — materials such as fiberglass or plastics. Because molds are expensive and consistent quality is difficult to achieve, manufacturers requiring molded parts have been forced to rely on a handful of suppliers, adding shipping time and coordination complexities to their burdens. So imagine how life would improve for some manufacturers if high-quality molded parts could be made in a portable minifactory placed in their own backyard.

Inventor Gene Kirila believes his system for producing molded plastic products may usher in a new era. Formerly the CEO of Pyramid Composites and now head of GK Ventures, Kirila is co-inventor with Robert McCollum of VEC cells. VECs are transportable factories for manufacturing molded parts. They run a patented manufacturing process whose operating system is best described as a controlled process-in-a-box. VEC cells provide operators with computer controls and simple visual and audio instructions that guide operators through the molding process. Because the software can be upgraded and controlled centrally over the Internet, VEC cells can deliver a consistent process regardless of where they are located or who is operating them.

Richard E. Morley, the inventor of the programmable logic controller, the device that launched the $5 billion industrial automation industry, believes that VEC cells represent a significant breakthrough. “The VEC takes manufacturing to the consumer; Bluetooth software [high-speed, low-power microwave wireless machine-to-machine communication] and minifactories will allow reconfiguring of a plant in seconds.” The first VEC cell, sold to boat-builder Genmar, is capable of making 17.5-foot molded boat hulls in 70 minutes, a process that takes several days using traditional methods.

Another obstacle to faster cycle times is the complexity of distribution networks. In the appliance business, for example, the intricate network of distributors, warehouses, dealers and producers makes it hard to get a product ordered and delivered quickly. Most appliance producers and their customers maintain inventories of finished goods, raw materials and components worth millions of dollars rather than rely on risky customer-demand forecasts. Now managers at Maytag think they have found a solution to the problem.

In summer 2000, Maytag set up an Internet-based communications link between the Cleveland, Tennessee, line that assembles small cooktops and Fleetwood Enterprises, a Riverside, California-based manufacturer of recreational vehicles — a major Maytag account. Previously, when Fleetwood vehicles on the assembly line were ready for a cooktop, a finished unit was pulled from a warehouse, a distribution center or a truck. Eventually, Maytag planners would resupply the warehouse, but the inventory network was expensive and slow. Today, a unit installed on Fleetwood's California assembly line automatically triggers electronic just-in-time replenishment of that cooktop by the Cleveland site. Planners have been able to phase out millions of dollars of inventory. And they see no limit to the cost savings from such pipeline shortening, which has let Fleetwood save, for example, on pre-assembly warehousing, too.

Manufacturers also are using innovations such as genetic algorithms to produce and deliver an ever greater variety of products more efficiently. A genetic-algorithm software package uses a type of evolutionary computation that finds the best solution to a problem. First, possible solutions are developed; next, the solutions are compared to find the best one; finally, the best solution replaces the poorer original solutions, and the process is repeated. In that way, the software can calculate the most efficient solution to complex coordination problems. For example, a genetic-algorithm package can be used to plot a truck's best route through five cities, with parameters such as time requirements, sequencing of package deliveries and minimization of gas expense. A less than optimal solution —one that zigzags the delivery route, for example — would be rejected because it would not fall within “best” cost and time parameters. The mechanics of natural selection are at work: Only the best solutions survive.

Agricultural-equipment manufacturer Deere, facing numerous demands for variations in its planters, used genetic-algorithm software to juggle the planter assembly schedule until the software arrived at the best sequence of options across the assembly line. The first Deere genetic-algorithm project was big and expensive, but the software package was quick and allowed the company to meet changing consumer demand speedily and provide valued market leverage. Planners could handle superhuman degrees of product variety without excess inventory or frequent, expensive line changeovers.

Now Deere and its supplier, Auburn Consolidated Industries, are using genetic-algorithm routines to build the best schedule to handle the huge variety in job-shop applications. At Auburn's Nebraska plant, Deere's laptop-toting engineers use a package called Evolver that tests and selects scheduling possibilities until the software has created a solution that meets all optimization parameters for timely deliveries, sequencing and routings.

Together, genetic-algorithm software, Internet links between customers' and suppliers' production lines, and VEC technology form a springboard for the next stage in manufacturing. However, as manufacturers have discovered, none of those sophisticated technologies is worth implementing if the production process is not lean. Once companies combine a lean process with the right technology tools, they can move manufacturing forward in giant leaps and bounds.

Reprint #:

42294

More Like This

Add a comment

You must to post a comment.

First time here? Sign up for a free account: Comment on articles and get access to many more articles.