THE END OF THE LINE FOR MASS PRODUCTION
Dr. David M. Anderson, P.E., CMC
1923 Ė THE HEYDAY OF MASS PRODUCTION
Mass production thrived in the bygone era of stable demand and little product variety. At its peak, the "any-color-as-long-as-itís-black" Model T Ford had a 57% market share. Despite its very low price of $245, however, it was pulled off its pedestal when General Motors offered variety such as color paint and other options. Ford was slow to respond, mostly because mass productionís keys to its success, hard tooling, labor specialization, and economies of scale, prevented it from offering variety or adapting quickly to emerging trends.
IF YOU BUILD IT, THEY WILL COME
In the mass production paradigm, the marketing department forecasted demand and these forecasts drove complex MRP systems to order the required parts and materials far enough ahead of time so that they would, hopefully, all arrive at the factory before the production run was scheduled. Assuming this was successful, equipment would be set up and a batch of products was built to satisfy the forecast. In order to make a batch of products, batches of parts would wait in queues before each operation Ė hence mass production is more aptly described by the less glamorous phrase, batches and queues.
If all the parts did not arrive on time or if market demand had changed since the original forecast, then more parts would have to be procured by expensive expediting. Finished products were then put in a warehouse or series of distribution centers where they awaited the anticipated orders, relying on the industrial version of the "Field of Dreams" philosophy: If you build it, they will come.
Not only has it become harder to satisfy customer demand from inventory, it is also becoming more expensive. For the last 25 years, the average inventory carrying cost has been 25% of inventory value per year. In other words, $4 million in inventory will cost the company $1 million per year to pay for the interest, space, insurance, and administration. From a competitive standpoint, it doesnít matter if the inventory is at the factory, distributors, or stores -- customers still have to pay for it. When inventory goes obsolete, manufacturers face a deadly dilemma: write it off or delay new products until the old products can be sold off, usually at huge discounts. Packard Bell suffered severe consequences when it got caught with a warehouse full of 75 MHZ computers when consumers wanted higher speed computers.
DEFECTS BY THE BATCH
The flow manufacturing movement is pointing out the inherent quality problems of batch production. In one-piece flow, as each part is completed, it is passed on to the next worker, who looks for any visible deviations. If the part doesnít fit or work in the next operation, the feedback will be immediate leading to quick rectification of the problem at the source. This is part of the continuous improvement, or kaizen, movement in which the production system is continuously learning and improving itself.
In mass production, batches of parts go from one machine or work station to another until they reach test, which may discover that a recurring error at any one of the steps ruined the whole batch. Then the entire batch would have to be reworked or scrapped, thus raising costs, delaying delivery, and reducing production capacity.
In addition to recurring defects, large WIP (work in process) inventory levels cover up many other quality and efficiency problems when buffer inventory abounds, "just in case."
ITS ONLY OVERHEAD
Based on its early success with long runs of identical products, mass production has taken on the image of being the lowest cost way to build any product. Industrial empires were built on the principle of economies of scale and Frederick Taylorís specialization of labor. And even today, many people think that the only way to get cost down is to get volume up.
However, as product variety goes up and batch size goes down, opportunities for economies of scale fade, except for standardized parts, which, ironically, are underutilized by most mass producers.
Today, labor and raw materials are only a small portion of the only cost that matters -- the customerís cost which is the manufacturerís price. But, since they are the only costs measured in most companies, they become the exclusive attention of "cost reduction" efforts, which often produce counterproductive results by buying cheap parts, awarding the lowest bidder, or moving production to countries with low labor costs.
The majority of total cost is "overhead," which is rarely quantified, instead being allocated (averaged) over all products. And when overhead is not quantified, no one understands it or tries very hard to do anything about it. One of the authorís clients had a controller who told engineers to "not worry about overhead -- itís a fixed cost." In reality, these so-called fixed costs, like floor space, are not fixed at all. Flow manufacturing can cut floor space in half, eliminate WIP inventories, and eliminate recurring defects. Quality can be designed into the product and built in with process controls. The costs of R&D and change orders can be minimized by good product development practices. The "cost of variety" (see below) can be minimized by mass customization. Setup costs, finished goods inventory, and most of the distribution costs and material overhead can be eliminated by spontaneous build-to-order.
TAKE ALL ORDERS
Companies that have revenue growth as their cornerstone encourage the mantra, "take all orders," even though this increases variety and makes their mass production operations less efficient. This increase in variety shrinks batch size to the point where setup becomes a major cost and eats away at plant capacity.
When it comes to customization, most companies take an ad hoc, case-by-case approach to product customization where the mantra expands to "take all customized orders," followed by heroic efforts in Engineering, Purchasing, and Manufacturing where employees jump through hoops to get every customized product out the door. Often, such "fire-drill" customizations dilute other important programs when they borrow people from new product development and factory improvement programs.
These custom products and the unusual standard products produced at low-volume are usually subsidized by the high-volume "cash-cows." This has two detrimental effects: (1) the customized products and unusual standard products are not making nearly as much real profit as assumed and many are really losing money, and (2) the cash-cows must be priced higher than they should be to pay for the subsidization, thus lowering their competitiveness.
The quick and efficient approach to product customization is mass customization, which is accomplished by proactively developing families of products around modular product architecture, implementing flow manufacture to achieve batch-size-on-one capability, establishing a spontaneous supply chain around standard materials, creating agile order entry systems based on configurators, and building parametric CAD templates with automatic CAD/CAM links to CNC equipment. For more on this see Mass Customization.
THE EFFECT OF VARIETY ON MASS PRODUCTION
Today, most companies offer hundreds or thousands of products or product variations, sometimes called SKUs (Stock Keeping Units). Even if most of the SKU variety is caused by "minor" labeling and packaging variations for sale in various regions of the globe, it still slows down the mass production and warehousing operations and makes them less efficient. And today, many companies are forced to customize products, to some degree, for increasingly selective customers or to compete in niche markets.
Unfortunately, most of these companies use a reactive approach to product variety, where streams of new products are introduced on a case-by-case basis without the benefit of systematic product portfolio planning or the standardization of materials, parts, modules, and processes. And, in the vast majority of companies, the old products are not discontinued -- they just accumulate and dilute company resources away from higher leverage new products.
Relying on forecasts to order parts is becoming more problematic as product lines grow, build quantities shrink, and markets become more turbulent and unpredictable. As selling from inventory becomes more costly and less effective, some companies try to become agile and build products "to order," but do this reactively. Their inflexible factories still need to order an enormous variety of parts and schedule their production in small batches on production equipment that is efficient only in large batches -- the legacy from the mass production era.
THE COST OF VARIETY
Chapter 3 of the authorís book, Agile Product Development for Mass Customization (McGraw-Hill, 1997) introduces the concept of the cost of variety, which is the sum of all the costs of attempting to offer customers variety with inflexible products that are produced in inflexible factories and sold through inflexible channels. This includes the actual costs of customizing or configuring products, all the setup costs, the costs of excessive parts, procedures, and processes, and the excessive operations costs caused by trying to offer the customer variety from inflexible factories. Another way of estimating the cost of variety is to compare a companyís current operation budget to the idealistic case of producing a single product with no variety manufactured in the same volumes as current operations. The difference between current operations costs and the single product scenario is the cost of variety.
As the graph shows, the variety cost for mass
production increases exponentially with increasing market variety
because of the compounding effect of all the
inefficiencies discussed above.
TRYING TO MAKE MASS PRODUCTION FLEXIBLE
Many companies blindly accept the shortcomings of mass production and then spend millions of dollars and precious resources trying to make an inherently inflexible process become flexible and responsive. One category of this is building "distribution centers," in essence, warehouses for finished goods inventory. The Gap is building a $150 million, 1,100 employee distribution warehouse for Old Navy Clothing. Similar steps are being taken by Williams-Sonoma for housewares and Dollar General, a Nashville-based retailer.
Some digital strategies propose a decentralized "web-based" version of selling from inventory and recommend offering "the illusion of build-to-order" by searching the web for inventory! Similarly, on the transportation side, overnight shippers come to the same conclusions when they recommend that the way to reduce inventory levels is to know where the inventory is and how long it will take to ship it.
Many companies, encouraged by ambitious suppliers, think that they can solve any problem by "throwing software at it" and nowhere has this been more prevalent than with all the attempts to improve mass production with software. A few decades ago, MRP systems were created to allow purchasing functions to cope with rising product variety, an unnecessary proliferation of parts, and an ever expanding vendor base caused by procurement practices driven by low-bidding. But everyone forgot the basic premise of Industrial Engineering 101: simplify before automating or computerizing. There are many effective ways to simplify the supply chain with standardization, automatic resupply techniques like kanban, and product line rationalization which can eliminate the unusual parts by eliminating or outsourcing the unusual products.
One of the biggest shortcomings of mass production is the reliance on forecasting as a basis to order parts and schedule batch production. So some companies are focusing their efforts on programs to improve forecasting. VF Corporation, makers of Lee, Wangler, Britannia, and Rustler jeans, is spending $100 million in information systems that include sophisticated software to aggregate information on demographics and point-of-sales trends to try to improve forecasting. However, a much better approach would be to adopt a production system that could build products spontaneously without forecasts (see new paradigm section below).
Some companies accept the built-in inefficiencies of mass production and then conclude that the only way to deal with rising labor costs and falling prices is to move manufacturing off-shore where labor costs are cheaper.
TIME FOR NEW PARADIGMS
Now it is time to shift the paradigm. The overall paradigm shift could be described as evolving from mass production to mass customization or changing from build-to-forecast to build-to-order. Part of this shift is the change from batches and queues to flow manufacturing, which is the subject of this journal.
Mass Customization is the proactive management of variety that can build customized products for individual customers or niche markets like specific countries or regions. Mass customization also enables a steady stream of "new" products that are really planned "variations on a theme."
Spontaneous Build-to-Order is the ultimate in supply chain management that allows companies to build standard or mass-customized products on-demand without forecasts, inventory, or purchasing delays.
Dr. Anderson is a California-based consultant specializing in training and consulting on build-to-order, mass customization, lean/flow production, design for manufacturability, and cost reduction. He is the author of "Build-to-Order & Mass Customization, The Ultimate Supply Chain Management and Lean Manufacturing Strategy for Low-Cost On-Demand Production without Forecasts or Inventory" (2008, 512 pages; CIM Press, 1-805-924-0200, www.build-to-order-consulting.com/books.htm) and "Design for Manufacturability & Concurrent Engineering; How to Design for Low Cost, Design in High Quality, Design for Lean Manufacture, and Design Quickly for Fast Production" (2008, 448 pages; CIM Press, 1-805-924-0200; www.design4manufacturability.com/books.htm). He can be reached at (805) 924-0100 or email@example.com; web-site: www.build-to-order-consulting.com.