Value Stream Mapping

Value stream mapping (VSM) is a concept popularized by the book Learning to See (Rother/Shook). VSM (and lean manufacturing) essentially says that customers are only willing pay for the "value-add" parts of any process. Everything else is "Muda," the Japanese word for waste. In VSM, the theory is that 85% of any process is Muda. Eliminate that, and you've achieved your improvement goal. The trick, therefore, is to first learn how to identify Muda and then eliminate it.

"The general feeling among manufacturing managers is that this concept of 85% waste is heresy. We've been so conditioned to simply look microscopically and inwardly at our own processes, searching for 5 or 10% improvements here and there. When someone comes in and says we're going to redesign for 85% improvement, they are incredulous," explains Jones. "Many times there is passive resistance. They don't 'see.' You first have to learn to see, and then once you 'see,' it will drive you nuts; you'll see massive improvement opportunities throughout everything in your life. In our business, we're now like the Pac-Man game, eating up value chain dots, because our customers need us to take on larger sections and without increasing durations. Internally, we call this high-velocity systems and we run around preaching this to our own suppliers and customers. If they want to play with us in the 85% improvement game, they have to learn to see what we see, and then they have to build their business directly in to ours. It is the only way."

The overall philosophy in the new world is that efficiency can only be significantly increased if we reach beyond our own finite element control and optimize processes and steps across major businesses and business systems, and real-time communications and data and information exchange is the key. This is consistent with the Inescapable Data mantra that businesses will be driven by efficiencies goals more than immediate bottom-line profit and that the new efficiency realizations will only occur if the breadth and depth of business operations spans organizations. This means opening up key internal aspects of your own operation and allowing your customers and suppliers to tie into your systems and processes to an unheard of depth. "Most North American companies would turn inventory 10 times a year. We now have customers that turn inventory 1,800 times a year, 5 times a day. Do you think a month-end inventory report to them now has any value? You can't get from 10 to 1,800 by squeezing. You have to start over and be willing to tear down walls of control and knowledge. Our customers want to know exactly the current inventory levels of any single part we're producing as well as its dependencies. This is unprecedented yet required in order to reduce that 85% of Muda," extols Jones.

Automotive Resequencing, Wirelessly

Automotive manufacturing is notoriously a floor-space-intensive operation. Building a Hummer aggravates assembly-line floor-space issues given the heft of the major components, such as those bulky seats, which make for local inventorying and staging nearly impossible. AM General manufactures the Hummer for General Motors out of its Mishawaka, Indiana, plant and the new H2 has extremely aggressive sales (and thus production) goals. Without being able to physically expand their plant, AM General had to redesign their operations to meet a much increased production load and lean manufacturing analysis led to innovative deployment of wireless networking technology. AM General deployed products from WhereNet and GE Fanuc Automation (CIMPLICITY) to build a novel real-time parts replenishment system.^[1]

Because of competitive pressures and plant closings, automotive manufacturers are forced to new levels of flexibility such as moving several different models of a vehicle down a single assembly line as well as line-time customizations of vehicles. Naturally, this creates significant challenges with coordinating parts at various line-staging cells. Wireless on-the-line devices and RFID allow a line to be rapidly reconfigured without restringing wires and to be real-time tied into the supply network. In the H2 case, AM General uses special wireless "WhereCall" buttons affixed to each operator work location (and easily moved around). Instead of having a cache of Hummer seats queued up at each cell (of various colors or other details), the operator is able to signal a forklift driver to retrieve the next-needed seat by pressing the appropriate WhereCall button.

Similarly, each H2 has affixed to it a "WhereTag" (RFID), and as it passes out of each manufacturing cell its location and status is real-time updated in the central manufacturing system, CIMPLICITY. This data is amassed in various databases and continuously examined, trended, and managed for exceptionswhich are wireless sent out via Nextel to manager's pagers. With such technology, a given assembly line can be "multitasked" (if you will) to build several different model cars as well as better handle customizations, greatly reduce floor space, and nearly eliminate transient inventory...all Muda.

^[1] http://www.managingautomation.com/maonline/magazine/read.jspx?id=1605638.

Is there really 85 percent Muda in an operation? Many factory owners would be inclined to think differently. They look around and see their 1,500 employees running around trying to keep up with machines and forklifts and even working overtime and weekends. Their common mistake is that busyness does not mean value-add from a paying customer's point view. (Customers will only pay for the values they receive.) Let's examine a simple hypothetical metal-stamping plant. The customer has ordered 10,000 units of some brushed-steel custom emblem. The manufacturer first buys the raw steel and because the customer ultimately receives steel he is willing to attribute some cost for this step. The manufacturer "receives" the steel, which is a process of coordinating a delivery, inspecting the raw goods (that should be exactly what was ordered), keying some payment system, and then moving the goods into some holding area. All Muda. The steel sits in inventory for some number of weeks and, when it is needed, it is first located by some floor personnel and then transported to the work machine. More Muda. At the machine, the material is re-inspected. (Is this the right raw material? For which customer? Which process steps must be performed?) Muda. Finally, the metal is cut and stamped and brushed, all values the customer is willing to pay for. On the outbound side, nearly the same set of operations take place with additional staging, delay (inventorying), and shipment to the buyer. Muda again. Indeed, there is at least 85 percent of nonvalue steps, time, and labor applied to the product creation.

Traditional factory optimization techniques would look at a single narrow activity, say that of transporting the raw material to the work area, and try for a 10 percent efficiency gain. Perhaps this year, they use propane-powered drivable forklifts instead of human-pulled hand-pumped ones. Inward focused, "squeeze" approach to improvement. Having the courage to "see" revolutionary approaches can be challenging. The revolutionary approach, in the metal-stamping example, would be more along the lines of: no inbound inspections of material quality because we're tied to the suppliers quality-control records, no inbound inventory holding because our supplier delivers the raw goods the moment we need it because he was tied into our "build schedule"records, no "what steps do we perform to this piece" at the machine because the piece comes with an electronic record of instructions, and so forth throughout the completion.

Another hypothetical example: ATM machines. Suppose a given regional bank wants to increase usage of its services and machines. It plans, like it always has, to add 10 percent more machines to a given large territory every year, year after year. If the bank steps far enough back, it's not machines that people want. People want an ability to buy things with the ease of cash without being encumbered to carry a large wad of cash. Eliminate the machines altogether. Instead, start a campaign to have the bank's ATM card accepted "as cash." Dramatic difference in capital and operational costs (and hence why we see a rapid adoption of debit cards at all retail locations).

ATM: Rise and Fall

Once upon a time, ATM machines were few and far between. One often had to drive to find one. Not now. One finds them in every conceivable location. The first proponents of ATMs were banks that realized that at 27 cents per ATM transaction, paying $1.07 per transaction at the teller window was less than cost-efficient. Next, retailers started installing them in the belief that easy access to cash translates to more money spent in the store (hence the proliferation of ATMs in convenience stores, malls, and markets).

However, debit cards challenge all that thinking. ATM use has fallen to 57 percent of U.S. households in 2003 from a peak of 65 percent in 2000^[2] (source: Synergistics Research, Atlanta). This is attributed to the rising use of debit cards, and is influenced by several factors:

• Consumer rejection of ATMs fees (considered too high by consumers)

• Perceived safety of carrying less cash

• Speed plus convenience of using the cards that act like cash

Even gas-station convenience stores accept debit cards today (or Speed Pass or other noncash fast-payment devices). The point is this: Adding ATM machines was solving the wrong problempeople are not looking for easier access to cash, they are looking for easier access to purchases.

^[2] http://www.mlive.com/news/grpress/index.ssf?/base/news-17/1095087375314090.xml.

Many new tools are available to help various companies interconnect their systems. Although e-mail, faxes, and even Web pages will continue to be useful for human communication, those are not the tools of the new trade. On the engineering side, virtual office and groupware allow for real-time collaboration on designs and documents through shared repositories and instant messaging. Asynchronous collaboration yields far higher efficiency and smoothes out time-zone bumps between manufacture and buyer and serial approval processes.

As wonderful as e-mail has been, it has a natural electronic boundary that solidifies the separation of two parties. Shared workspaces are an opposite mentality that essentially advertises that separate teams are virtually co-located without the barrier that a message-exchange mechanism introduces. On the manufacturing and fulfillment side, software exists today that directly attaches multiple companies' business systems together as a result of XML and Web services. The inventory level is directly known. The product location is directly known. The order quantity can be directly changed. Product options can be directly changed. No delays and no human-to-human involvement. Our point is not to detail any particular type of solution, but rather the more general notion that all the electronic streamlining you have grown to look for within your walls can now be done across your chain, if you're willing to take the walls down. Tie the business systems directly together.

Changing a business operational mentality can be tough. Let's examine inventory tracking for an example. If you want to know an inventory level in your supply chain 150 years ago, you wrote a letter and waited. Fifty years ago, you picked up the phone and queried. Twenty-five years ago, you might receive a fax on a regular schedule. Five years ago you might receive the information via an e-mail. Today, you might dial a special Web site and self-serve. All are incremental improvements but completely miss the real objective. Like the ATM example, the real objective is not to learn about inventory levels. What is even done with that information? Skip the intermediary step of discovering inventory levels (which is Muda) and jump directly to automated fulfillment, procurement, and receivables. Don't get cash to buy goods, go directly and buy the goods. Getting inventory knowledge was not the goal, and so improving the speed (of getting inventory knowledge) was missing the larger opportunity. The new opportunities are created by driving business systems together directly at an automated electronic non-humaninvolved levelbarreling through traditional walls.