Article                           Back to Home  Back to Articles

Flow Manufacturing is the Essential Component in Your Supply Chain Strategy
By Gerald Najarian

Well, it’s the dawn of the new millennium and the age of the Supply Chain.  We’ve successfully survived the Y2K non-crisis and, in the process, gotten a new ERP System to back bone the company’s information management infrastructure. Now we are entering the new age of Supply Chain Management trying to tie all the nodes of the supply chain together so that materials flow seamlessly along a path beginning with a customer’s order and pulling through from the raw material supplier all the way through the process.  All of this is good.  The previous ten years were turbulent ones for the manufacturing/materials executive as we absorbed the new disciplines of the period and saw the old ones disappear: ERP replaced MRP ll, the Just-in-Time wave from the far east morphed into Demand Based Flow, shop organization went from functional to product based and, the multitudinous materials management functions began to creep together into a homogenous unit known as the Supply Chain.

Supply Chain Management is to the twenty first century manufacturing enterprise as MRP was to the factory of the 1970s; the overriding framework that points the business toward the customer.  The customer of the MRP and post MRP era could put up with the fixed (and mostly long) lead times, inflexible product structures and, the high cost of batch/subassembly/final assembly manufacturing and inventory management of that time.  Not so the customer of the Supply Chain Management era.  As we begin the Supply Chain Management era with the market’s demand for flexibility, velocity and, minimal waste (not necessarily no waste), we now focus on this new overriding framework that is under-girded by ERP and empowered by modern Flow Manufacturing. 

Every supply chain is different and the driver of differentiation is the products or product lines in the particular supply chain.  In fact the same company can have more than one supply chain depending on the value chain perceived by the customers of a particular supply chain and it’s products.  So if supply chains exhibit different characteristics depending on their place in the customers’ value chain, it stands to reason that the heart of the supply chain, the flow manufacturing operation, should be specific to the particular supply chain.  Taking a hard look at products and supply chains and categorizing them permits us to start tailoring the flow operation to each one and thereby make the individual supply chains more sensitive to the customers’ value chain.

“Commodity” products.   These types of products exhibit characteristics of price elasticity and sensitivity and are usually only differentiated by customer service in the supply chain.  They have inherently low margins. The supply chain for these products needs to have all the waste squeezed out of it, be one that moves product through at high velocity, has short lead times and, is ripe for a low cost producer strategy.  Depending on the complexity of the product line, demand forecasts for these products should be reasonably reliable and finished goods inventory can be deployed in small buffer quantities in limited strategic locations.  If the product line has been designed with some interchangeable parts, modest WIP kanbans of  “bright stocked” parts can be maintained at the end of feeder shops.   Examples of such products abound: inkjet printers, Q-tips, store brand aspirin.  They are typically manufactured in a discrete or repetitive environment.

“Specialty” products.  Specialty products display inelastic pricing and, while not insensitive to competitor’s prices, are much less so than commodity products.  Margins in such products should be higher than those of commodity products.   In the customers’ value chain, these products are usually unique enough to forbear some supply chain inefficiency but reliable fill rates and lead times will still be required.  Since specialty products have inherent finished goods risk, inventory should be maintained as far back in the supply chain as possible.  These will usually exist in an assemble-to-order environment of one sort or another.   Some examples are:  laptop computers, specialty chemicals, specialized sporting equipment.

  MANUFACTURING CHARACTERISTICS

There are four manufacturing environments – make to stock (MTS), assemble to order (ATO), engineer/make to order (ETO and (MTO).  And there are four basic types of plants to accommodate these various manufacturing environments – V plants, A plants, T plants, I plants.  Recognizing the issues inherent in each of these types of plants and how those issues affect the manufacturing middle of the supply chain can make or break the velocity/flexibility/waste efficient supply chain we are trying to create.

“V” plants.  Many end items with a small number of raw materials characterize V plants and all end items are manufactured using the same basic processes: hence the name -“V” plant.  These plants and processes often utilize specialized capital-intensive equipment to manufacture commodity type products usually in an ATO/MTS environment.   The most notable characteristic of V plants is that of a process dominated by “divergences” as the product moves through the stages of manufacture.  Divergence means that the relatively few raw materials diverge into a greater number of intermediate items and those intermediate items diverge into still more items and so on until final process or assembly. 

Facilitating flow in such plants is focused on creating cells in the final process/assembly areas and managing queues in the homogenous process centers.  The problem is that queues will accumulate in front of bottleneck centers and the bottleneck will shift as mix changes cause divergence imbalance.  So, if we do two level scheduling – a final assembly schedule and a homogenous process master schedule at the last divergence point before the final process – we then have to manage the process bottleneck centers almost perfectly and make the assembly centers very flexible.  An “Advanced Planning & Scheduling” system that will effectively manage the queues accomplishes the former and set up reduction accomplishes the latter. 

To illustrate, imagine a videotape cassette plant in which there are make to stock and assemble to order end items.  Tape is made in an extremely complex process and then is slit to width and cut to length before assembly into a wide number of cassette cartridges.  To get the velocity demanded in this commodity type business, tape has to be scheduled to a small buffer of roll stock using an APS system to schedule materials and capacity simultaneously without “waste” of the capacity.   The slit and cut tape then feeds an automated cassette assembly operation that requires rapid breakdown and set up to assemble to customer order and to a minimal finished goods inventory.

“A” plants.   Few end items and many raw material and parts items in an ETO/MTO environment are the main characteristics of an A plant.  Here, the raw materials are often commodities, manufactured parts are unique to each end item and are manufactured in general purpose process centers, and the routings for these parts are dissimilar.  Unlike the V plant, the flow of materials and parts “converge” as they approach final assembly.  Convergence means that fabricated parts converge into a subassembly, which, in turn, converges into another subassembly and so on until final assembly.  The flow and supply chain issue in A plants is one of resource utilization and capacity planning in the general purpose work centers.  For materials to flow smoothly, it is vitally important that these centers not be overused by trying to amortize often-lengthy setups. Otherwise parts will move from center to center in bursts causing a center to be overloaded one day and underutilized the next with consequent negative effects on final assembly due date performance. 

Balancing flow in an A plant, therefore, is dependent upon reducing set up time in the process centers and shrinking intermediate parts lot sizes.  In addition, the bottleneck center(s) will have to have buffer stocks of the common parts in front of the center(s) to allow the center flexibility to produce non-common parts when needed.  Forecasting and master scheduling must be done at a very early convergence point in the process, maybe even at the raw materials level. 

Consider for example, a metal buildings plant; a custom engineering intensive business in which lead time performance is critical.  Many kinds of raw steel are fabricated into different building components in cutting, general machining, welding centers and then are “assembled” to order on a flatbed for shipment to a customer. Successful supplier management to receive steel timely and in small lots is dependent upon master scheduling at the first level in the “common” bills of material and at the raw steel level and letting MRP plan the inbound steel shipments.  The bottleneck centers, usually welding, need buffer stocks of often- used joists and beams to free them to produce custom components when needed.  Last but not least, set up reduction in all the centers will permit the small lots that preclude wave like materials flow.  With all these in place, buildings can be final assembly scheduled according to the lead time and the middle nodes of the supply chain guarantee that goods will flow through it without impediment.

“T” plants.  “T” plants are characterized by a large number of end items made from a relatively few common parts.  These plants are similar to “A” plants except that the process is dominated by divergences almost all occurring at or near the final assembly point.  Such plants are usually ATO or MTS plants in which the catalog of finished products is so diverse that carrying much finished stock is a risk.  What inventories there are must be in the common parts that can be configured into multitudinous end items.

Facilitating flow in a “T” plant is almost entirely scheduling and schedule integrity dependent.  With a large number of common parts being final assembled into many end items in small lots and many shop orders, the availability of parts at final assembly is crucial.  When parts are in short supply, those that are committed to pre-existing shop orders may be “appropriated” for another order considered more important by it’s foreman.  Such parts stealing impedes flow by introducing the chaos of missed final assembly order due dates.  To optimize flow, scheduling of end items to customer order in ATO environments or to small stock buffers in MTS shops and pull system “scheduling” of the common parts to time buffers in flow/JIT fashion is the path to follow.  Where bottleneck centers are extensive among the common parts fabrication centers, an APS would be the preferred approach.  Which ever is chosen, maintaining the time buffers is the overriding manufacturing management discipline to be enforced in a “T” plant.  

A good example of an “T” plant is a wiring devices plant – an MTS environment - in which sockets, receptacles, connecters are final assembled from numerous stamped/tapped metal parts, plastic molded items, purchased parts into thousands of different end items.  A plant such as this requires that parts from fabrication shops be available to assemble when stocks of finished goods diminish to the point where customer service may be jeopardized.  The constant danger is that inventories of these parts will grow and become a working capital burden on the company.  The tendency of these parts to grow in such a business is directly related to the complicated and often long machine setups in the fabrication shops and the natural desire to amortize the setup.  The key to optimizing flow (without large, financially burdensome inventory) is to insure a constant stream of fabricated parts in small quantities in a controlled environment to flexible flow assembly lines.  The critical needs are for setup reduction to allow small lot production and possibly for controlled storerooms to protect parts from misappropriation.  The scheduling options are threefold in this type of plant: use a pull system to schedule fabricated parts to time buffers and master schedule end items; use MRP to schedule fabricated parts and final assembly schedule end items to order; use a pull system to schedule fabricated parts and end items to time and stock buffers respectively. 

“I” plants.  Make to stock with many components and either few or many end items characterize “I” plants.  Typically, the process consists of many functional departments making common parts and subassemblies.  The parts and subassemblies “converge” throughout and especially at or near the final assembly point.  The stocked finished goods strategy often is to have product deployed in multiple locations.   These plants have issues similar to those of V and T plants where there are many common components made in large functional work centers resulting in queues in front of bottleneck centers and the temptation to “appropriate” parts. 

The way an “I” plant is scheduled is one of the keys to creating a smooth flow; the other is close tolerance finished goods inventory management.  Since these are usually MTS plants and most MTS businesses incur inventory risk when they carry any significant amount of finished goods, the natural desire is to deploy as small quantities of stock as possible.  Although there are other approaches, a good scheduling combination is to master schedule end items using a “close in” forecast or VMI and use a pull system to flow common parts to final assembly.

The “I” plant is often found in the consumer goods industries - electronic gadgets, sporting goods, tools, toys, home computing accessories – in which lead times are short and customers will fill shelves with another product if yours isn’t available.    A jigsaw puzzle plant is a good example.  “Parts” are imprinted and die cut in common process shops in numerous varieties and meet up with a significant variety of boxes in which to be packed at the final assembly line.  These are toys and are subject to changing tastes from season to season so carrying inventory presents risk.   Once the decision as to the product mix for the season has been made, a “campaign” can be launched to build some stock for the initial periods of the season but then the process centers that imprint and die cut have to be managed to keep a flow of small lots moving to the flexible assembly lines.  In an operation such as this, quick changeover and an APS in the process shops will keep inventory risk to a minimum.

THE “HEART” OF THE SUPPLY CHAIN

All the emphasis on the “extended enterprise supply chain” is appropriate to successful competition in the modern world economy.  Making suppliers adjuncts of their customer’s factories was the vision of the JIT movement and was clumsily done when it was done at all.  That vision is currently becoming reality with new and better attitudes about supplier management and relationships.   Now there is software to enable implementation of the supply chain management paradigm from beginning to end.  Certainly the new ERP packages and, the bolt on APS software and extended enterprise SCM software provide all the systemic tools we need for success in supply chain management.

The heart of the supply chain is it’s center, the shop floor, and for goods to flow in a seamless fashion from raw material supplier to customer, they have to get though manufacturing in a smooth flow.   A manufacturing operation that still processes goods in batch mode, however disguised, will be the impediment to successful supply chain management.  Identifying the supply chain needs of the market in relation to the company’s manufacturing environment and then executing a scheduling and operations strategy to meet the needs of both (supply chain/manufacturing strategy table) will transform even the most stubborn batch operation into the key contributor to a smooth and seamless supply chain.  

Related Services:
Supply Chain Management
Manufacturing Organization
Constraints Management