Getting Lean for Optimal Return
The goal for all manufacturing companies has remained constant for many years
— profits now and into the future. However, there has been significant debate
about what manufacturing strategy should be employed in any given environment
to achieve that goal. There is no debate about the importance of increasing throughput,
return on assets, and decreasing operating expense. This is exactly what Lean
manufacturing is all about. When you can get more throughput out of a manufacturing
plant with no increase in overhead or assets, significant results are seen on
the bottom line. This same idea holds true across the supply chain. The Lean approach
provides improved visibility, velocity, and reduces variability across the supply
chain, resulting in improved profits — now and into the future.
What Is Lean Manufacturing?
First of all, what is "Lean manufacturing"? Lean, sometimes known as "flow," is a philosophy that goes well beyond being merely a method of production and inventory control. It is essentially characterized by doing more with less — getting more quality output without a corresponding increase in resources. Many people mistakenly believe that Lean can be achieved only in environments where repetitive products are built consistently and in sufficient volume so that production lines may be dedicated to products or at least families of products. In practice however, Lean is more concerned with establishing an effective program of matching production rates with demand while eliminating all non-value-added activity. Lean has both a planning and an execution component. Planning is built around level or at least smooth production (make a little of what is needed each day rather than in large lots or waves). Execution "pulls" resources — driving replenishment based on sales (of finished goods) and usage (of components).
Lean leverages these consistent patterns of product-based scheduling so that this smooth drumbeat of production can be transmitted to suppliers as well as in-house operations.
The use of Lean techniques and principles are proven to shorten the production, purchasing, and supply-chain paths in a way such that throughput, and therefore return, is increased for the same asset base. This happens because inventory flows faster when there is less of it and when lead times are composed largely of value-adding activities.
When Can You Use Lean?
Is Lean for everyone? No. There are businesses that intentionally make it their policy to operate in ways that are counter to Lean theory, approaches, and practices. Then, there are other companies where Lean would be a great fit that have not achieved the proper disciplines necessary.
Since the most important prerequisites to employing Lean manufacturing are volume and product/component/methods standardization, companies that are true "job shop" operations may not fit the Lean model. This is because they typically employ one or more of the following: engineer-to-order sales and production methods, families-of-resource organization and layout, high setup times, and lead times that are usually long and variable. These companies sell their capacity. They do not forecast products — they forecast resource utilization.
Sometimes, a company meets the criteria of volume and standardization, but is not pursuing process improvement in the necessary areas sufficient for Lean to be effective. These areas include a pursuit of:
- Zero setup/changeover
- Zero defects
- Zero lead time
- Zero queue (and buffers)
- Zero inventory
We have listed these in descending order of importance to a Lean implementation. While quality is, of course, always of supreme importance, the Lean principles can still be practiced even in those situations or environments where yields are low. Regarding the inventory attribute, you could argue that reduced inventory is a result. However, when lot sizes are artificially high, excess inventory (and queues) are for no good reason added to the system.
On the other side of the spectrum are those companies that make standard products of sufficient volume and intentionally drive setup, lead time, queue, inventory, and defects to ever more-demanding new lows. Toyota and its famed Toyota Production System is an example of such a company.
JIT, Flow, TPS, Lean, and Focused Factory — So What's New?
We have talked about why and when, so now — what is it? Lean is a set of principles that emphasize these attributes:
- Lean production
- Visible production control
- Demand driven and demand smoothing
- Pull versus push
Beginning with the Toyota Production System, many offshoot modes of operation have emerged. Among these are: Just-in-time (JIT), Kaizen (continuous improvement), Kanban (pull), continuous replenishment, and many others. To a large degree, they are all part and parcel of the same thing.
Lean is a child of these movements. It embodies the spirit and approach of each of them. With Lean, there is a bias toward keeping demand consistently level, with incremental and smooth adjustments. There is a preference toward "visible" production control versus complex scheduling logic. The ideal Lean plant involves these aspects: simple, visible production control methods that mirror the shop floor (and supplier) environments they manage; placing a limit on the amount of inventory that can be in the chain (in-process, in inventory, or in a buffer); all supply chain, production, and procurement operations are clearly defined and composed of standardized, repeatable procedures.
Additionally, Lean brings its own heritage of methods and terms. Probably the most important of these is "TAKT" time. TAKT is a German word that literally means the "beat" that an orchestra plays to, as given by the conductor. The primary goal in Flow is to achieve consistently spaced production signals for all items based on their leveled demand.
The key point here is that demand is the driver — not clumsy and wasteful supply policies like EOQ (economic order quantity), most of which are attempts to balance two or more undesirable affects like carrying cost and setup cost. But all businesses have demand, so why can't everyone do Lean? Job shops have demand, but it is not consistently reliable, and so much buffering (inventory and/or lead time) must take place to protect key assets (like machine utilization).
Lean is also characterized by minimal reporting. When operations are synchronized, there are fewer control points to measure. When good, effective standards are in place, there is less pressure on tactical performance measurement (like labor efficiency). The TAKT sets the pace, and the pace (and its results) is visible.
Paradigm Shift Moving to Lean Manufacturing
In a batch or discrete manufacturing environment, work is pushed throughout the plant. A work order is created to signal production of a product with a defined quantity. The pick list and schedule is then generated and associated with the work order to communicate delivery of the raw materials required to create the product. Raw material is stored in a centralized location typically a distance away from the point of use. The schedule is created to deliver raw material in the most efficient manner and defines what parts should be picked, the routes they follow, and the estimated time they are expected to arrive and depart each work center throughout the production process. In an effort to optimize material handling, parts may be picked together and moved to their appropriate work centers prior to the operator's need.
The raw material then waits in queue until the operator needs the material. Once the parts are needed, the operator accesses the order in the system and the material is consumed. A transaction is processed for moving and completing the work. The work order then moves throughout the remaining steps, experiencing queues between move and completion transactions. When the product is finally completed, it is moved into finished goods inventory where it waits to be delivered to the customer. Implicit in this approach are management complexities derived from scheduling material handling and inefficiencies experienced in non-value-added move and wait steps.
In a Lean environment work is triggered by demand, not pushed via schedules. In a pure Lean environment, there may be no work orders, pick lists, or transactions for moving material inside the production process. This is accomplished by putting the emphasis on visual management. Raw materials used by the operators are stored near their point of use. Material is pulled from convenient locations when it is needed. Kanbans are used instead of schedules to signal work and control WIP. Dedicated product lines are balanced with respect to production demand so each operation performs within the TAKT time. Management is committed to JIT principles and supports facilities redesign to accomplish a flowing product view of the plant versus a discrete process view.
Process Engineering and Labor Flexibility
Facility redesign and increased labor flexibility are typically the most noticeable changes for operators and supervisors moving to a Lean methodology. In a discrete plant or job shop, departments are segregated by process (milling, stamping, welding area, etc.). The methods and metrics they focus on are with respect to their own departmental efficiencies. This can lead to major inefficiencies for the plant with regard to product cycle time, inventory buildup, and quality. In a Lean environment, machine and storage equipment is reorganized to support dedicated product family lines. Often new machine and storage equipment is purchased to support this approach but the benefits outweigh the costs as quality, lead time, and inventory are improved.
Overcoming the methods and attitudes practiced in discrete environments is the biggest obstacle for implementing Lean manufacturing. Information systems on their own will not solve these problems. As stated earlier, it takes management commitment, extensive training, and skillful process design. Information systems can support in certain areas and provide a point of control for maintaining the methods. Defining and maintaining process definitions, demand planning, line balancing, and Kanban management in an ERP system leverages the efforts of key personnel to make Lean manufacturing work. Operators, supervisors, engineers, buyers, planners, vendors, and customers all need to share in the commitment to obtain the highest return.
Reorder Point, Pull, and Lean In: What's the Difference?
Reorder point (ROP) has been around as long as inventory has been controlled. The idea is simple, if not always easy to practice successfully. For each item or product in each stocking point (or point of use), have a systematic signal or tickler alert the user or responsible inventory control person when it is time to replenish. The main point is that inventory level itself provides the basis for this signal — when inventory reaches a prescribed level (quantity), place a replenishment order. The tricky part is this: What should that level be?
The theory behind the reorder points goes like this: assuming demand is relatively constant, and assuming lead time is reliable and not variable, then when inventory reaches a point where you have enough to cover lead time (usually plus some safety stock quantity), it is time to order. For this reason, the reorder point is also known as (and calculated by) "demand during lead time."
As far as a theory goes, ROP has been successfully practiced over the years. But it has also met with failure too. When it failed, it was largely based on three factors:
- Demand was not constant
- Lead time was not reliable or varied significantly across planning periods
- Demand was calculated exclusively based on historical sales and usage
The first two problems were considered to be outside the control of most production and inventory managers: machines will break down; supplier priorities are not easily controlled and monitored; customer orders are based on their economies and not ours; and so on.
The advent of JIT and other continuous improvement methods focused largely on the lead-time component. Rather than push inventory through the system, inventory that was often just not needed (like when lot sizes were inflated beyond true needs), material would be pulled based on actual usage. The Pull system was born. The idea is this: Even if you think you will sell/use 100 units this period (day, week, month, whatever), break this into smaller sub-lots, make one, and then wait until it is used before making another. Many benefits came from this simple but extremely elegant idea: shorter lead times, improved quality, more reliable scheduling, lower inventories, less non-value-add time throughout the chain of operations affected.
A special form of Pull system was recognized as being a model for efficient, yet simple, operations: the "two bin system." This system worked like this: For a given item/location combination, set up two bins such that each bin contains a demand-during-lead-time quantity of parts. Then, when one bin becomes empty, it becomes the signal (the Kanban). While that empty bin is being replenished (it's lead time including move, queue, setup, run, and post-op), the full bin would serve the usage point(s). In this way, the two bins would rotate on a lead-time cycle for each (plus some safety stock, of course!).
In effect, the two-bin system was a form of ROP and Kanban.
But now the problem became how to solve the fact that demand was calculated exclusively based on historical sales and usage. Since master scheduling and MRP technologies and techniques had been invented, in part, to cure this backward-looking problem, could they help the Pull system too?
The answer was yes. And this is exactly what the Lean solution does: Give you the tools to establish a future-demand-based (what one practitioner once coined as "future history") two-bin system operating in a more reliable and stable environment.
Planning and Execution
Traditional production and inventory control systems have approached their goals this way: Establish plans and then drive execution to match these plans. With Lean manufacturing it doesn't work this way. Instead, the strategy is to establish strong execution systems (like supplier relationships and source inspection and pull philosophies) and then use the planning system to support execution. Of course everything must tie into the business plan. But the departure from traditional thinking is profound: maintain stable operations by introducing change carefully and incrementally.
The Lean solution advances your operations to a high state of flexibility by planning TAKT time, and by establishing what is called an n-bin system across a bill-of-lines.
Demand Is Everything
With Lean, demand is everything because the ideal goal is to make what you need, not building anything in advance of actual requirements. In this respect it is very much like JIT. All of the principles of Lean are aligned with this objective. This is also why Lean is easiest to implement in environments of high volume and standard products (or at least configured products made from standardized components). However, Lean applies also in operations where the process is repetitive and the product varies widely.
High-Quality Demand
Even when the production mode is make-to-stock and product lifecycles are short, it can be difficult to accurately forecast production requirements. One of the prerequisites to Lean is a program that includes these attributes:
- Shortening of the supply chain in order to accelerate the capture of demand information
- Ongoing attempt to improve forecast accuracy
- Demand-smoothing program
Techniques for supply-chain streamlining are outside the scope of this paper, as are approaches toward increased forecast accuracy. Remember that the best forecast is no forecast. This is possible when the production lead time has been reduced to less than the customer demand time. Demand planning and smoothing, however, are key aspects of the solution.
Demand Smoothing
As mentioned before, having a level demand is very important to achieving the benefits of a Lean implementation. A Lean solution assists with this by capturing demand at the item and business-unit level for a given time horizon.
As demands change, a tolerance can be employed to permit a certain level of variation from planning period to planning period with no adverse effects. This is known as a "flex fence."
Supply chain planning can help you establish a smooth, level demand for input into final assembly scheduling and sequencing. That way you can move demand from line to line to meet capacity and resource loading needs and policies.
Calculation of TAKT Time
If level schedules are important to passing stable pull signals down through all preceding operations, then the calculation of TAKT time is also important, and sometimes even a prerequisite. A Lean system takes statements of demand by product and location (or business unit) and "smoothes" that demand — in this way computing a level frequency of production for a specified time horizon.
This process starts with the final assembly schedule and then proceeds in an explosion through all feeder lines that are also operating in a Lean mode.
One of the most important elements of success with this phase of the plan is a stable master schedule. As lead times are reduced (using Pull and supplier JIT contracts), the pressure is reduced and flex fences can be employed to dampen the effect of variable demand like forecasts, and to permit more flexibility in the business planning process.
Calculating TAKT times for all Lean-coded items builds inherent stability and regularity into lower level schedules, and support further lead time reduction. In this way you achieve flexibility through synchronicity — eliminating the "tsunami effect" as a disruptive source of noise throughout the manufacturing and even procurement processes. Supply points, both inside and outside, are more capable and can more readily meet their delivery requirements when such regularity becomes systematic and formalized. Of course, all of this also depends on well-engineered, standardized operations.
Bill-of-Lines
With a Lean solution, you can structure the planning process so that planning parameters (like replenishment lead time, reorder point, pull, and demand) are unique by item and production line. This permits you to plan the pull system based on item usage in every point-of-use.
You can also review the results of the planning process after each line is planned and before sending Kanban plans down to the next level. This ensures adequate safety has been built into the plan.
Standard Work: Operational Method Sheets
Operation method sheets are graphical illustrations used to convey assembly instructions. This is a key element to ensure quality between operations on the assembly line. Method sheets replace traditional sub-assembly drawings and text intensive standard operating procedures as they are easier for the user to view and use on a regular basis. Method sheets can be supported in PeopleSoft manufacturing as attachments or integration to document management systems. The end user can make reference to the flow-routing steps on the method sheet and attach a pictorial to define the items and quality steps used in the assembly or test process. In addition to standard bitmap attachments, multiple formats of multimedia are supported for audio and video.
The n-Bin System
The PeopleSoft system supports what we call an n-bin system. Based on the traditional two-bin system, the Kanban planning component gives the planner the flexibility of entering Kanban container size and then having the system calculate the number of Kanban containers required to meet demand. This panel shows demand expressed as TAKT time, as well as the Kanban signal to replenish (reorder point), and the order quantity (in number of Kanban containers).
By calculating the number of Kanbans, a Lean system integrates the important principles described earlier: a future-demand-based n-bin system that integrates the use and benefits of Pull with the stabilizing effects of high-quality planning.
Lean and ERP in a Mixed Mode
With the right solution, you get the best of both worlds: demand-driven pull technology coupled with future-oriented planning. The setups and planning parameters provided give planner's the "control knobs" they need to stay in control with minimal maintenance — keeping overhead low. Here is an example: You can code an item in a point-of-use location as "Auto" kanban mode. The system will perform these functions:
- Calculate the number of Kanbans and Kanban containers needed
- Signal replenishment when a Kanban is available
- Create a schedule for completion reporting and component backflushing with no planner intervention
- Collaborate with suppliers in a Lean environment
Kanban Planning and Capacity Management
Lean can peacefully coexist with MRP. Lean can optionally pass reorder signals to MRP for procurement or even production optimization.
Capacity planning can be used with Lean in several ways. One way is to use capacity planning logic for rough- cut planning to test the feasibility of the business plan or the demand plan before exploding to lower level plans.
Another way is to synchronize production planning so that cells and fab/machining centers that support flow lines have the necessary dependent demand so they can be optimized as needed. This is particularly important when lower level work centers are still setup intensive and queue oriented.
Summary
Technology can only provide value if it addresses a limitation to success that is currently facing the company. In today's rapidly changing market, having an agile response time is critical for driving a sustainable competitive advantage. This response time is driven by the real-time knowledge of customer demand supported by a Lean enterprise, and this Lean enterprise must also be enabled through a lean, agile supply chain. With PeopleSoft, the technology enables these win-win relationships from customer to supplier's supplier so that each link can be demand driven and realize its profit potential — now and into the future.
