Six Sigma has evolved into a business improvement methodology that focuses on how variation is affecting organizational desired results. Six Sigma project teams follow the DMAIC model to drive rapid improvement. DMAIC is an acronym for Define-Measure-Analyze-Improve-Control.

Define: Typically in this stage a team is assembled, a project charter is developed, customer Critical to Quality (CTQ) requirements are defined, and a process map is created. The charter will clearly define the business case for doing the project, state the problem, define the scope, set goals, and milestones, and spell out the roles and responsibilities of team members. In identifying the CTQ issues, we must define customer characteristics that have the most impact on quality. The process map, called SIPOC (Suppliers, Inputs, Process, Outputs, Customer), defines a high-level process map of the project focus.

Measure: In this step, we define what to measure- develop a data collection plan and perform a baseline capability study to calculate the baseline sigma.

Analyze: It is important not to jump to improve before verifying why the problem exists. The main areas to look for causes of defects are data analysis, process analysis, and ultimately root cause analysis.

Improve: This step takes all the data from the D, M, and A steps and develops, selects, and implements solutions that will reduce the variation in a process.

Control: Sustain the new process through a robust monitoring plan.

The main purpose of the DMAIC process is for process improvement. When a process is at its "optimum" and still doesn't meet expectations, a redesign or a new design is needed. This is called Design For Six Sigma (DFSS). DMADV (Define-Measure-Analyze-Design-Verify) is a common acronym used today for DFSS.

Motorola was one of the first companies to realize that a metrics and methodology approach was still not enough to drive "breakthrough" improvement. They continued the Six Sigma evolution into what is called the Six Sigma Management System. A Six Sigma Management System is a structured process to ensure that all improvement efforts are aligned to business strategy. Six Sigma has become a top down approach to execute strategy through the alignment of all improvement activities to assure fast, sustainable growth.

Theory of Constraints (TOC)

The basic concept of TOC is often introduced through the chain analogy. A chain is only as strong as its weakest link. Improvement that does not improve the performance of the weakest link most likely does not improve the system and can be considered waste. Many claim TOC is just common sense, but it is surely not common practice.

Introduced by Eli Goldratt in the mid 1980s, a wide awareness and understanding of parts of the TOC methodology was first accomplished through people reading the book, The Goal (Goldratt and Cox 1984). Although many of the TOC basic concepts were discussed in The Goal, the complete body of knowledge was not.

Some people think of TOC as simply finding and speeding up Herbie (the fictional Boy Scout in The Goal), the bottleneck. Then they find the next Herbie and the next Herbie, etc. TOC is not about chasing Herbies. More accurately, TOC is about how to improve and manage how the system constraint (Herbie) performs in the context of the total system. This is quite different. It is about managing the total system, which is comprised of interdependencies, variability, and constraints, to ensure maximum bottom-line results for the organization. TOC is about focusing first on the system's leverage points and then on how all parts of the system impact the operation of the leverage points. This is the way to achieve total system improvement, not just localized improvements.

TOC applies the logical Thinking Processes (TP) used in the hard sciences-cause-and-effect-to understand and improve systems of all types, but particularly organizations. The process a doctor would follow if you went to him with an illness, first Diagnosis, then Design of a treatment plan, and then Execution of the treatment plan, is the same process followed by TOC with the use of three questions, What to Change, What to Change to, and How to Cause the Change.

One of the core beliefs of the hard sciences is that for many effects there are very few causes. Using the construct of "cause and effect" becomes increasingly important as we perform scientific analysis. All too often, we see organizations treating many "symptoms" instead of addressing the root causes. TOC looks for the core conflict that holds the root causes in place.

Think of an organization as a "money-making box" (see Fig. 36-3). It is first primed with investments in equipment and Inventory (I). Money is continually poured in as Operating Expense (OE) to pay for people and other ongoing expenses. The people process the Inventory and sell their products to make a larger amount of money called Throughput (T) (money generated by the system through sales).

The TOC systems approach requires that you first understand the system, its goal, and measurements. Then you can apply the Five Focusing Steps2 (Goldratt, 1992, 307): 1. Identify the constraint(s).

2. Decide how to exploit the constraint(s).

3. Subordinate/synchronize everything else to the constraint(s).

4. If needed, elevate the system's constraint.

5. If the constraint has been broken, go back to Step one. Do not let inertia become the constraint.

The application of these steps in a situation where the system constraint is physical is usually obvious and straightforward. However, often it is not a physical constraint. The nature of many constraints in organizations is policy constraints. In that case, the Five Focusing Steps break down into three questions (Goldratt, 1990, Chapter 2): FIGURE 36-3 Money Making Box (Adapted from the "Structured Presentation" 1990, Avraham Y. Goldratt Institute) Used with permission, Avraham Y. Goldratt Institute, a Limited Partnerhip.

1. What to Change?

2. What to Change to?

3. How to Cause the Change?

The TOC methodology looks at the world through the eyes of cause-and-effect logic and focuses on managing system constraints, interdependencies, and variability.

Discords that can Block the Effective Integration of TOC and Lean Six Sigma (LSS)

There are many synergies between the methodologies. They are all customer focused and want to provide the best value for the customer. Lean and TOC focus on creating a pull system to increase flow through the process and shorten the lead time to market. However, there are several discords between the methodologies that if not handled carefully will diminish the gains the organization can achieve from their improvement efforts.

In the early stages of the "design" of a system, there is a difference in approach between Lean and TOC.

Most Lean designs calculate Takt time, the rate at which you need to produce to meet customer demand, and then attempt to balance resources and equipment to that rate. Capacity in any operation that is greater than the amount needed to satisfy demand is considered waste. Improvement initiatives then focus on how to eliminate the waste in order to "balance" out the capacity and be equal to the demand. Due to variation, most Lean designs today will make sure that the cycle time of each operation is some percentage below Takt time, but the goal for the design of the "ideal" system is to have a balanced line with little or no "excess" or waste. In this "ideal" system, the capacity of each operation in the system would be balanced to support a cycle time just slightly shorter than the Takt time. Note that in this case, every operation in this ideal system could become the system's constraint if there is any variation in demand, product, or processes.

FIGURE 36-4 Balanced or unbalanced.

The TOC approach believes that there is a constraint in every system, and the constraint dictates the output of the organization. An hour lost on the constraint is an hour lost for the entire organization; thus, we don't want to "starve" the constraint. A TOC design would have some sprint or protective capacity on non-constraints to ensure that the constraint can be exploited to the fullest extent possible. This "unbalanced" capacity allows all operations to focus on how they are impacting the operations of the constraint and thus how their actions are impacting the Throughput of the total system. Figure 36-4 shows the difference in how a balanced line and unbalanced lines are set up. When integrating TOC and Lean, the correct choice must be made.

If there is no variation, in either process times or demand, a balanced line can work. This is obviously not very likely and Dr. Deming suggests there will always be variation. An unbalanced line enables one to protect Throughput from that variation. Variation anywhere in a balanced line can immediately have a negative effect on the Throughput of the organization. Continued variation at different operations in a balanced line will dictate that you eliminate variation in the entire line in a very quick manner, which would often be a very huge and costly task. "Focus on everything, and you have not actually focused on anything" (Goldratt, 1990, 58).

The unbalanced line approach focuses on the constraint and ensures that non-constraints have enough protective capacity to catch up to the constraint when "Murphy" strikes. Eliminating variation is still a priority in an unbalanced line. The difference is that the focus of the improvements are directed to what will rapidly improve and protect Throughput while reducing Inventories (or other investments) or Operating Expenses.

In summary, both designs are set up to meet customer demand. The balanced line works well when there is little or no variation in product mix, process times, or demand. The unbalanced line works well in the presence of variation in product mix, process times, and demand. While variation reduction is a priority in both designs, the difference is where and how many places one must focus and what the impact will be on the Throughput of the total organization. The constraint in the unbalanced line is managed very tightly. Efficiency and predictability at the constraint are important metrics. The non-constraints are measured on their effectiveness in keeping the constraint supplied-this is called time buffer management. The output of the total system is the overall top metric.

Work Behaviors

The balancedunbalanced design decision dictates how resources will be measured and ultimately how they will behave. Lines with balanced capacity expect workers to work to Takt; unbalanced lines would have workers working to the "relay runner"3 work ethic. Figure 36-5 depicts the discord between working to Takt and working to the Relay Runner ethic.

FIGURE 36-5 TAKT or Relay Runner work ethic.

Once Takt is determined and the line is balanced, the operator is to work to Takt. This works well when there is little or no variation in product mix, process times, or demand. However, if you have negative variation in the actual versus planned processing time of an operation, the work is blocked from moving to the next operation at Takt time. This results in a negative impact on Throughput and typically calls for inserting coping mechanisms on the shop floor. When there is positive variation, the worker has no incentive to pass the work on quickly so there is little opportunity to do other value added work.

Behaviors common in a work to Takt time environment are the student syndrome and Parkinson's Law. With the student syndrome, you think you have ample time to finish the task and therefore hold off starting the work until the last minute. If variation occurs after the last minute start, the work is finished late. Parkinson's Law states that, "Each task will expand to fill the allotted time available."

In this environment, improvements are masked due to these policies and behaviors. Early finishes of each operation are not passed on and late finishes by any operation can disrupt meeting the Takt time of the total system. This is the result of having protection that by policy is isolated within each operation and therefore cannot be aggregated to protect the total flow time. When Takt time is violated in one operation, the entire line suffers the consequences.

The relay runner ethic emulates a finely tuned relay race team. When work is present, the operator works head down at a fast pace that is consistent with quality and safety until the work is completed or he is blocked. Should the operator become blocked, he works on the next sequenced job until the previous work becomes unblocked. This eliminates the student syndrome and Parkinson's Law effects while exposing improvement opportunities. In the relay runner environment, early finishes are passed on immediately and are aggregated to form time buffers that protect the constraint and the delivery to the customer from variation in process time or demand. Thus, the on-time delivery and Throughput of the system are protected even in the presence of significant variation.

In summary, note that how you designed your line-balanced to Takt or unbalanced-will dictate if the system works to Takt or applies the relay runner work ethic. In recent years, there have been many "workarounds" offered to try to make a balanced line work to Takt time in the presence of variation. These "workarounds" often redesign the line to an unbalanced state.

Material Release

Another subtle difference in applying TOC or LSS to a system is how material is released into a system. Both systems are pull systems based on responding to customer demand. The main difference is that the TOC signaling method is based on time, while the LSS method is based on inventory.

As shown in Fig. 36-6, when there is demand on the time-based system (known as Drum-Buffer-Rope[DBR]) there is a signal sent to the constraint for scheduling purposes to meet a shipping request, and a signal is sent from the constraint to the beginning of the line (production control) for timing the release of material. As discussed earlier, this is an unbalanced line. The non-constraint resources have "catch-up" capacity to assure orders get to the constraint on time and to the customer on time even in the presence of variation. Buffer times are calculated from the constraint to the shipping point, called the shipping time buffer, and from material release to the constraint, called the constraint time buffer. These buffers will absorb variation in getting to the constraint and to the customer, thus protecting Throughput. Material is released based on the time buffers and the actual run time of the constraint. Material is only released into the system when there is a pull from the customer; therefore, the WIP in the system is based on customer need and what the constraint can produce. There is no standard number of units of WIP, but the WIP is based on the amount of processing time that it will take on the constraint resource.

In the time-based system, high variation in demand, product mix, and process times are accommodated through adjustments to the two time buffers. These time buffers act as shock absorbers to all of the operations preceding them. Instead of providing large buffers to accommodate variation at each individual operation, the Relay Runner work ethic allows the buffering to be aggregated just in front of the constraint and in front of the customer. The protective capacity of non-constraint resources coupled with the Relay Runner work ethic allows them to catch up when there are disruptions any place in the system. Some protective capacity is usually available at the constraint resource as well. This allows it to catch up when it is the cause of disruptions.

As shown in Figure 36-6, an inventory-based release system (Kanban Manufacturing System) is activated when there is a customer demand. A signal to produce, called "kanban" is sent upstream link by link, as material is pulled to satisfy and protect customer requirements. This process is continued until all supermarkets needing replenishment are filled. The Kanban system is a system of visual signals that triggers or controls material flow. The Kanban in each supermarket is set to restock each part to its "Standard Level" once the signal is sent to reorder. Kanbans synchronize work processes across a system. In this system, nothing is produced unless there is a signal to produce.

FIGURE 36-6 Release of material-time or inventory.

FIGURE 36-7 Replenishment system-time or inventory.

In systems with high variation in demand, product mix, or process times, the inventory-based system will not work effectively. In the inventory-based system, high variation in demand, product mix, or process times can lead to high variation in the Takt time, which can require frequent rebalancing of a balanced line. Variation can create wandering bottlenecks, which can disrupt the flow through the line and have a negative impact on the Throughput of the system.

Replenishment System

Another subtle difference between TOC and LSS is determining the size of raw material and finished parts inventories and the difference in the mechanism for triggering the need to resupply them. Figure 36-7 illustrates a traditional replenishment system4 and a TOC replenishment system. In a traditional replenishment system, the size of the parts inventory is based on a min-max type of system with the reorder point to resupply based on a predetermined physical quantity remaining, often known as the reorder point. TOC sizes the buffers based on demand patterns during the time to reliably replenish (TRR). The TRR includes a fixed reorder time interval (e.g., once a day, once a week, etc.) and that time interval is the signal to resupply the parts inventory with what has been consumed.

This is a time-based replenishment system versus an inventory-based replenishment system. The batch size is variable based on the demand during the fixed reorder interval. The inventory-based system has a fixed minimum batch size (the maximum level minus the reorder point) and the time interval to trigger resupply varies. The "time"-based system handles variability much better than the "inventory"-based system because the time-based system's replenishment time is bounded. In the inventory-based system, the time to trigger the replenishment is unpredictable and can be very long.

The time-based system will work effectively in any environment. The focus is on managing the flow of parts in time versus managing levels of material. It really comes down to what makes you pull the replenishment trigger-time or parts.

Figure 36-8 reveals the design differences that you must be aware of when integrating TOC and Lean.

FIGURE 36-8 TOCLSS design choices.

The design choice between a balanced or unbalanced line will lead to different resource behaviors and replenishment systems. Despite what some say, the designs are not "the same just different"; the design intent is different and you will get different results depending on the environment. The "balanced" design works very well in the absence of demand, process time, and product mix variation. The unbalanced line, typically thought of as the best way to go in low volume high variability environments, actually works best in all environments.

How effectively we integrate the three methodologies depends on the design choice path that is taken. If the Lean design path is taken (balanced line, work to Takt, inventory release and replenishment), then only two of the TOC Five Focusing Steps can be applied-Step 1: Identify and Step 4: Elevate. These two steps will need to be applied continuously to identify and eliminate each new constraint. During this effort, the process will not be stable or in control. If the organization wants to experience the full power of the TOC Five Focusing Steps, the other design path (unbalanced line, relay runner work ethic, and time-based release and replenishment systems) must be followed. This path provides early system stability and focused system improvement.

TOCLSS-Fully Integrated TOC, Lean, and Six Sigma

The most powerful way to integrate TOC, Lean, and Six Sigma begins with strategy. The strategy provides the roadmap to improve business results. This strategic roadmap provides the direction for the areas of the organization that can most benefit the total system by applying improvements first. The system design of the first area provides predictable and stable system performance by focusing on protecting and managing the constraint(s) of the total system. Once this is achieved, process improvement efforts can be applied in a focused way to provide even more bottom line results for the organization. Finally, the improvements must be sustained in order for the organization to achieve real bottom line results over time.

In Fig. 36-9, the SDAIS Model illustrates the deployment framework to ensure business success by driving effective focused process improvement through TOCLSS, from a stable operational platform.

FIGURE 36-9 The Velocity Approach ((C) Avraham Y. Goldratt Institute. LP 20062010.) The Velocity Roadmap to continuous business success has three major parts: the constraint-based system architecture and the TOCLSS improvement architecture, combined with the SDAIS deployment framework. In order to begin to really improve what is important, you have an understood direction and an aligned stable platform that delivers reliable, consistent Throughput.

Strategy- The output of a good strategy session is a clear, agreed upon roadmap to improve business results. The TOC strategy process involves using cause-and-effect logic to understand the core conflict of an organization, validate the conflict, and then develop the future reality, which breaks the conflict and adds other "injections" needed to improve the system. Roadblocks are removed and the result is a strategic roadmap to the future. This is done using rigorous cause-and-effect logic, which not only shows the sequence but also the interdependencies in the plan. This is much different from most strategic plans that end up being no more than an isolated list of actions from each department. The focus is on optimizing the performance of the total system versus improving the individual departmental functions in isolation.

Design- Operational/functional leaders and subject matter experts design their operations to align their business processes to achieve the identified strategic bottom-line results. During the design process, they reconfigure the operational model, policies, measurements, roles and responsibilities, and information systems within the context of strategy and proven TOC solutions and execution management tools.

Activate- During the activation process, the newly defined policies, measurements, roles and responsibilities of the operational model and the information systems, and execution management tools are implemented to make the design operational.

This constraint-based system architecture will produce a system where business processes are designed, aligned, and operated in a stable, predictable manner.

Once a system is stabilized and is delivering stable predictable results, ongoing focused system improvements are applied that result in increased sustainable bottom-line results. TOCLSS uses the synergy of TOC, Lean, and Six Sigma to coherently achieve focused system improvement (FSI) beyond what might be accomplished by applying each method individually with a traditional continuous process improvement (CPI) approach.

Improve- Once a more stable operational system exists, the energy is turned to focused improvement efforts to drive the operational system to achieve the desired effects and strategic objectives identified in the organization's strategy session. Improvement efforts are evaluated based on their ability to increase Throughput, and to reduce Inventory and Operating Expense and advance overall system performance (Jacob, Bergland, and Cox, 2009). Key performance indicators (KPIs) are examined to identify gaps between present and desired performance levels. The gaps are analyzed further and opportunities are assessed to focus improvement efforts at the business process level to achieve the desired outcomes. Improvement experts determine which improvement technique(s) are needed and then identify improvement project priorities. Some useful improvement techniques include 5S System, Standard Work, Rapid Setup Reduction (SMED), elimination of non-value added waste, Total Productive Maintenance (TPM), Point of Use Storage (POUS), Mistake Proofing (Poke Yoka), Visual Tactics, Control Charts (SPC), Capability Studies, and Design of Experiments.

Sustain- Organizational memory is created and supported by establishing the documentation of the strategy, operational design and the focused system improvements details. The organization continually reviews key measurement results to assess, address, and institutionalize the policies, measurements, and behaviors to guarantee that the results are sustained and do not degrade. The organization ensures that they have continued capability to achieve buy-in and maintain expertise.

Following the SDAIS process eliminates the need for an organization to have to "choose" a methodology, or to use the "toolbox" approach randomly. The organization can utilize the full integration of TOC, Lean, and Six Sigma in order to obtain focused system improvement that achieves real, sustainable breakthrough performance.

References

Blackstone, J. H. Jr. 2008. APICS Dictionary. 12 ed. Alexandria, VA: APICS.