The critical feature of a T-plant is that the final products are assembled using a number of component parts and these component parts are common to many different end items (in contrast to an A-plant). Because of this sharing of components, the assembly part of the product flow has the structure shown in Fig. 8-16. Note that the number of end items is larger (much larger) than the number of component parts. This creates the sudden explosion of the PFD to create the T-shape. To illustrate the magnitude of this explosion, consider a case where there are six component parts and each part has four variations, giving a total of 24 different components. The number of possible end products is 4 4 4 4 4 4 = 4096!

Most manufacturers of consumer products are T-plants. Consider the production of personal computers. The basic elements-hard drive, processor, memory, display, etc., are available in a few variations each. For example, the hard drive is available in 40, 60, and 80 G sizes. The processor might be available with speeds of 1.8, 2.0, or 2.4 GHZ. As illustrated above with just a few such variations, the number of distinct computers that the manufacturer produces can be very large indeed.

The characteristics of a T-plant are: 1. A number of common manufactured and purchased parts are assembled together to produce the final product.

2. The component parts are common to many different end items.

3. The production routings for the fabricated component parts are usually quite dissimilar.

FIGURE 8-16 Product flow diagram for a typical T-plant.

The dominant characteristic of a T-plant is that the assembly point is actually a divergence point. The same component part (80 G hard drive) can be assembled into a very large number of different end units. Unlike a V-plant where the divergence points are spread through the operation, the divergence in a T-plant is concentrated in the assembly area. The impact of this is devastating. We have seen the impact of a simple divergence point in the case of V-plants. In a T-plant, the divergence is assembly and this means that not one but all components are diverted to the wrong product if assembly produces the wrong item. This significantly magnifies the impact and spreads through the whole system like wildfire. This is illustrated by the simple case shown in Fig. 8-17 involving four component parts A, B, C, and D, and four assembled Products E, F, G, and H. The arrows show how the products are made and the figure indicates the inventory available for each part. Now suppose that an order for 100 parts of Product E is due to be assembled and shipped. The assembly of Product E requires 100 units of part A and 100 units of part B and is next on the assembly schedule. However, as shown in Fig. 8-17, part A has zero inventory. An expediter will have to be dispatched to expedite 100 units of part A. In the meantime, the assembly operation is going to be idle. However, it is possible to make 100 units of Product H, which requires part B and part C. In most cases, the assembly will not be left idle. Product H will be produced, since it is an active part and might very well have an order due next week. Note that this action consumes the available stock of part B, while at the same time creating finished inventory of Product H. Alternately, Product E is behind schedule while Product H is ahead of schedule. However, the real damage is revealed when part A finally arrives at the assembly area. It is still not possible to assemble Product E because we are now short of part B that was consumed in the production of Product H.

FIGURE 8-17 Example of the phenomena of "stealing."

The concerns or issues that are shared by T-plants in general are: 1. Large finished goods and component inventories.

2. Poor due date performance (30 to 40 percent of the orders early and 30 to 40 percent of the orders late).

3. Excessive fabrication lead times.

4. Unsatisfactory resource utilization in fabrication.

5. Fabrication and assembly act as separate unsynchronized plants.

DBR in T-Plants

In a T-plant, we have two situations. The most common situation is that most T-plants tend to belong to the MTS environment. Typically, buffer stocks are maintained at both the component level (just prior to assembly) and at the finished goods level. In this case, there are no real constraints (see discussion of MTS in Chapter 10) and the proper system to implement is the S-DBR system discussed in Chapter 9.

If this is not the case and there are capacity constraints, then the key factor is that the assembly operation must be managed properly. As long as stealing occurs at assembly, T-plants will be chaotic and flow will be difficult to manage. However, once stealing is eliminated through tight control of the assembly operations, then a T-plant becomes an A-plant and the DBR system discussion in A-plants should be followed.

The schedule control points are materials release, divergence, convergence, and the physical resource constraint (if one exists).

I-Plants

I-plants are the simplest of the production flows. The major issue with I-plants is the sharing of resources between the different products. Each product follows the same sequence of operations. There is little or no assembly and there are no divergence points.

The characteristics of an I-plant are: 1. All parts have similar routings.

2. Resources are shared between different parts, while raw materials are not.

3. There is very little assembly involved.

The typical I-plant product flow is shown in Fig. 8-18 and the shape makes the name obvious.

I-plants are the simplest of plants to manage. Nevertheless, traditional focus on resource utilization results in the use of production batches that are much larger than required to maintain a smooth flow. As a result, WIP piles can be created and the wave-like flow of A-plants can be observed. Consequently, I-plants have the following concerns/issues: 1. Low due date performance.

2. High WIP inventories.

3. Level of output below theoretical line rates.

FIGURE 8-18 Product flow diagram for a typical I-plant.

DBR in I-Plants

I-plants are straightforward to manage from a product flow standpoint. The DBR system as described in the previous sections can be designed and implemented with little complication. Identification of the constraint is simple-all personnel will be aware of this resource and the inventory buildup should confirm the section of this resource. Simple steps to improve productive use of this resource (see the section on Step 2-exploiting the constraint) should be followed by the implementation of the DBR system.

Most academic research has been conducted on I-plants (primarily on lines of 10 or less work centers) as indicated in Chapter 7. It is by far the simplest to simulate and study. In contrast, most plants are V, A, T, or combinations of these structures.

Summary

This chapter covered the basic terminology and concepts related to the TOC production solution. As such, it provides the foundation for a deeper understanding of DBR in an MTO environment, S-DBR in an MTA environment, and supply chains linking manufacturing to the downstream links. The various types of buffers are defined and illustrated, as are the various types of plants with their control points. A discussion of implementing DBR in each environment is provided.

References

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

Ford, H. 1928. Today and Tomorrow. Garden City, NY: Garden City Publishing, Goldratt, E. M. 1990a. The Haystack Syndrome: Sifting Information Out of the Data Ocean. Crotonon-Hudson, NY: North River Press.

Goldratt, E. M. 1990b. What's This Thing Called Theory of Constraints and How Should It Be Implemented? Croton-on-Hudson, NY: North River Press.

Goldratt, E. M. 2003. Production: The TOC Way. Rev. ed. Great Barrington, MA: North River Press.

Goldratt, E. M. 2009. "Standing on the shoulders of giants" The Manufacturer June. http://www.themanufacturer.com/uk/content/9280/Standing_on_the_shoulders_of_giants. (accessed February 4, 2010).

Goldratt, E. M. and Cox, J. 1984. The Goal: Excellence in Manufacturing. Croton-on-Hudson, NY: North River Press.

Schragenheim, E. and Dettmer, H. W. 2001. Manufacturing at Warp Speed. Boca Raton, FL: St. Lucie Press.

Senge, P. M. 1990. The Fifth Discipline: The Art and Practice of the Learning Organization. New York: Doubleday Currency.

Srikanth, M. and Umble, M. 1997. Synchronous Management: Profit-Based Manufacturing for the 21st Century. Vols. 1 and 2. Guilford, CT: Spectrum Publishing Company.

Sugimori, Y., Kusunoki, K., Cho, F., and, Uchikawa, S. 1977. "Toyota production system and Kanban system materialization of just-in-time and respect-for-human system." International Journal of Production Research, 15(6):553564.

Sullivan, T. T., Reid, R. A., and Cartier, B. 2007. TOCICO Dictionary. http://www.tocico.org/default.asp?page=dictionary.

About the Author.

Dr. Mokshagundam (Shri) L. Srikanth obtained his PhD in physics from Boston University. After a brief tenure as Associate Professor at Boston University, he joined Dr. Eli Goldratt in 1979. He is a partner in the Goldratt Group, an international organization headed by Dr. Eli Goldratt and dedicated to helping organizations and individuals achieve breakthrough improvements through the creation and dissemination of new knowledge. He is currently head of Goldratt Schools for North America.

He has nearly three decades of experience with industrial enterprises and ways to improve their performance. Dr. Srikanth was a Senior Director of the Center for e-Business Excellence at i2 Technologies. Prior to this position, he was a Director in i2's Product Management group. Before joining i2 Technologies, he was cofounder and managing principal of Spectrum Management Group.

Dr. Srikanth has helped companies improve delivery performance, reduce lead times, and reduce investment in inventories and resources. His experience covers a broad cross-section of industries including aerospace and defense, automotive, furniture, textiles, consumer, and industrial products. Companies range from Fortune 100 companies such as General Electric, Ford, General Motors, and United Technologies to small family-owned organizations.

Dr. Srikanth has authored several books including Regaining Competitiveness: Putting 'The Goal' to Work, with Harold E. A. Cavallaro, 2nd Revised Edition (North River Press, 1993); Synchronous Manufacturing: Principles for World Class Excellence, with Professor Michael Umble (Southwestern Publishing, 1991); Measurements for Effective Decision Making, with Scott A. Robertson, (Spectrum Publishing Company, 1995); and Synchronous Management: Principles for Profit-Based Manufacturing for the 21st Century, Vols. 1 and 2, with Professor Michael Umble (Spectrum, 1997). He is a contributor to Srinivasan, Mandyam, Streamlined-Principles for Building and Managing a Lean Supply Chain (Cengage Learning, 2004).

CHAPTER 9.

From DBR to Simplified-DBR for Make-to-Order

Eli Schragenheim

Introduction.

Drum-Buffer-Rope (DBR) is the name given by Dr. Eli Goldratt to a simple and effective production planning method. The root of the name is based on the analogy of the scouts tour described in The Goal (Goldratt and Cox, 1984, Chapters 1315). DBR was at the time the cornerstone of the Theory of Constraints (TOC) and continued to be the best known application of the theory until the appearance of Critical Chain (Goldratt, 1997), outlining the concepts for planning projects.

Simplified Drum-Buffer-Rope (S-DBR) is a variation on the original DBR methodology. It was suggested by Schragenheim and Dettmer (2000) in Manufacturing at Warp Speed as a valid, simplified replacement especially suited when the implementation has to use the common material requirements planning (MRP)/enterprise resource planning (ERP) software. Since then, the basic principles of S-DBR were adopted by Dr. Goldratt. Important improvements were added and dedicated software for S-DBR has been developed by Inherent Simplicity Ltd. under the close supervision of Dr. Goldratt. S-DBR has now replaced the older DBR as the preferred planning method with one exception, which will be explained later in this chapter.

Another important realization concerning production planning has to be mentioned. Both S-DBR and DBR were assuming a make-to-order (MTO) environment. During the rethinking of the TOC-focused planning methodology, it was recognized that the make-to-stock (MTS) production environment should be based on different principles. The author dedicates Chapter 10 to MTS, or rather to make-to-availability (MTA)1 environments, to emphasis the clear distinction.

Another comment should be made. While DBR and S-DBR are planning methods, they are not stand-alone methods. Buffer Management (BM), the TOC control mechanism, should be viewed as inseparable from the planning method. Thus, Chapters 9 and 10 deal with both the DBR/S-DBR planning as well as with BM as an absolutely necessary part of both planning methodologies.

The purpose of this chapter is to explain the S-DBR/BM concepts, logic, and procedures through the development of the ideas over time. Thus, the emphasis is on the historical development, which is critical to the full realization of the continual paradigm shift we have gone through during the last 25 years since the introduction of DBR.

A Historical Background and Perspective

In the mid-1980s, DBR represented a huge advancement in providing a robust plan for the production floor. DBR was developed as a major departure from the concept, created by its own developer, of very sophisticated and detailed planning of the shop floor. In the late 1970s through the first half of the 1980s, Dr. Eliyahu M. Goldratt led a software company, Creative Output Ltd., in developing a sophisticated program called OPT (Optimized Technology) to plan manufacturing orders in great detail for any kind of production shop floor. OPT was a true advanced planning and scheduling (APS) program even though the term was coined years later. At the time, the name given to such programs was "finite-capacity scheduling system" and that name hinted at a contrast with the MRP II programs of the time, which were known as "infinite-capacity scheduling systems."