Applying Six Sigma Methods to Materials Management
Dan O'Leary, CQA, CQE, CQMgr, CRE
Dan O'Leary, CQA, CQE, CQMgr, CRE, Business Process Executive, Picker International, Inc. (www.picker.com) Cleveland, OH 44143, 440-473-5776, firstname.lastname@example.org
85th Annual International Conference Proceedings - 2000
Abstract. This paper characterizes Six Sigma as a breakthrough technique in contrast with restorative and continuous improvement techniques. Materials Management, the movement and control of goods into, inside, and from a company, is a functional area in comparison with a department on an organization chart. Understanding the functions, processes, and procedures of Materials Management sets the stage for measurement and consequent improvement. The application of Six Sigma to Materials Management requires projects selected to support the company's strategy. The project improvement teams apply statistical methods to understand a process, set the process parameters for improved results, and institute methods to hold the gains achieved.
Characterizing Six Sigma. Six Sigma is a "new kid on the block" as an improvement program, but the methods and techniques find their roots in the statistical process control approaches of Shewhart, Deming, and Juran plus the experimental design approaches of Box, J. Hunter, and W. Hunter. Motorola initiated Six Sigma and played a pivotal role in their winning the Malcolm Baldrige National Quality Award. The current resurgence of Six Sigma originated with Allied Signal and the subsequent adoption by General Electric. Six Sigma programs are becoming common and, as a result, have many different definitions. For our purposes, Six Sigma is, "the use of statistical methods to achieve breakthrough performance."
Some current Six Sigma debate focuses around the question of its status in a company's business improvement strategy. The debate has Six Sigma ranging from the overarching strategy to an element in a tool box. Part of the answer, of course, depends on the maturity of the company, its experience with business improvement activities, and the success of those activities. Fortunately, we do not have to settle the debate to gain insight and value from Six Sigma.
To understand Six Sigma, we break the characterization down into two parts:"what we want to achieve" and "how we intend to achieve it." Below, we examine each part and analyze the Six Sigma components.
In a Six Sigma program we want to achieve breakthrough performance. Breakthrough is one kind of quality technique in contrast with restorative quality methods. Each is the solution to a particular kind of quality problem. A sporadic problem is a sudden and adverse change to the status quo and requires that the system return to its former state (the status quo ante). Restorative quality methods resolve sporadic problems. A chronic problem is a long standing adverse situation; unfortunately, a chronic problem easily becomes the status quo. The solution is to move the system to a new state, to a new status quo. See Juran and Gryna, Chapter 5, for a discussion of this distinction.
Assume that in company X the normal cycle time from receipt of a shipment at the receiving dock to transaction entry in the MRP system is six hours. If it jumps to ten hours, then restorative quality methods can determine the root cause and return it to six hours. However, if management determines that six hours is too long, then breakthrough methods can reduce the time, perhaps achieving ½ hour.
We should also contrast breakthrough improvement with continuous improvement. In continuous improvement, we make changes in small bits instead of the big bites taken in breakthrough improvement. Continuous improvement looks for incremental changes -- small changes that accumulate to large impact over time. Breakthrough improvement looks to large changes followed by a period of maintenance. Continuous improvement is often described using the Japanese word kaizen. In Imai, pg. 23 the difference as "two contrasting approaches to progress: the gradualist approach and the great-leap-forward approach. Japanese companies generally favor the gradualist approach and Western companies the great-leap approach." Six Sigma clearly falls into the great-leap approach and is a part of the occidental tradition.
In a Six Sigma program, one achieves breakthrough performance using statistical techniques. Statistical techniques apply a standard set of statistical methods to measure performance, collect data, analyze processes, propose alternatives, and standardize the results. Statistical techniques use measurements, data, and analysis instead of opinion and conjecture.
The typical approach to using statistical techniques lies in modeling a real problem as a statistical problem. For example, we can ask, "How can we improve the cycle time from receipt of a shipment at the receiving dock to transaction entry in the MRP system?" We may convert the real world problem into a statistical problem that looks like
Cycle Time = Y = f(x1, x2, . . , xn).
We solve the statistical problem by identifying the factors, the xi, that control cycle time and tuning the values that optimize it. We bring the statistical solution back to the real world by first adjusting the process settings and then setting up controls (often Statistical Process Control) to hold the new settings.
The common approach for Six Sigma uses experts working on specific problems. The experts are often selected and trained specifically for Six Sigma assignments and called "black belts", "process improvement masters", etc. Projects are typically selected to improve one of three metrics: first pass yield, capacity, and the cost of poor quality.
The first pass yield of a process is the product of the first pass yield at each process step. If a process has three steps, each with a first pass yield of 90%, the process first pass yield is 0.9 0.9 0.9=0.73 or 73%. Only 73% of the product gets all the way through the process without being rejected. The process yield is a percent or a ratio.
Capacity is the amount of product that can flow through the process in a given time. Each process step has a capacity and the process step with the lowest capacity sets the overall capacity. The lowest capacity step is the process bottleneck. Capacity is the maximum product per unit of time. In a warehouse it might be the number of parts that can be picked in a day.
Cost of poor quality is the cost of not producing correct parts and includes rework, sorting, scrap, etc. Often expressed in dollars, such as the value of scrap, it is better normalized as a ratio or percentage. For example, we might figure out the percentage of incorrect parts that must be thrown away for the number of parts received.
Six Sigma projects are selected to improve at least one of these three metrics. For example, if a Six Sigma project can improve the capacity of a process step by 15% then the whole factory capacity may be increased. Naturally, the process step selected for capacity improvement must be a bottleneck. A Six Sigma project may increase capacity with almost no capital expenditure. Using the statistical techniques, the "process improvement master" determines the factors that affect capacity and changes the process settings for improved results.
Characterizing Materials Management. Unlike Six Sigma, Materials Management has been on the scene for some time. Unfortunately, however, Materials Management still does not have a clear and generally accepted definition. The APICS dictionary defines Materials Management as, "The grouping of management functions supporting the complete cycle of material flow, from the purchase and internal control of production materials to the planning and control of work in process to the warehousing, shipping, and distribution of the finished product."
A complementary view of Materials Management is offered in Tersine, pg. xiii.
The management of materials concerns their flow to, within, and from the [company]. The efficiency and efficacy of the flow can substantially influence costs and revenue generation and thus hold serious implications for marketing, finance, and production. Materials management seeks a balance between shortages and excesses in an uncertain environment. As it does, marketing is influenced through revenue and customer relations, production through efficiency and cost of operations, and finance through liquidity and operating capital.
In both views of Materials Management we encounter the idea of a business function. In the former view functions are explicit while the latter includes examples such as marketing, finance, and production. We need to clarify the distinction between function and organization. We turn to Dobler, et al. pg. 10 where, in discussing purchasing they say, "There is a fundamental distinction, however, between the purchasing function and the purchasing department. They are not necessarily the same. As a function, purchasing is common to all types of business enterprise. The purchasing department, however, is an organizational unit of a firm whose duties may include part or all of the purchasing function. This distinction between function and department is not always appreciated or understood by top management."
The distinction drawn by Dobler, et al. extends beyond purchasing to all the functions in a modern business. It is the difference between the work that must be performed and the organizational structure created to accomplish that work. This distinction is drawn more sharply by Martin in making the contrast between functions, processes, and procedures. Martin offers the following definitions. Business Function: A group of business activities that together completely support one aspect of furthering the mission of the business. Business Process: A task or group of tasks carried out as a business function. Procedure: A method by which one or more processes may be carried out.
While the formal definitions are useful a little more explanation sheds light on the intention. A business function is concerned with what has to be done to operate the business. It does not include how it is done nor what part of the organization is responsible for doing it. A function is ongoing and continuous. A business process is also concerned with what has to be done and, like the business function, does not include how it is done nor the organizational component responsible. A process is executed repeatedly, has a beginning and an end, and is described in terms of inputs and outputs. A procedure is concerned with how something is accomplished and typically changes as the technology changes. Often there are several procedures that could be used. While procedures may change, the functions and processes must continue to run the enterprise.
Functions may be grouped into functional areas that combine functions. For our purposes, Materials Management (as well as marketing, finance, and production) is a functional area composed of functions, processes, and procedures. Since functions (and functional areas) are independent of the organizational structure, we do not need to know the organization chart to effectively discuss Materials Management. Adapting Dobler, et al., we characterize Materials Management (a functional area) as an integrated approach to planning, acquisition, conversion, flow, and distribution of materials. The functions associated with Materials Management, consequently, include procurement, inventory management, receiving, stores and warehousing, materials handling, production planning, traffic and transportation, and surplus and salvage.
Applying Six Sigma to Materials Management. Having characterized both Six Sigma and Materials Management, we now consider the issues of applying Six Sigma. The first step is to develop the metrics that apply to Materials Management's functions. We described the typical Six Sigma metrics in three broad areas: yield, cost of poor quality, and capacity.
The use of Six Sigma is best started with the collection and management of both data and information. As part of the company's total improvement program, there should be a strong emphasis on measurement. Materials Management, along with the other functional areas in the company, needs a well established set of metrics, i.e., defined, reported frequently, reported consistently, and meaningful to both workers and management. The metrics most valuable for Six Sigma are not highly aggregated, but are very close to the functions and processes they represent.
Moreover, the metrics selected should include more than just cost; a complete picture may also require measurements of variability, cycle time, timeliness, and defects as well as the metrics discussed above (cost of poor quality, yield, and capacity). While metrics at a point in the process are important, one should not overlook metrics that take into account the whole process. We saw an example above of the difference between first pass yield at a process step and first pass yield for a complete process. Additionally, the metrics should roll up across the plant, the organization, and the functions. Tracking the metrics affords a longer term view over time (months, quarters, and years).
Consider the advice in Cartin, pg. 52, "Process performance (vitality) must be measured before intelligent decisions can be made toward improving it. The measurement's dimensions should be the important indicator for organizational success and meaningful process improvement.
Measuring cost alone is often not a useful index. Costs are the result of the variation in other factors. If the TQM emphasis on process improvement (variation reduction) is to be maintained, important characteristics, such as cycle time quality, and productivity should be measured. Every process can be measured in terms of its output and the resources used. It is also important that measurements be tracked over time so that characteristics such as trend and capability can be determined."
Having established process and function measurements, management must select projects for improvement. The typical selection criteria involve not only poorly performing processes, but also processes that directly support the company's strategic plan. For example, if a company wants to increase its capacity, then Six Sigma projects aimed at capacity or yield are good choices. A company that competes in markets with short customer response times may want to select projects that improve cycle time or yield. Companies that compete on cost may select projects to improve yield or the cost of poor quality.
Having selected a project, the company must have two essential elements in place. The first is management commitment, while the second is the skill set to understand and change the process. Each of these is essential to success in any improvement effort.
Process improvement requires commitment from some member of the company's management. We made a distinction between functions and organization, recognizing that Materials
Management (a functional area) may have functions distributed across the organization. Each function should have an owner -- the person in management who is responsible for the proper operation of a function, its processes, and its procedures. In addition, each function has stakeholders who have an interest in how the function performs. As pointed out in Tersine, pg. xiii, Materials
Management's stakeholders include marketing, finance, and production. The company must name the owner and the stakeholders for each function. If the function's owner is not committed to the improvement effort, i.e., to making major changes in the function, processes, and procedures, there is little likelihood of success.
The improvement project selected must have an improvement goal based on the measurements. For a breakthrough the goal should be an improvement of at least two times. Less ambitious goals often lead to lack of success and subsequent rationalization. An improvement of two times in yield might, for example, move a total process yield from 90% (10% rejects) to 95% (5% rejects). Similarly, a cycle time improvement project might move the total process cycle time from 20 days to 10 days.
Six Sigma improvement projects are often based on a core of experts called "black belts", "process improvement masters", etc. Each project, while team based, has one person responsible for the project and achieving the improvement goal. The team's expert must have the necessary knowledge to lead the team. Since the improvement project is based on the application of statistical tools, data analysis, etc. the team's expert must have the requisite skills and practical experience.
See Table 1.
The statistical knowledge was often based on a high degree of education and required specialized knowledge of statistical tools. Today, an adequate working knowledge of the statistical methods is accessible in about four weeks of intensive training. The tools, which often rely on calculations based on data sets, are accessible in standard, and readily accessible software, such as Excel, QuattroPro, and Lotus.
A typical project uses several steps, but all are based on the Shewhart cycle of Plan-Do-Check-Act. In the Six Sigma context, projects are often laid out as Measure-Analyze-Improve-Control. Measurement determines how processes are operating and provides the raw material to determine which processes to improve. Analyze uses the data (and information) produced by the process to determine the improvement opportunity, i.e., what process parameters control the process and how they should be set. Improve implements the knowledge from the analysis stage to reset the parameters. Control establishes the methods for the process operators to monitor the process, preserve the gains established, and continuously improve the process.
|Table 1 Some Statistical Tools for Six Sigma|
|Process mapping||Cause and effect diagrams (Isakawa)|
|Measurement analysis||Failure modes, effects, and criticality analysis|
|Graphical techniques||Statistical process control charts|
|Box and whisker plots||x-bar & R chart|
|Histogram||;x & MR chart|
|Scatter diagrams||p chart|
|Pareto charts||np chart|
|Check series||u chart|
|Design of experiments||Statistical distributions|
Conclusion. Six Sigma techniques, often applied to manufacturing can lead to good results in Materials Management. The company should define Materials Management as a functional area, and identify the associated functions. Each function should have an owner, a well developed set of metrics, and a clear understanding of the stakeholders needs.
Materials Management improvement projects must be selected in line with the company's strategic goals and the markets in which it operates. The goals must be aggressive enough to provide real breakthrough; less aggressive goals are more suited to continuous improvement.
The improvement team must a have a solid grasp of statistical techniques and enough practical sense to apply them well. When the team is complete, the process parameters are set to new values and the process operators have the tools to ensure the achieved gains are held for the long term.
APICS Dictionary, 9th ed. APICS -- The Educational Society for Resource Management, 1998.
ISBN 1-55822-162-X. Cartin, T.J. Principles and Practice of TQM. ASQC Quality Press, 1993. ISBN 0-87389-153-8.
Dobler, D.W., D.N. Burt, and L. Lee. Purchasing and Materials Management: Text and Cases, 5th ed. McGraw-Hill Inc., 1990. ISBN 0-07-037047-8.
Imai, Massaki. Kaizen: The Key to Japan's Competitive Success. McGraw-Hill Publishing Company, 1986. ISBN 0-07-554332-X.
Juran, J.M. and Frank M. Gryna. Quality Planning and Analysis: From Product Development Through Use, 2nd edition. McGraw-Hill Book Company, 1970. ISBN 0-07-033178-2.
Martin, J. Information Engineering, Book II: Planning and Analysis. Prentice-Hall, Inc. , 1990. ISBN 0-13-464885-4.
Tersine, R.J. Principles of Inventory and Materials Management, 4th edition. Prentice-Hall, Inc. 1994. ISBN 0-13-457888-0.