The Scientific Approach of Quality Continual Improvement

1. THE SCIENTIFIC APPROACH

The scientific approach is one of the fundamental concepts that separates the total quality approach from other ways of doing business. Scholtes and his colleagues describe the sci­entific approach as “making decisions based on data, looking for root causes of problems, and seeking permanent solu­tions instead of relying on quick fixes”7

For putting the scientific approach to work in a total quality setting, these four strategies are helpful: (1) collect meaningful data, (2) identify root causes, (3) develop appro­priate solutions, (4) plan and make changes.8

1.1. Collect Meaningful Data

Meaningful data are free from errors of measurement or pro­cedure, and they have direct application to the issue in ques- tion.9 It is not uncommon for an organization or a unit within it to collect meaningless data or to make a procedural error that results in the collection of erroneous data. In fact, in the age of computers, this is quite common. Decisions based on meaningless or erroneous data are bound to lead to failure. Before collecting data, decide exactly what data are needed, how they can best be collected, where the data exist, how they will be measured, and how you will know the data are accurate.

1.2. Identify Root Causes of Problems

The strategy of identifying root causes is emphasized through­out this book.10 Too many resources are wasted by organiza­tions attempting to solve symptoms rather than problems. The total quality tools are helpful in separating problems from symptoms.

1.3. Develop Appropriate Solutions

With the scientific approach, solutions are not assumed.11 Collect the relevant data, make sure they are accurate, identify root causes, and then develop a solution that is ap­propriate. Too many teams and too many people begin with “I know what the problem is. All we have to do to solve it is…” When the scientific approach is applied, the prob­lem identified is often much different from what would have been suspected if acting on a hunch or an intuition. Correspondingly, the solution is also different.

1.4. Plan and Make Changes

Too many decision makers use what is sometimes called the “Ready, fire, aim” approach rather than engaging in careful, deliberate planning.12 Planning forces you to look ahead, an­ticipate needs and what resources will be available to satisfy them, and anticipate problems and consider how they should be handled.

Much of the scientific approach has to do with establish­ing reliable performance indicators and using them to mea­sure actual performance. In his book Total Manufacturing Management, Giorgio Merli lists the following examples of useful performance indicators:13

  • Number of errors or defects
  • Number of or level of need for repetitions of work tasks
  • Efficiency indicators (units per hour, items per person)
  • Number of delays
  • Duration of a given procedure or activity
  • Response time or cycle
  • Useability/cost ratio
  • Amount of overtime required
  • Changes in workload
  • Vulnerability of the system
  • Level of criticalness
  • Level of standardization
  • Number of unfinished documents

This is not a complete list. Many other indicators could be added. Those actually used vary widely from organiza­tion to organization. However, such indicators, regardless of which ones are actually used, are an important aspect of the scientific approach.

2. IDENTIFICATION OF IMPROVEMENT NEEDS

Even the most competitive, successful organizations have limited resources. Therefore, it is important to optimize those resources and use them in ways that will yield the most benefit. One of the ways to do this is to carefully se­lect the areas of improvement to which time, energy, and other resources will be devoted. If there are 10 processes that might be improved, which will yield the most benefit if improved? These are the processes that should be worked on first.

Methodologies for identification of improvement needs were discussed in Chapter 16. Another approach is offered by Scholtes and his colleagues. They recommend the follow­ing four strategies for identifying improvement needs:14

  • Apply multivoting. Multivoting involves using brain­storming to develop a list of potential improvement projects. Team members vote several times—hence the name—to decide which project or projects to work on first. Suppose the original list contains 15 potential proj­ects. Team members vote and cut the list to 10. They vote again and cut it to 5. The next vote cuts the list to 3, and so on, until only 1 or 2 projects remain. These are the first projects that will be undertaken.
  • Identify customer needs. An excellent way to identify an improvement project is to give the customer a voice in the process. Identify pressing customer needs and use them as projects for improvement.
  • Study the use of time. A good way to identify an improve­ment project is to study how employees spend their time. Is an excessive amount of time devoted to a given process, problem, or work situation? This could signal a trouble spot. If so, study it carefully to determine the root causes.
  • Localize problems. Localizing a problem means pin­pointing specifically where, when, and how often it hap­pens. It is important to localize a problem before trying to solve it. Problems tend to be like roof leaks in that they often show up at a location far removed from the source.

3. DEVELOPMENT OF IMPROVEMENT PLANS

After a project has been selected, a project improvement team is established. The team should consist of representa­tives from the units most closely associated with the prob­lem in question, including the process operator. It must in­clude a representative from every unit that will have to be involved in carrying out improvement strategies. The project improvement team should begin by developing an improve­ment plan. This is to make sure the team does not take the “Ready, fire, aim” approach mentioned earlier.

The first step is to develop a mission statement for the team. This statement should clearly define the team’s pur­pose and should be approved by the organization’s govern­ing board for quality (executive steering committee, quality council, or whatever the group is called). After this has been accomplished, the plan can be developed. Scholtes and his colleagues recommend five stages for developing the plan:

  1. Understand the process. Before attempting to im­prove a process, make sure every team member thor­oughly understands it. How does it work? (This usually requires the development of a process flowchart: see Chapter 15.) What is it supposed to do? Why is that step necessary? What are the best practices known pertain­ing to the process? The team should ask these questions and others, and pursue the answers together. This will give all team members a common understanding, elimi­nate ambiguity and inconsistencies, and shine light on any obvious problems that must be dealt with before proceeding to the next stage of planning.
  2. Eliminate errors. In analyzing the process, the team may identify obvious errors, or potential errors, that can be quickly eliminated. These should be eradicated before proceeding to the next stage. This stage is some­times referred to as “error-proofing” the process.
  3. Remove slack. This stage involves analyzing all of the steps in the process to determine whether they serve any purpose and, if so, what purpose they serve. In any organization, processes exist that have grown over the years with people continuing to follow them without giving any thought to why things are done a certain way, whether they could be done better another way, or whether they need to be done at all. Few processes can­not be streamlined.
  4. Reduce variation. Variation in a process results from either common causes or special causes. Common causes result in slight variations and are always present. Special causes typically result in greater variations in performance and may not always be present. Strategies for identifying and eliminating sources of variation are discussed in the next section. See also Chapter 18.
  5. Plan for continual improvement. By the time this step has been reached, the process in question should be in good shape. The key now is to incorporate the types of improvements made on a continuous basis so that continual improvement becomes a normal part of doing business. The Plan–Do–Check–Adjust cycle, discussed in Chapter 16, applies here. With this cycle, each time a problem or potential improvement is identified, an improvement plan is developed (Plan), implemented (Do), monitored (Check), and refined as needed (Adjust).

4. COMMON IMPROVEMENT STRATEGIES

Numerous different processes are used in business and in­dustry; consequently, there is no single road map to follow when improving processes. However, a number of standard strategies can be used as a menu from which improvement strategies can be selected as appropriate. Figure 19.2 shows several standard strategies that can be used to improve pro­cesses on a continual basis: (1) describe, (2) standardize, (3) eliminate, (4) streamline, (5) reduce, (6) bring under sta­tistical control, and (7) improve.15

The strategy of describing the process is used to make sure that everyone involved in improving a process has a de­tailed knowledge of the process. Usually, this requires some investigation and study. The steps involved are as follows:

  1. Establish boundaries for the process.
  2. Flowchart the process (as it is, not as it should be).
  3. Make a diagram of how the work flows.
  4. Verify your work.
  5. Correct immediately any obvious problems identified.

To continually improve a process, all people involved in its operation must be using the same procedures. Often this is not the case. Employee X may use different procedures than Employee Y. It is important to ensure that all employees are using the best, most effective, most efficient procedures known. The steps involved in standardizing a process are as follows:

  1. Identify the currently known best practices for the process and write them down.
  2. Test the best practices to determine whether they are in fact the best, and improve them if there is room for improvement these improved practices then become the best practices that are recorded).
  3. Make sure that everyone is using the newly standardized process.
  4. Keep records of process performance, update them continually, and use them to identify ways to improve

The strategy of eliminating errors in the process involves identifying errors that are commonly made or which could be made (potential) in the operation of the process and then getting rid of them. This strategy helps delete steps, proce­dures, and practices that are being done a certain way simply because that is the way they have always been addressed; and those that could be done incorrectly due to ambiguous or in­complete process procedures; or even faulty process design. Whatever measures can be taken to eliminate such errors are carried out as a part of this strategy.

The strategy of streamlining the process is used to take the slack out of the process. This can be done by reducing inventory, reducing cycle times, and eliminating unnecessary steps. After a process has been streamlined, every step in it has significance, contributes to the desired end, and adds value.

4.1. Reduce Sources of Variation

The first step in the strategy of reducing sources of variation is identifying sources of variation. Such sources can often be traced to differences among people, machines, measurement instruments, material, sources of material, operating condi­tions, and times of day. Differences among people can be at­tributed to levels of capability, training, education, experi­ence, and motivation. Differences among machines can be attributed to age, design, and maintenance. Regardless of the source of variation, after a source has been identified, this information should be used to reduce the amount of varia­tion to the absolute minimum. For example, if the source of variation is a difference in the levels of training completed by various operators, those who need more training should re­ceive it. If one set of measurement instruments is not as finely calibrated as another, they should be equally calibrated.

The strategy of bringing the process under statisti­cal control was explained in detail in Chapter 18. For this discussion, it is necessary to know only that a control chart is planned, data are collected and charted, special causes are eliminated, and a plan for continual improvement is developed.

There are many different ways to design and lay out a process. Most designs can be improved on. The best way to improve the design of a process is through an active pro­gram of experimentation. To produce the best results, an experiment must be properly designed, using the following steps:

  1. Define the objectives of the experiment. (What factors do you want to improve? What specifically do you want to learn from the experiment?)
  2. Decide which factors are going to be measured (cycle time, yield, finish, or something else).
  3. Design an experiment that will measure the critical fac­tors and answer the relevant questions.
  4. Set up the experiment.
  5. Conduct the experiment.
  6. Analyze the results.
  7. Act on the results.

4.2. Additional Improvement Strategies

In his book Total Manufacturing Management, Merli lists 20 strategies for continual improvement that he calls “The Twenty Organizing Points of Total Manufacturing Management.”16 Eighteen of these strategies are still valid and are explained in the following paragraphs:17

  • Reduced lead time. Raw materials sitting in a store­room are not adding value to a product. Efficient man­agement of the flow of materials is essential to competi­tiveness. Lead time can be reduced by evaluating the following factors: order processing time, waiting time prior to production, manufacturing lead time, storage time, and shipping time.
  • Flow production. Traditionally, production has been a stop-and-go or hurry-up-and-wait enterprise. Flow production means production that runs smoothly and steadily without interruption. An example illustrates this point. A large manufacturer of metal containers had its shop floor arranged by type of machine (cut­ting, turning, milling, etc.). All cutting machines were grouped together, all turning machines were grouped to­gether, and all milling machines were grouped together. However, this isn’t how the flow of work went. Work flowed from cutting to turning, back to cutting, and on to milling. Arranging machines by type caused a great many interruptions and unnecessary material handling. To improve production efficiency, the machines were re­arranged according to work flow. This is often referred to as cellular production. Flow production smoothed out the rough spots and made work flow more smoothly.
  • Group technology. Traditional production lines are straight. With group technology, processes are arranged so that work flows in a U-shaped configuration. This can yield the following benefits: shorter lead times, greater flexibility, less time in material handling, minimum work in progress, flexibility with regard to volume, less floor space used, and less need for direct coordination.
  • Level production. This involves breaking large lots into smaller lots and producing them on a constant basis over a given period of time. For example, rather than produc­ing 60 units per month in one large lot, production might be leveled to produce 3 units per day (based on 20 work­days per month). This strategy can yield the added ben­efit of eliminating the need to store the materials needed for large lots. This, in turn, makes it easier to implement just-in-time manufacturing.
  • Synchronized production. Synchronized production involves synchronizing the needs of the production line with suppliers of the materials needed on the line. For example, assume that a line produces computers in a va­riety of different internal configurations. The difference among the configurations is in the capacity of the hard drive installed. Such information as what type of hard drive is needed, in what quantities, at what time, and at what point on the line must be communicated to the hard-drive supplier. The supplier must, in turn, deliver the correct type of hard drive in the correct quantity at the correct time to the correct place on the line. When this happens, synchronized production results.
  • Overlapped/parallel production. This strategy involves dismantling long production lines with large lot capaci­ties and replacing them with production cells that turn out smaller lots. This allows production of different con­figurations of the same product to be overlapped or run parallel, that is, concurrently.
  • Flexible schedules. Production cells and the ability to overlap production or run it parallel allow for a great deal of flexibility in scheduling. The more options available to production schedulers, the more flexible they can be in developing schedules.
  • Pull control. Pull control is a concept applied to elimi­nate idle time between scheduling points in a production process, the need to maintain oversized inventories to off­set operational imbalances, and the need to plan all tar­get points within a process. With good pull control, work moves through a process uninterrupted by long waiting periods between steps.
  • Visual control. Visual control is an important aspect of just-in-time manufacturing. It is an information dis­semination system that allows abnormalities in a pro­cess to be identified visually as they occur. This, in turn, allows problems to be solved as they occur rather than after the fact.
  • Stockless production. Stockless production is an approach to work handling, inventory, lead time plan­ning, process balancing, capacity utilization, and schedule cycling that cuts down on work in progress. With stockless production, it is necessary to eliminate process bottlenecks, balance the process, and have an even work flow that eliminates or at least minimizes work in progress. Stockless production and just-in-time go hand-in-hand.
  • Jidoka. Jidoka means halting an entire process when a defect is discovered so that it won’t cause additional prob­lems further down the line. Jidoka can be accomplished manually, or the line can be programmed to stop auto­matically, or both.
  • Reduced setup time. This strategy consists of any activ­ity that can reduce the amount of time required to break down a process and set it up again for a different produc­tion run. Such things as quick changeovers of tools and dies are common with this strategy.
  • Control of work-in-process. Work-in-process (WIP) often means work that is sitting idle waiting to be processed. Controlling the amount of idle WIP in­volves organizing for a smoother flow, small lot sizes, process flexibility, pull control, and rapid breakdown and setup.
  • Quality improvement. In addition to improving pro­ductivity using the various strategies discussed in this chapter, it is important simultaneously to improve quality. This book is devoted to an approach for con­tinually improving quality. An important point is that productivity and quality improvements are reciprocally supportive.
  • Total cost cycles. This strategy involves basing deci­sions on the total cost cycle rather than isolated pieces of it. It is not uncommon for decisions to be based on reducing the costs associated with part of a process, although another part of the process may have its cost increased by the decision. True improvements have not been accomplished unless overall costs have been reduced.
  • Cost curves. A cost curve is a graphic representation of a time-based process wherein manufacturing costs accu­mulate relative to billing. Two types of costs are shown on a cost curve: materials and conversion costs. A cost curve shows graphically how much cost accumulates until the customer is billed for the product. It is a tool to help man­agers set the optimal point of production.
  • Supplier partners. This strategy amounts to involving suppliers as partners in all phases of product develop­ment rather than keeping them in the dark and revealing your activities to them only through the low-bid process. If tested and trusted suppliers know what you are trying to do, they will be better able to maximize their resources in helping you do it.
  • Total productive maintenance. Total productive maintenance (TPM) means maintaining all systems and equipment continually and promptly all of the time. In a rushed workplace, one of the most common occur­rences is slacking off on machine and system mainte­nance. This is unfortunate because a poorly maintained system cannot achieve the quality and productivity needed to be competitive. Poor maintenance can result in the following problems: shutdowns from unexpected damage, increased setup and adjustment time, unused uptime, speeds below the optimum, increased varia­tions, increased waste from defects, and production losses.

Source: Goetsch David L., Davis Stanley B. (2016), Quality Management for organizational excellence introduction to total Quality, Pearson; 8th edition.

Leave a Reply

Your email address will not be published. Required fields are marked *