Six Sigma in research and development firms

MANAGERS AT WORK

Albert Johnson

LESSONS LEARNQI FROM SIX SIGMA IN R&O "Eliminating defects" is difficult in R&D because the work defies systematic improvement. In the workshops of the Industrial Research Institute's Six Sigma in R&D Project, dozens of organizations have shown how Six Sigma concepts are taught to technical staff, how to involve R&D in corporate initiatives, and how to avoid spending a fortune on relatively little-used tools.

This article conveys insights shared and extracted from the IRI workshops on both Six Sigma and Design for Six Sigma (DFSS) in R&D. With appropriate corporate strategy. Six Sigma and DFSS in R&D help to generate superior products and competitive profitability.

Defining Terms

As popularly used in industry. Six Sigma has come to mean a management strategy for quality and perfor- mance improvement that uses a metric of 3.4 defects per million opportunities as an anchor or guideline for per- formance. This metric provides a clear, focused and easily-understood standard for eliminating defects from work activities and communicating improved results.

Random, seasonal and biased processes in R&D organi- zations are difficult to associate with concepts of statis-

Albert Johnson is a senior analyst in the Science & Tech- nology division. Corning Incorporated, Corning, New York. He manages research contracts, consortia and other affiliations for Coming's research organization, conducts and coordinates Six Sigma projects, and par- ticipates in business analysis for strategic planning. Johnson has an M.S. in industrial administration and a B.S. in management science from Carnegie Mellon Uni- versity, a certificate in management of research, devel- opment and technology-based innovation from the Massachusetts Institute of Technology, and has completed the Advanced Licensing Institute of the Franklin Pierce Law Center. He is on the Board of Directors of the Industrial Research Institute, and is past chairman of its Research-on-Research Committee. [email protected]

I March—April 2006 O895-63O8/O6/$5.00 © 2006 Industrial Research Institute, Inc.

tical control. But not all processes in R&D organizations are random or biased—some are repeatable. Defect elimination focuses in part on the repeatable work in an R&D organization, and is one area of impact for quality improvement. Furthermore, some R&D work is focused on outcomes related to Six Sigma, such as inventing and designing products and processes that will perform at or better than Six Sigma levels.

Thus, in an R&D context. Six Sigma represents a mindset that is a consequence of adopting 3.4-in-a- million as the guiding performance standard for the repeatable work. Building on the strategic and opera- tional meanings of Six Sigma, Design for Six Sigma (DFSS) means to design products so that they are manu- factured, and services so that they are measurably rendered, with defect rates at or below Six Sigma.

Insights from the Workshops Companies have formally implemented Six Sigma for various reasons. These reasons, predictably, focus on decreasing cost, increasing speed to market, and improving product and process quality—quality as defined by customers. Because of the focus and sponsor- ship that these initiatives require, most use a somewhat similar top-down approach for implementing company- wide Six Sigma programs.

A major component of this approach involves training a relatively large number of employees in methodically defining problems and applying quality tools. This methodical approach is the essence of the cultural shift for most organizations, which is why many people describe Six Sigma as a management approach, or paradigm for thinking about product and service quaiity. Further, the required methodology also constrains the implementation in a way that makes the approaches used in different organizations similar in methods, tools and tactics.

Because Six Sigma implementations involve resolving persistent management issues—again, issues involving cost avoidance, improving revenue and customer- defined product quality—the problems it addresses and its problem-solving tools are similar to those of other

recent quality improvement strategies, such as TQM. In our Six Sigma workshops, R&D professionals have talked about solving problems in meeting customer requirements, achieving new product development goals, as well as optimization of administrative, technical, production, maintenance, and managerial systems.

Because the problems they are solving are usually pervasive, expensive and persistent, most organizations rely on extensive implementations of Six Sigma tech- niques, led from the most senior level and involving rela- tively large numbers of their professionals in order to communicate and support the necessary cultural and behavioral shifts. These company-wide implementations can be a sore point for R&D professionals, in part because they are already quite familiar with those methods, tools and tactics in a different context. Conse- quently, researchers who are "forced" into a Six Sigma implementation may openly wonder about the rationale for all the fuss over something so obvious. To many R&D professionals. Six Sigma seems analogous to "old wine in new bottles," in part because it extensively re-uses quality improvement knowledge that was popu- larized during the TQM days.

Linking the R&D organization to the corporate Six Sigma program is neither as easy nor as difficult as it might seem at first. It is not as easy because although there are broad limits to Six Sigma's usefulness in R&D, if senior corporate leadership fails to engage research leadership in defining context, then those limits are quickly reached. It is not as difficult as it may seem at first because the context for Six Sigma in R&D can be easily drawn by focusing on problem definition and problem solving.

Formalisms and Scientific Method

To apply Six Sigma, problems have to be stated formally—an approach that researchers are quite familiar with. In the case of Six Sigma, this means a complete problem statement defines "x" as a cause, input-and-process, or problem, and " / ' as a symptom, output or effect. With some care, then, formal statements of 3̂ = f(;c) are made to complete the problem definition.

Formalisms used in the Six Sigma approach to organiza- tional problem solving strongly resemble those used in the scientific method. Consequently, the similarity between scientific method and the fundamental Six Sigma approach is not only very important to R&D pro- fessionals but is also easily explained. Scientific meth- od is:

1. Observing one or more phenomena;

2. Forming an explanatory hypothesis;

3. Using the hypothesis to predict other phenomena, or fiiture observations;

4. Performing independent, controlled experiments to test the predictions.

Six Sigma's fundamental method, DMAIC (Define, Measure, Analyze, Improve, Control), is:

1. Defining the problem(s) in an existing process;

2. Measuring the process' outcomes;

3. Analyzing the process, including formulating and running experiments to test claims about it;

4. Making improvements based on the analysis;

5. Implementing systems and metrics to sustain and monitor the improvements.

Both scientific method and Six Sigma involve measure- ment, hypothesis formation and testing, analysis, synthesis, and evaluation—putting R&D professionals into a situation in which they are essentially learning somewhat different formalisms and vocabulary to augment an already-familiar approach to problem iden- tification and solution.

Broad Applicability

What all this means is that once R&D professionals grasp the analogy between Six Sigma and scientific method, they are apt to find it has broad applicability in R&D because R&D work is fundamentally a series of problem-defining and problem-solving processes. Because formalizing problem statements is routine for most R&D professionals. Six Sigma approaches can easily be used to improve the repeatable portions of the R&D organization's activities.

In the workshops, we learned of many successful projects. R&D leaders from a couple of companies have reported on building DFSS requirements into their standards for product development, for instance; others

Q Research • Technology Management

have reported successful projects in laboratory automa- tion, as well as successfiil processes for the "fuzzy" early stages of research and innovation.

A major area of DFSS impact in research is designing and developing manufacturing processes that operate within Six Sigma tolerances. This positions R&D orga- nizations as key drivers of long-range, sustainable, top- line growth via customer satisfaction through DFSS, in contrast to the relatively short-term, cost-reduction, pro- duction, and administration emphasis of Six Sigma's fundamental DMAIC method.

In general, DFSS replaces the "find and fix" steps of Six Sigma with process development, implementation and verification activities. DFSS is generally used to develop new products and processes, or in situations in which existing products or processes cannot be improved enough with DMAIC methods to reach operational or competitive goals. The point is to devise processes and systems that consistently reach Six Sigma performance goals.

In general, R&D professionals use House of Quality, control charts, and other tools that look familiar to people previously involved in corporate quality initiatives (Table 1). The difference between Six Sigma approaches and past approaches became apparent only in the context of implementation examples from various industries.

Creating the Future

In one example, the representative of a major industrial electronics and heavy machinery firm explained how its large central research organization is using Design for

Tabte 1.—Six Sigma and Design for Six Sigma Toots.

Balanced Scorecard Baldrige Criteria Benchmarking Business Process Reengineering Deming (14 Points of Management) Document Control DMADV/New Products & Services DMAIC/Product & Service Improvement Financial Analysis/Cost ofQuality ISO 9000 Management (leadership, motivation, change) Metrics (specific, measurable, actionable, relevant, and

timely) Plan, Do, Check, Act-PDCA Process Management (model-driven business process improvement) Process Modeling Project Selection (manageable project, measurable impact, capable organization) Simulation Taguchi Methods TL 9000 Total Quality Management (TQM)

March—April 2006

Six Sigma to "create the future" by building a defined and rapid process that uses ethnographic as well as technical information to devise novel and manufactu- rable solutions to poorly-stated customer problems. This organization possesses some of the world's best engi- neering know-how, but has found that the principles of DFSS are necessary to link that know-how to the tech- nology roadmapping and creative problem-solving processes that bring focus and actionability to vaguely- defined problems.

Another major industrial research organization uses Six Sigma and Design for Six Sigma methods and tools to focus "new opportunity" teams—people looking at creating wholly new products and services—simultaneously on opportunity identification (the "what") and idea creation (the "how"). This firm uses relatively familiar business strategy tools in new ways to understand how those new products and services will serve customers, exploit internal business capabilities, and effectively address the likely initiatives and counter-moves of competitors.

Fitting the R«&D Environment

Implementing these and other tools in an R&D context provides discipline, methods and infrastructure appropri- ate for problem-solving in the R&D environment; con- sequently, although research activities change as projects progress, approaches to problem-solving are dramati- cally improved using Six Sigma tools while keeping an appropriate research focus.

In using Six Sigma and DFSS to improve fundamental research processes, the focus must be to define and improve the repeatable processes that add value to knowledge production. Among these are processes that ensure appropriateness, reliability and reproducibility of experimental and research methods, define and protect intellectual property, and manage the physical plant.

One challenge for research leadership related to Six Sigma is to show the association between the impact of Six Sigma projects and customer satisfaction. TJiis can

be an issue because fixing things via Six Sigma usually removes customer dissatisfaction but by itself does not guarantee satisfaction. However, the nature of people in research organizations encourages use of information to prove benefits—whether from ad-hoc studies or from well-designed experiments. Consequently, it is important to choose or devise relevant metrics to track the customer satisfaction impact of each Six Sigma project.

Usually, R&D leaders find that several types of metrics are relevant—at least to answer questions about effi- ciency and effectiveness that arise during Six Sigma projects. One useful measure is "knowledge generated," which holds projects accountable for the explicit indica- tors such as numbers of patents per project or per unit of time. Another useful measure is products developed, trying to assess whether the products that are created really satisfy customer needs in effective and efficient ways. Also useful are measures of cycle time, such as time to market and the rate of change in sales per unit time, both of which help R&D leaders assess customer acceptance of new or updated products.

Table 2.—Companies Presenting Results in IRI Wortisttops on Six Sigma and DFSS in R&D

Avery Dennison Air Products & Chemicals Bethlehem Steel Boeing Callaway Golf Caterpillar Conoco Coming Incorporated Chemtura Delphi Automotive Dow Chemical Dow AgroSciences Dow Coming DuPont Eastman Kodak Eaton Ford Motor Company General Electric Honeywell Johnson & Johnson Lockheed Martin Lubrizol Maytag Motorola Nokia Omnova Owens Coming Praxair Rohm & Haas Samsung Seagate Sherwin-Williams Timken TRW W. R. Grace Xerox

Facts, narrative examples and specific quantifiable or soft benefits generated by practitioners of Six Sigma and Design for Six Sigma help R&D leadership communi- cate the needed insight and focus for success. Sometimes, detailed stories of other organizations' successes are enough to motivate interest and engage- ment in quality improvement efforts. In other cases, R&D leaders found that some Six Sigma project experi- ence within the organization was necessary. That project experience not only generates success stories internal to the company, but also in the case of project failure, gives the opportunity to celebrate the process of trying hard and trying well, without unreasonably penalizing people for failure to achieve the results desired.

A few companies implementing Six Sigma in R&D have suffered competitive setbacks—making amazing quality and productivity gains with the method but being out- flanked by competitors' product innovation strategies. However, to many innovative R&D organizations Six Sigma matters very much, and is done in conjunction with other changes in competitive strategy. The posters and experience reports at the IRI workshops have described projects of paradigm-changing strategic intent whose impact was achieved or amplified via Six Sigma and Design for Six Sigma.

One experience report outlined how DFSS accelerated development and improved the quality of large one-of-a- kind systems. The key lesson from that report was that DFSS proved very useful in a time-bound, deadline- driven systems engineering environment. A poster showed how a company making a relatively ordinary industrial commodity used Six Sigma methods and tools to re-think and re-engineer the product in dramatic ways, even including using nanotechnology to implement new and truly useful functions into their old-line product.

Several other presenters and posters showed how Six Sigma enhanced execution of business strategies that were intended to accelerate innovation, and in some

Research • Technology Management

cases amplify the impact of strategic paradigm shifts. These examples taught participants that although Six Sigma is no substitute for creative competitive strategy, it would likely amplify its effectiveness.

The most experienced practitioners state that although Six Sigma is necessary for improving core R&D processes, R&D generates the most commercial impact from DFSS projects. DFSS emphasis on customer needs (Voice of the Customer), multigenerational product planning, and early-stage resource investment leads to generating the most value from Six Sigma thinking and tools. This emphasis led to success in one or more areas of impact: increased sales from new product introduc- tions, improved ratio of desired compared to actual feature attainment in products, and reduced slippage of actual versus planned research and new product develop- ment projects.

No Passing Fad

Experiences and insights shared in the IRI workshops have shown that Six Sigma and Design for Six Sigma in R&D are important and effective. They are not the passing fad that some still believe them to be.

Several presenters pointed out elements that distinguish successfiil Six Sigma initiatives: broad executive spon- sorship, adequate resources, and inclusion of Six Sigma competencies in corporate goals for succession planning (e.g., the next and future CEO/CTOA^P will have been a Master Black Belt). So, although tools that are familiar to people previously involved in quality initiatives are used, the context of successful Six Sigma efforts broadly and deeply includes the people, systems and relationships across each company.

For many companies. Six Sigma in R&D often generates amazing results to share with customers, the public and employees. One company reported more than 20 percent reduction in the time needed to find and fix design fiaws in components of their "white goods." Another company reported solving a persistent problem with focusing cre- ativity in invention and development. Others have, over the course of several workshops, described how product quality as well as R&D productivity and impact (measured by percentage of sales from new products) have consistently improved since the technical staff was required to use Six Sigma and Design for Six Sigma.

R&D activities naturally defy systematic improvement efforts, and Six Sigma is not the only ingredient in an effective recipe for competitive advantage and profit- ability. The examples shown at successive IRI workshops help to explain how Six Sigma and Design for Six Sigma, linked with corporate strategy in an R&D context, have helped companies generate superior products and competitive profitability. These outcomes will at least help companies "stay in the game" in the face of aggressive international competition. ©

Acknowledgement

The author acknowledges and thanks the more than 550 R&D professionals who have participated in the Indus- trial Research Institute Six Sigma in R&D Project, espe- cially the presenters whose companies are listed in Table 2. This article is dedicated to our memory of one partici- pant and presenter, Charles B. (Chip) Huber, former executive director of DFSS for Seagate Technologies. Chip passed away at home on October 29, 2004.

Information for Authors RESEARCH • TECHNOLOGY MANAGEMENT welcomes manuscripts that deal with enhancing the effectiveness of technological innovation.

Manuscripts are reviewed by the Board of Editors, which looks for ideas and information to help industrial R&D/technology leaders run their operations more effectively. This means an emphasis on "real-world" experience that can be put to use by practitioners across a spectrum of industries. Articles based primarily on research studies should, therefore, deemphasize methodology in favor of explaining: 1) what the investigators learned, and 2) why those findings could be useful to industry managers.

Manuscripts submitted for review should be on good paper (not faxed), double-spaced, and paginated. References should be numbered in the order in which they are cited, and listed together at the end of the manuscript.

Illustrations should be individually numbered, furnished one per page on 8-'/2 x 11-inch white paper, and be suitable for black-and-white reproduction, without redrawing.

Send manuscripts to the Editor at [email protected], or RESEARCH • TECHNOLOGY MANAGEMENT, Industrial Research Institute, Suite 1102, 2200 Clarendon Blvd., Arlington VA 22201-3331.

March—April 2006