ACHIEVING PRODUCT INNOVATION BY REDUCING VARIABILITY

Transcription

1 ACHIEVING PRODUCT INNOVATION BY REDUCING VARIABILITY John L. Schlafer Quality circles, Japanese automobiles, and Dr. Deming s teachings are frequently in the news, but one of the prime reasons products are improving is almost never mentioned: striving for quality frequently leads to product innovation. In the 1960s, for example, when electronics designers developed a circuit that unintentionally generated high heat, they designed a heat exchanger to conduct the heat away. Today, designers simply design a circuit that generates less heat, thus reducing the energy the circuit consumes and lengthening its life. This innovative focus leads to simpler designs and greater reliability. Simpler designs result in less statistical variability among sample products because the performance variability of the product tends to diminish as the designer evolves the design into its more basic rudiments. One key measure designers often use to evaluate their designs is variability from product to product frequently statistical measures such as standard deviation and variance are used and as the variability reduces, the design quality and reliability improve. This chapter describes how an R&D group at EcoWater Systems reduced performance variability in the monitor and communications system for a water softener commonly used for treatment of water in the home and thereby improved the product. The communication system provides a means for service and design engineers to gain information directly from the customer. The water softener contains a resin bed that filters and treats the water entering the home. The resin bed must be reconditioned every three to seven days by backwashing and treating with salt or a similar no-salt regenerate. The regenerate is held in a tank that is filled as needed by the homeowner with bags of regenerate. The monitor informs the homeowner when the level of regenerate is low and more regenerate needs to be added so that the softener continues to properly treat the water. In the past, this filling had to be done by the owner every two to four weeks. This frequency was quite regular and owners usually did not forget to do it. But in the past five years, the regeneration process design has improved JOHN L. SCHLAFER, Ph.D., P.E., is manager of product engineering at EcoWater Systems, Inc., in Woodbury MN.

2 in efficiency to the point at which refilling of the regenerate tank is necessary only every four to eight months. As a result, many owners do not remember to fill the regenerate tanks, and a monitor designed into the softener is necessary to remind them. AVOIDING PAST MISTAKES WHILE DEVELOPING A ROBUST DESIGN In the past, various monitors were developed, but they could not survive the severe environment of salt and water. Even today patents are being issued for monitors that are vulnerable to the same historical problems. The design usually includes a float, a weighing device, or a sensor in the regenerate feeder stream. All are subject to rapid corrosion or fouling, and thus, to early failure. This design approach is referred to as an intrusive device in that it intrudes into the regenerate in order to measure it. No known satisfactory means existed in the consumer market place to monitor regenerate level. This gap between a monitor the customer now wants and what has failed in the market place in the past had to be solved. 1 The R&D group had to find an innovative way to deliver the feature with the benefits the customer expects. It was clear that a nonintrusive monitor was necessary. After several brainstorming sessions, the R&D team realized that the softener s microprocessor has available to it the necessary information to be a regenerate level monitor. It records the amount of regenerate use each time the softener regenerates information it can accumulate over time. The problems of knowing how much regenerate is added by the customer each time he fills the storage tank and the variability of the salt and water meter measurements had to be understood and reduced. Despite these major concerns, developing a monitor that was software and electronics based had major advantages. Ecowater Systems electronic design and manufacturing capability is state-of-the-art for consumer products. This gives Ecowater a manufacturing-for-design advantage. The company can manufacture prototype designs quickly and easily being confident that regardless of the design requirements, production capability would not be a limitation. 2 This is a competitive advantage in that new designs can be developed rapidly with the confidence that manufacturing can produce them without the trial and error of finding a design that is manufacturable. This shortens the time to market. In other words, the manufacturing capability can match or is technologically ahead of the design capability. This is the inverse of the popular design-for-manufacture approach where the design process is limited in some degree by the manufacturing capabilities. The first limitation inherent in the design was that the accuracy and precision of the regenerate metering method was inadequate. The R&D team had to develop a robust design that would minimize the variability of the product s performance in the customer s environment. 3 Otherwise, wide

3 variation in regenerate usage would be obvious to the customer causing concern and expensive, time-consuming service calls. The variations anticipated were due to changes in pressures and temperatures of customers water systems and variation from one unit to the next. The microprocessor s predictive algorithm would be vulnerable to these variations and mislead the customer with early or late warnings. Luckily, the R&D group had already started to improve the quality of this phenomenon. A year earlier the team had gone through an extensive sequence of continuous improvement projects to improve laboratory test methods and the product design. 4 So new test methods and automated test equipment were in place identifying the parameters causing the problem. The continuous improvement efforts that had already taken place and the manufacturing-fordesign advantage in electronics set the stage for the rapid development of this innovation. Because the R&D group had learned how to resolve the variabilities in the product, design improvements tightened up the variabilities and made possible the nonintrusive monitor. When the microprocessor measured the regenerate metering, the group knew the measurement was much more accurate and precise. Misleading early or late warnings to the customer were now correctable, and the mechanical system s vulnerability to failure was eliminated. This process was predominately a learning and discovery process. Basically, product innovation and new product development are learning processes. 5 People involved must have the capability and drive to learn. The more innovative and new the product, the more learning is required. Aggressive learners evolve aggressive innovations. In addition, the learning process must be essentially complete before the introduction of the product into the market; otherwise, some learning occurs after the customer receives the product and that is perceived as poor quality. Although this situation can be resolved with service flashes, recalls, or warranty adjustments, it is still detrimental to a quality image. EXPERIENCED LEARNERS EcoWater Systems assembled experienced learners to carry out this project: that is, mechanical, electrical, chemical, software, and human factors specialists. Experienced learners are specialists who have learned from experience, practice, and education the breadth and depth of their favored specialty. For example, the software specialist who innovates an algorithm that corrects for a mechanical weakness or an electrical designer who can correct for a chemical sensor weakness with an innovative circuit are experienced learners. The mechanical engineer who can find the better fail-safe design that is inherently simpler, is manufacturable, and adds to customer perceived value is also an experienced learner. The ability to apply their specialty across technical disciplines is an important attribute. They continue to study their professional expertise and are not hesitant to experiment. They view problems as opportunities to seek out a new innovation.

4 For its part, management must provide room in some cases, literally for experimentation. Laboratory facilities, CAD systems, prototyping capabilities, and capable associates in all other relevant functions are key to product development. Programs must include time in the planning process for experimentation and for the opportunity to learn from mistakes. Some of the leading companies known for new products and quality provide room in the development process for up to three generations of prototypes. 6 This provides development people with the opportunity not only to stretch their designs, but also to correct them when an excessive stretch would have resulted in failure. If a new design is developed without any failure in the prototypes, the design is probably too conservative and the benefits to be gained by it too limited. Generally, the learning process carries with it a degree of unpredictability risk is inherent in the discovery of the unknown. Management must provide the environment for the development people to stretch their learning process and discover the unknown. In addition, if this type of learning process continues for the long term, many concepts and proven ideas can be banked for the next generation of products. If a need suddenly arises, such as a competitive threat or a regulatory requirement, these banked ideas can be applied during a short time frame resulting in a fast new product development. Experimentation yields two results: new information about the expected and discovery of the unexpected. The expected results lead to planned and orderly evolution of a product. Expected results are those an engineer can foresee or can design controlled experiments for in which to quantify results. In a sense, the process is one of confirming and detailing the expected. Unexpected results are where the unforeseen happens and a new theory, further experiments, or a new approach is required for understanding. The unexpected can yield surprising innovations, major breakthroughs in design, and new products. The experienced learner in a supportive environment finds the surprising and new innovations that get market attention. The monitor project had to be limited to what was known or could be confidently developed at the time because the timing of the project was fixed by an early commitment to the customer. The R&D group had to use the banked knowledge and techniques gained a year before on the meter testing and use brainstorming with capable people to meet the time frame. Thus the team focused on known technology with minor improvements: electronics, software, what the microprocessor could do, and our manufacturing-fordesign advantage. THE MONITOR INNOVATION The microprocessor accumulates in memory the amount of regenerate used since the last addition to the storage tank by the customer (see Exhibit 1). After adding regenerate, the customer reads a yard stick on the side of the tank and enters the reading into the microprocessor. The challenge was simplifying the customer-microprocessor interaction. The company did not want

5 Exhibit 1 Software and electronics may be used when the variability in the meter, clock, and regenerate added measurements are improved. to have the problems the VCR industry has had with complexity and confusion. Extensive efforts went into the display and method of customer interaction so that it was intuitive, simple, and fault tolerant. A previous generation of products was successfully developed with similar customer interaction, so the group was confident but not complacent. Early attempts at a similar monitor had an obvious problem. When the regenerate level got too low, the design called for an intermittent buzzer to sound reminding the customer to add regenerate. Most softeners regenerate at 2:00 am, and that is when the monitor buzzer would activate. It only took one field test in a home with small children who were awakened at 2:00 am before the warning device was quickly changed to a subdued flashing LED on the face plate of the monitor. Regardless of the confidence in the innovation, customer field testing is a vital element in new product development. During the development of a product, engineers often become too close to the design to have a customer s perspective. Therefore and as in this case, customer exposure during the development process can lead to critical learning. SECOND INNOVATION Water softeners and filters must meet customer needs for treated water when the customer demands it. The unit is also expected to minimize water wasted during regeneration, which includes backwash water and the regenerate (e.g., salt or non-salt regenerates). This has become particularly true with the increasing awareness of the harmful effects of waste on the environment. When a softener regenerates, first water is used to backwash the resin to clean out the filtrate accumulated. This backwash water is sent to drain as waste. In

6 the past, excess water was wasted just to be sure the softener was clean. With the need to conserve water more and more today, this backwash water must be minimized. This is also true for the regenerate. After backwashing, a salt or non-salt regenerate was used just to be sure the resin was regenerated. In addition the need to reduce chemicals in the environment makes it necessary to also minimize the amount of regenerate used. To minimize water waste and chemical use and design the unit to the best advantage, the R&D group had to learn the customers water use habits. Because customers do not know how much water they use each day, a customer survey would be ineffective. The team needed a way to get that information without customer disruption, cost-effectively, reliably, and without intrusive equipment. In addition customer water usage patterns are very diverse. Some people launder on Mondays and clean on Wednesdays. Others do those activities on Saturday. Whatever the case, the softener must be in ready condition for the next day as needed without fail. This wide variability in customer expectations presented a challenge for the design group. Determining how to get answers to these questions led to an innovative communication system embedded in the softener control. The water softener s meter and a clock residing in the controls provide data from which water usage each day can be determined. The audible beeper in the electronics that normally informs the users that they are properly interacting with the controls can also display contents from the memory. Software was written that transforms the needed information into a series of beeps similar to a Morse code that can be sent over any standard phone line to a service center computer as illustrated in Exhibit 2. This computer decodes the beeps into a complete report of information, including water use patterns and how the softener is responding to that pattern. Thus, the company has complete information to evaluate the gap between customer-measured needs and how the unit meets those needs. With sufficient population of units in homes with broad demographics, this innovation is now permitting the group to develop controls that are closing the gap. 7 We now have a control system that has minimal cost, is convenient for customer use, and will help us evolve an environment- and customer-compatible product. The top featured model softener carries within the product everything necessary to communicate in this fashion with the customer. CONCLUSION By focusing on developing improved accuracy and precision in the control system of a water softener, EcoWater Systems has been able to develop a nonintrusive regenerate monitor that is unique in the industry and adds very little cost for the features added. In addition, focusing on the customer water usage patterns and how the product responds resulted in a unique communication system between the customer product and the design function. The company and the R&D group have learned several key points during these programs:

7 Exhibit 2 System to Measure and Evaluate Product Capability against Customer Usage Innovation is a learning process and management needs to assemble participants who know how to learn and to provide an environment that stimulates learning and allows the freedom to make mistakes. Experimentation with prototypes provides two results: information about the anticipated and discovery of the unanticipated. The unanticipated results can lead to innovations with surprising market impact. Manufacture-for-design is a favorable company strategic position that permits rapid acceleration of new product development. When prototypes are made by the manufacturing operation, everyone gets familiar with the new product early. In addition, designers can be confident early in development that a design will be manufacturable. Focus on continuous improvement, including study of variability in products, leads to knowledge about the product which in turn opens up potential for innovations in product design. Teamwork of people from many different technologies in these cases software, electrical, mechanical, chemical and human factors leads to innovation due to the breath of knowledge applied. This is particularly true when team members have considerable experience with the technologies, maintain their expertise, and are experienced learners. Closing a customer need gap for example, where designs are unreliable but the customer has an important need for one is an opportunity for innovation and can be an important competitive advantage.

8 The new products have had very good response from the field. They are meeting customer expectations, and customer acceptance has been very good. But it is important to keep learning from the customer, and the new communication innovation between the customer and the design function is the key to this company s future learning. As always, those of us in R&D cannot be complacent. Notes 1. Szakonyi, R., Technology Management Vol. 2 (Auerbach Publications, 1993) pp Clark, K.B. and Fujimoto, T., Product Development Performance (Harvard Business School Press, 1991), p Phadke, M.S., Quality Engineering Using Robust Design (Prentice Hall, 1989), p Scherkenback, W.W., The Deming Route to Quality and Productivity (CEE Press Books, George Washington Univ., 1988), pp Reinertsen, D.G. and Smith, P.G., Developing Products in Half the Time (VanNostrand Reinhold, Inc., 1991), p Monden, Yasuhiro, The Toyota Management System (Productivity Press, 1993), pp Szakonyi, R., Technology Management Vol. 2 (Auerbach Publications, 1993) pp Acknowledgments The Sears Roebuck and Company Buying Department and Laboratory personnel deserve a special thank you for providing customer insight and guidance throughout the development and introduction of these products. The EcoWater Systems marketing department and colleagues in the product engineering department also deserve a thank you for having successfully labored through the learning process to deliver these innovations of which we are proud.