Kevin Otto

Extensions to the Taguchi Method of Product Design

The Taguchi method of product design is an experimental approximation to minimizing the expected value of target variance for certain classes of problems. Taguchi’s method is extended to designs which involve variables each of which has a range of values all of which must be satisfied (necessity), and designs which involve variables each of which has a range of values any of which might be used (possibility). Tuning parameters, as a part of the design process, are also demonstrated within Taguchi’s method. The method is also extended to solve design problems with constraints, invoking the methods of constrained optimization. Finally, the Taguchi method uses a factorial method to search the design space, with a confined definition of an optimal solution. This is compared with other methods of searching the design space and their definitions of an optimal solution. ∗EDRL-TR 90f: Manuscript to appear in the ASME Journal of Mechanical Design. †Graduate Research Assistant ‡Associate Professor of Mechanical Engineering, Mail Code 104-44, Caltech, Pasadena, CA 91125

Function Based Design-by-Analogy: A Functional Vector Approach to Analogical Search

Design-by-analogy is a powerful approach to augment traditional concept generation methods by expanding the set of generated ideas using similarity relationships from solutions to analogous problems. While the concept of design-by-analogy has been known for some time, few actual methods and tools exist to assist designers in systematically seeking and identifying analogies from general data sources, databases, or repositories, such as patent databases. A new method for extracting functional analogies from data sources has been developed to provide this capability, here based on a functional basis rather than form or conflict descriptions. Building on past research, we utilize a functional vector space model (VSM) to quantify analogous similarity of an idea’s functionality. We quantitatively evaluate the functional similarity between represented design problems and, in this case, patent descriptions of products. We also develop document parsing algorithms to reduce text descriptions of the data sources down to the key functions, for use in the functional similarity analysis and functional vector space modeling. To do this, we apply Zipf’s law on word count order reduction to reduce the words within the documents down to the applicable functionally critical terms, thus providing a mapping process for function based search. The reduction of a document into functional analogous words enables the matching to novel ideas that are functionally similar, which can be customized various ways. This approach thereby provides relevant sources of design-byanalogy inspiration. As a verification of the approach, two original design problem case studies illustrate the distance range of analogical solutions that can be extracted. This range extends from very near-field, literal solutions to far-field cross-domain analogies. [DOI: 10.1115/1.4028093]

A multi-criteria assessment tool for screening preliminary product platform concepts

Platform concept evaluation is a more challenging task than evaluating a single product concept since a platform must effectively support multiple product variants over a prolonged period of time. Existing platform methods develop specific criteria in depth, yet an evaluation of alternative platforms should be based on a broad set of criteria. Based on expert interviews, personal experience, and a literature search we propose a platform assessment tool consisting of 19 criteria for platform evaluation. The criteria are group into six categories: customer satisfaction, variety, after-sale, organization, flexibility, and complexity. The tool is focused on the early platform architecture phase, before proof-of-concept prototyping. However, it can also be used subsequently for platform refinement when more data becomes available. We demonstrate our platform assessment tool through an example with a cordless drill platform.

Evaluating Process Capability During the Design of Manufacturing Systems

This paper introduces the concept of a process capability matrix-an ordered set of dimensionless parameters that capture information on a manufacturing systems response to disturbances. The matrix is similar to the process capability indices C p and C pk , but is extended to multiple acceptance criteria and multiple causes of variation. Equations are presented that use the matrix to estimate yield in manufacture of products with multiple acceptance criteria. The surface mount of large body electronic packages serves as an example of the effectiveness of the process capability matrix as a tool for design decision making.

Manufacturing System Robustness Through Integrated Modeling

Robust design techniques are often applied to the design of manufacturing processes to determine the most robust operating points for a production system. However, such efforts have traditionally been focused on treating the output of each manufacturing operation in isolation. This approach ignores the fact that the sensitivity of each operation to input variation is a function of the operating point, which can only be changed in conjunction with the operating points of all other operations in that system. As such, applying robust design to each operation within a system individually does not guarantee lowest end-of-line variation. This is contrary to commonly held beliefs. What is needed instead is a method for conducting a system-wide parameter design where the operating points of each operation are optimized as a complete set to reduce final product variation. The logistics of such an integrated parameter design scheme become difficult or impossible on processes that may occur in different geographical locations. In this paper we outline the use of mathematical models to conduct system-wide parameter design. We demonstrate this technique on a model of a sheet stretch-forming manufacturing system. Through this example, we show that selecting operating points while considering the entire system results in a greater reduction in variation than Taguchi-style robust design conducted independently on each of the operations within the system.

Key Inspection Characteristics

This paper introduces a technique that can be used in the early manufacturing process design stage to select from among the set of quality characteristics a smaller set that is adequate to ensure a product meets yield specifications. The critical subset of quality characteristics needed to ensure yield are identified, based upon a conditional probability model that ensures all specifications are met, given that the critical subset is met. The approach is demonstrated using a sheet stretch forming manufacturing system from the aerospace industry.

A Systematic Method for Design Prototyping

Scientific evaluation of prototyping practices is an emerging field in design research. Prototyping is critical to the success of product development efforts, and yet its implementation in practice is often guided by ad hoc experience. To address this need, we seek to advance the study and development of prototyping principles, techniques, and tools. A method to repeatedly enhance the outcome of prototyping efforts is reported in this paper. The research methodology to develop this method is as follows: (1) systematically identify practices that improve prototyping; (2) synthesize these practices to form a guiding method for designers; and (3) validate that the proposed method encourages best practices and improves performance. Prototyping practices are represented as six key heuristics to guide a designer in planning: how many iterations to pursue, how many unique design concepts to explore in parallel, as well as the use of scaled prototypes, isolated subsystem prototypes, relaxed requirements, and virtual prototypes. The method is correlated, through experimental investigation, with increased application of these best practices and improved design performance outcomes. These observations hold across various design problems studied. This method is novel in providing a systematic approach to prototyping.

METHODS FOR PROTOTYPING STRATEGIES IN CONCEPTUAL PHASES OF DESIGN: FRAMEWORK AND EXPERIMENTAL ASSESSMENT

Prototyping may be simultaneously one of the most important and least formally explored areas of design. Over the last few decades, designers and researchers have developed many methodologies for ideation, product architecture, design selection, and many other aspects of the design process. However, there have been relatively few methodologies published regarding the efficient and effective development of prototypes for new products. This research explores a methodology for enhancing the prototyping process. It is founded on extensive literature review of the best practices of engineering prototype development. These findings have been aggregated and form the foundation of a methodology for formulating prototyping strategies. This methodology has then been experimentally evaluated in a controlled design environment, and its effect on the performance of prototypes has been demonstrated. The method consists of a set of guiding questions with corresponding flowcharts and foundational equations that assist the designer to make choices about how to approach the prototyping process in an efficient and effective manner.

Platform Concept Evaluation

A platform must support several product variants at any point in time and it must survive several life cycles into the future. The technology composing the platform itself is usually the embodiment of the core value-added capability of the developing company, yet what makes a good platform? This question often arises, for instance, when comparing two alternative platform concepts or deciding whether to update or replace a platform. The decision is more complex than a standard concept comparison exercise, involving forecasts of several applications and alternative technologies. Multiplicity and uncertainty characterize platform concept evaluation.

SIMULTANEOUS ENGINEERING OF QUALITY THROUGH INTEGRATED MODELING

Simultaneous engineering has become the means to ensure quality through incorporation of manufacturing. This proves difficult without quantitative support tools. We present here a modeling based approach to simultaneously design a product and its production process. We demonstrate an approach of examining sensitivities of output to all inputs of a system, from product design specifications, process variables, and material specifications. We further develop mathematics of estimating the effects of in-line process control changes to improve quality. We demonstrate a method to choose noise sources to measure and process variables to control in-line based upon the se measurements, and estimate the error reduction that such process control changes will provide. The too l allows simultaneous engineering of the product and process to improve quality. NOMENCLATURE