Rob Stone

Evaluating TRIZ as a Provider of Provocative Stimuli

The Theory of Inventive Problem Solving (known by its Russian acronym TRIZ) is used across the globe to help engineers working on product design and development. In previous papers the authors developed a Function-Based TRIZ method and began a validation process. A single innovative feature was identified in a sample of innovative products, and these features were traced backwards on the Function-Based TRIZ matrix. The Function-Based TRIZ matrix was deemed successful in cases where any inventive principle suggested by the matrix could have produced the innovative feature. During this process, the authors observed that no matter which principle was used, some mental leap was needed to apply it. Additionally, many different inventive principles can lead to the same concept. This paper examines a new hypothesis: that the provocative stimulus presented by any randomly selected inventive principle facilitates concept generation just as effectively as using the historical contradiction matrix to guide inventive principle selection. This paper presents a study on the benefits of using the TRIZ contradiction matrix to select the “correct” principles during concept generation. During this study, participants were asked to come up with concepts using one of two TRIZ matrices: the real one that contains historical knowledge from an extensive patent search, or a randomly populated one. The results of this exercise were then examined using modified versions of two concept evaluation metrics set forth by Shah: quantity and variety. This paper offers two contributions to the field. The first is a step toward understanding the role and importance of conflict mappings in TRIZ and TRIZ-like problem-solving methodologies. The second is a method for evaluating process variety (as opposed to outcome variety) when TRIZ is used to generate ideas.© 2012 ASME

Customer Needs Extraction Using Disability Simulation for Purposes of Inclusive Design

Current product design methodologies are not conducive to creating inclusive products (products that meet the needs of persons with and without disabilities). In this paper, empathic design principles and modular product design strategies are explored as part of a novel approach to inclusive design. A disability simulation suit is used to test if empathically derived customer needs from persons without disabilities can serve as a proxy for the customer needs of persons with disabilities. This data collection methodology both increases the safety and ease by which customer needs can be gathered and gives designers an empathic design experience with the products they develop. This paper presents the techniques involved in and the preliminary data regarding the collection of customer needs on known product pairs. These product pairs perform the same function, but one is designed inclusively and one is not. Prior work on module identification from customer need statements is extended to specifically address the design of inclusive products. As part of larger research, this data will be used to make generalizations about the customer needs for inclusive products within the context of modular product design to create inclusive design guidelines, which will reduce the effort and expense involved in creating inclusively designed products.Copyright

A Multiagent Approach to Identifying Innovative Component Selection

Though there is a clear correlation between profitability and innovativeness, the steps that lead to designing a “creative” product are elusive. In order to learn from past design successes, we must identify the impact of design choices on the innovativeness of an entire product. This problem of quantifying the impact of design choices on a final product is analogous to the problem of ‘credit assignment’ in a multiagent system. We use recent advances in multiagent credit assignment to propagate a product’s innovativeness back to its components. To validate our approach we analyze products from the Design Repository, which contains thousands of products that have been decomposed into functions and components. We demonstrate the benefits of our approach by assessing and propagating innovation evaluations of a set of products down to the component level. We then illustrate the usefulness of the gathered component-level innovation scores by illustrating a product redesign.

A Validation Study of Disability Simulation Suit Usage as a Proxy for Customer Need Statements From Persons With Disabilities

Current product design methodologies do not typically address the creation of inclusive products (products that meet the needs of persons with and without disabilities). In this paper, empathic design principles and modular product design strategies are explored as part of a novel approach to inclusive design. The use of disability simulation as a data collection methodology both increases the safety and ease by which customer needs can be gathered and gives designers an empathic design experience with the products they develop. A disability simulation suit is designed to support empathic gathering of customer needs. The suit is subjected to validation trials to determine how accurately it mimics the physical and perceptual behaviors of persons with disabilities in users of the suit. Initial results show that the disability simulation suit provides a valid empathic design experience that yields similar customer needs and reduced dexterity as persons with disabilities.Copyright

Investigating Divergent Thinking in Creativity Exercises Through Alternative Uses Tests

Creativity is a valuable skill for today’s workplace and one that universities should be emphasizing in the classroom. Teaching creativity usually involves the completion of “creative exercises” that help an individual understand how to think outside the box. Often individuals that are considered creative “practice” creativity on a daily basis, either through their own will or through their occupation, which increases their creative potential. Creativity is shown to be divided into multiple aspects, one of which is divergent thinking. This study examines participants’ divergent thinking skills over nine weeks as they perform a simple design task each week. The participants are split into two groups as they perform an alternative uses test on a weekly basis. Each week a new item is presented and the results are collected and entered in a database. The number of entries per card is analyzed to determine if the participants have increased their divergent thinking ability throughout the nine weeks.Copyright

Dynamics Modeling to Inform Design Optimization

This paper presents the methods employed in modeling a vibratory conveyor for use in model-based design optimization. The conveyor, essentially a large table whose top oscillates at an angle off of horizontal, uses springs between the drive mechanism and the tabletop to directly apply a sinusoidal excitation. These springs prevent the system from losing response amplitude as load is increased. The manufacturer is having difficulty optimizing performance and reliability in newer designs, and has requested a model-based approach to the design optimization. This study discusses the initial steps taken in modeling the original mechanism design, specifically the dynamic model and experimental determination of the necessary spring constants. The first full iteration of the model starts with low detail and simplified geometry, with a plan to add complexity as needed to improve accuracy. In the initial model, the parallel springs in the tabletop suspension are combined, bypassing the spring mounting geometry, and tested as one large spring. The drive mechanism springs, bars of fiber reinforced plastic (FRP), are more meticulously tested in a tensile testing machine. The resulting spring constants are used in the initial model to calculate the sinusoidal response of the tabletop at any given input frequency. The deflection response per time of the tabletop is then measured and compared to the model. Conclusions detail the initial model’s accuracy and Future Work examines how to bring it in closer agreement with the real machine’s sinusoidal response.Copyright

Ai edam 25th anniversary issue reflections

Reviewing the research activities of the engineering design community over the past few years (many of which find their home in AI EDAM) reveals that the field is on the cusp of a major breakthrough in computer-assisted design research. I believe we are on the edge of making a shift to computer-directed design. Consider the following: methods from casebased research or design by analogy are gaining acceptance and need more accessible stores of design knowledge from which to extract analogies. Automated concept generation algorithms exist that are capable of parsing hundreds of thousands of alternatives across a multitude of domains in a matter of seconds. Failure and risk prediction is possible now at the earliest stages of product design with appropriate statistical data on relevant prior designs. Design is more frequently grappling with large, complex systems that go beyond one designer’s, team of designers’ or single firm’s knowledge and ability to predict interactions of the vast number of components. Fully utilizing these findings and solving the remaining integration issues may well advance the field to the next level and era of computer-directed design. My particular interest in this area has been to imbue the engineering design process with artificial intelligence (AI), that is, what I am calling computer-directed design. As a designer, my favorite activity is concept generation. It is exciting to identify the real needs of a customer, to imagine new solutions to a problem, and to consider all of the permutations; but it is also time consuming to complete, impossible to be completely thorough, and difficult to know everything you need to about a particular problem. That is where it has been an absolute joy to be both a contributing party and a spectator of the AI in design community. Over the 15 or so years that I have been involved in this community, I have watched the research move from defining basic units for design computation (e.g., functional, behavior, and component lexicons) to creating more and more advanced ways of generating concepts (e.g., generative grammars, matrix-based synthesis, mining of biological analogies) and now to evaluating concepts (e.g., design preference modeling, uncertainty modeling). All of these activities are moving design computation further into the “front end” of design. We are producing algorithms and techniques that truly are AI and, although not replacing the human designer, can handle the more tedious tasks involved with producing and evaluating concepts. Although I do not advocate a future where HALor SkyNet-like computers run our lives, the promise of computerdirected design should allow the human designer to innovate further and faster based on the legwork that the AI methods analyze, synthesize, and present back to the designer. For 25 years, AI EDAM has been the place where the signature works in the AI in design community are published. As a member of the Editorial Board, I will make sure that remains true and look forward to being a part of and seeing the results of this dynamic community of researchers.