TRIZ for the Real World

The Day 2 keynote by Prof. Pierre Collet from Strasbourg University was designed to challenge the TRIZ audience by showing us what is happening in the fast-moving, creative world of computer-generated engineering (yes, the computer as the creator!) He took us quickly through the history of (electronic) computing, showing the experiments that were done with evolutionary simulations as early as 1953. Now, “AE” = artificial evolution and is used widely in industry—everything from calculating train schedules to optimizing diamond cutting. (Frequent readers will remember the work of Prof. Noel Leon and his students, applying AE algorithms and TRIZ problem solving to the generation of walking/hopping transport devices, reported at both the Iberoamerican Innovation Congress and the Computer-aided Innovation meetings, 2008-10)


Collet chose a real-world example: how to evolve French crepes to get the best possible recipe, and showed us how the bio-concepts of random variation and of crossing of good variants can be very efficient generators of creative options. (Not so clear how the “best” solutions are chosen if the result is truly unique.) He also showed how this system could create an omlette, not a crepe, quite easily, if the variant on egg/flour ratio goes to an extreme.

A more technical world example was the NASA ST-5 mission, using 3 micro-satellites (tiny computers, tiny batteries, tiny solar cells…) but how do they send data back to earth? A nice TRIZ contradiction: powerful, unfocused or weak, highly focused antennas. 3 weeks of computation “Evolved” an antenna design –the original 10 meter package was smaller than 2 cm. Prof. Collet charmed the audience with the story, showing how minor changes in the mission requirements could not be met by minor changes in the antenna—a full new evolutionary simulation was required for each design. Our classical design paradigm found it embarrassing that they could not explain the designs, but the antennas were flown quite successfully.

He concluded by showing some of the massively parallel technology now available in Strasbourg very cheaply, and the wide variety of design projects that can be tried using AE. There will be much follow-up discussion of the TRIZ-AE interactions.

The first paper in the scientific technical session is “Substance Field Analysis and Biological Functions” presented by Sara Greenberg . (Reminder: I report on my experience at the conference, so except for the keynote sessions, at least half the papers won’t get mentioned! ) The aim of using su-field analysis was to develop a language that could be used for both biological and technical systems. The analysis uses conventional Su-Field analysis and constructions, with the differences becoming obvious when they aggregate the biological structures that accomplish the same functions.

I jumped to the practioner session for “Systematic innovation in enabling hybrid based preventive maintenance” by Nagappan Annamalai from Intel Malaysia. Key to the successful use of TRIZ is the comprehensive analysis of the system, including statistical designed experiments to understand and validate the cause-effect relationships within the system. Dramatic savings resulted from resolution of the contradictions by switching from schedule-driven to usage-driven PM, but management had a lot of concerns (maybe due to past history of non-scheduled maintenance?) Further application of the contradiction resolution tools resulted in a hybrid PM system , using both schedule and volume triggers. Preliminary results: 40% reduction in headcount, spares savings US$millions (exact number not specified), schedule improved by 75%.

“Identification and realization of innovation potentials at a drum brake using WOIS and TRIZ methods” presented by K. Hiltmann, showing a case that local industry FTE Automotive had asked the university students to solve. They use VDI 2220, a structured design process—one weakness is that it starts with a “problem.” WOIS starts with a need for innovation, which is a broader challenge, looking at strategy, market, target, customer, and system to decide if there is a project opportunity. The students had a very real world experience: the client said don’t make any changes. The full analysis of the squeaking cause/effect chain showed the opportunity for a very elegant TRIZ solution—same geometry, same material , same supersystem, but the conduction of the sound is disrupted by use of a 2-layer structure

I missed the gossip during the coffee break-the Samsung team is preparing for their “TRIZ Festival” next week and wanted to interview me. I’m not sure what benefit I can give to people who are already using TRIZ so much, but I sent them my advice about how to improve TRIZ and my good wishes for their festival.

Karen Gadd’s story of the implementation of TRIZ in BAE Systems (100,000 people, global) over the period 1998-now captivated the audience. Straight-forward training, pilot projects, etc., have evolved into a self-sustaining system that uses humor seriously as a teaching/learning method. Anticipated benefits were the standard TRIZ list (solve difficult problems!, take advantage of people’s natural creativity!) but the unanticipated list (everyone speak the same language, solves management problems, fun, happy & productive teams) has been the key to the acceptance of TRIZ throughout the company. Use TRIZ to propagate TRIZ: using the resources already present in BAE, such as the life cycle management system, the education portals, the case study libraries, and the growing cadre of enthusiasts. Karen spoke frankly about culture issues, and the need to avoid “gatekeepers” and non-TRIZ facilitators who try to limit and simplify the system. Her very real world learning (and BAE’s willingness to tell the story) were appreciated by all 3 ETRIA audiences—academic, industrial, and consulting.

Val Souchkov presented the “Function Value Map” after explaining that it is different from the tool he usually teaches-he is presenting on behalf of his colleagues Ives de Sanger and Kim Rutten. He reminded us that historically TRIZ had strong roots in manufacturing troubleshooting, and that developing this tool for process modeling is returning from the product phase to the process phase of TRIZ evolution. It is similar to Su-field modeling, with emphasis on critical activities defined as specific verbs, time studies and the financial impacts of the time and resources. This map has been very effective for explaining the opportunities for improvement (and TRIZ) to people who have had discomfort with their processes but not enough specific knowledge to move ahead.

Petr Shimukovich tried an innovative approach to presenting his paper “New method for TRIZ contradictions” – an English-language recording of the paper was played, while he managed the slides, then the questions. We’ll see what the audience thinks. The method examines 10 aspects of a system, then has a series of algorithms, such as a series of changes to the system requirements, which create opportunities for improvement. Shimukovich has extensive lists of the types of transformation that can be used to satisfy the requirements of each of these changes.

ANNOUNCEMENT: Next year’s conference will be in France, in Paris, at ParisTech, hosted by TRIZ France. Laboratoire Conception de Produits et Innovation. http://www.ensam.eu/

Afternoon program:  Keynote speaker for the afternoon program is Denis Alves Coelho, professor at Universidade da Beira Interior of Portugal, speaking on TRIZ and Human Factors and Ergonomics. He explained the current focus of HFE as cognitive work, Production processes (automation and globalization) and product design (being taken over by crowd sourcing.) Ergonomics enters the refinement phase of the design, when the moderators become the detail level designers. Likewise in the production phase, both HFE and TRIZ are needed to get from the concept to the implementation. Denis explored the simplest applications of the 40 principles to validate the idea of interlacing the two disciplines, and suggested the need for a collection of case studies to explore the overlap further.

Achille Souli from Strasbourg presented “A lexico-syntactic pattern matching method to extract IDM-TRIZ knowledge from on-line patent databases” using natural language processing. Historically, engineers have ignored important data from the patent databases because extracting it is time-consuming (and expensive if they need expert help or facilitation) and subject to error because of bad or confusing information in some of the patents. The approach proposed uses both super-marker and polyvalent markers, which are both nouns and verbs with very specific characteristics, which are used to split text in ways that can then be re-formulated as a graph. The geometrical properties of the graph are then interpreted. Achille concluded by telling us about additional research on TRIZ terminology, and inviting conference participants (and our readers) to comment on the work at www.souili.com/trizterm

In the short paper section I was able to hear part of Olga Bogatyreva’s paper on a TRIZ-based algorithm for biomimetic design. Her main theme was that with a proper theory, projects would be much easier and therefore cost less and have less risk. Lack of a common language for engineers and biologists is a significant roadblock to the development of the theory and methodology.

Fons Sweeger from Philips showed us “Revealing end-user driven insights using TRIZ tools.” There is a methodology called “End-user driven innovation” (not surprising, but new to me that it is a structured methodology.) He coupled this with the trends of evolution to show from the customers’ point of view what some probably next steps are. His “trend of sinning” – laziness, vanity, greed – got a lot of audience interest and started discussion of many different overlapping trends and analysis methods.

Pavel Livitov & Olga Livitova (far right in picture) introduced us to “The Principle of Feeling” and the history  of conflict between the systematic/logical generation of ideas using TRIZ with the intuitive, non-systematic decision making and implementation processes. The TRIZ audience had very mixed feelings about their method: learn to feel, not think. Accept feelings as data. They introduced “systemic constellations” which is a map made of living people representing a situation, and the assumption that the elements of the system will find their own solution (using themselves as resources!) Pavel gave us a TRIZ example, a projector that overheats. 50+ideas were generated using TRIZ. Then people were assigned roles in the systemic constellation, and the group waited for something to happen. For example, the LCD says “I feel uncomfortable if I am too close to the Lens” and then says “I ask the light source not to look directly at me.” We had a very good time with 5 heroic volunteers who enacted the feelings of a jar of strawberry jam, and the lid, and the seal, and the label, as they expressed their feelings about their relationships. (See photo, taken in dark conference room--man on left is the jar, next left is the jam, then seal, etc.)

The afternoon ended with Gaetano Cascini’s paper on” ARIZ85 and patent-driven Knowledge Support.” He started by identifying design methods driven by problems, by information, by solutions, or by knowledge. Ongoing research has been on information retrieval and information extraction, but not explicitly for the support of the problem solving process. ARIZ 85 has steps which request information both explicity and implicitly, from patents and from other sources, and there are steps where new information is potentially useful even if not requested. Detailed analysis of which steps of ARIZ are supported and which are not has been conducted, and exploratory tests with students. The students developed 100 search strategies, of which 10 were successful. He concluded by calling for research to develop robust algorithms and coordinating the research that is now going on in a very un-coordinated way.

An innovation in ETRIA organization—tonight there is an extra social event, a bus trip to hear Fado, the unique Portuguese singing. Thanks to the organizing committee for arranging it so that the visitors can appreciate this music/art in the busy schedule of the conference! See ETRIA’s Facebook page for photos—I plan to relax and enjoy the singing.