Image Image Image Image Image Image Image Image Image Image
Scroll to top

Top

TRIZ on Design-oriented Systems

| On 02, Mar 2009

By Vicente Chulvi and Rosario Vidal

Abstract


The future of computational synthesis models involves taking a creative advantage on computer-aided inventing (CAI) tools. The Theory of Inventive Problem Solving (TRIZ) is a set of standards that was recently applied to CAI tools including the CREAX Innovation Suite, Goldfire Innovator, Innovation Workbench and TechOptimizer. It has been studied in order to find a way to integrate TRIZ within computational synthesis models. Most of this research has been theoretical, but there are studies that have sought to put a model into practice. Several of these studies describe proposals for CAI models based on the architecture of a TRIZ system as well as proposals to allow it to be integrated through topological optimization systems.


The purpose of the present study is to establish a relationship between the National Institute of Standards and Technology’s (NIST) functional basis, which is a set of generic taxonomies of engineering functions and flows and TRIZ taxonomies as a first step toward the creation of a library of TRIZ-based algorithms. These algorithms are used to integrate TRIZ tools into a knowledge-based system (KBS) that is focused on functional design through a function-behavior-structure (FBS) framework.


Keywords


TRIZ, knowledge-based systems (KBS), function-behavior-structure (FBS), function, taxonomies, functional basis


Introduction


The Theory of Inventive Problem Solving (TRIZ) is a set of inventive tools that is frequently used as a CAI tool.3 It has been studied in order to find ways to integrate it within computational synthesis models.7,9,18 Several of these studies discuss the theoretical way to integrate TRIZ creativity with semantic knowledge in order to provide assistance in the design phase.19 But other research has also sought to put a model into practice.25 In order to link TRIZ with a functional model an understanding of the function within the FBS framework is needed.24,26For this work the authors used a derived model, known as the behavior-driven function-environment-structure (B-FES).23,27


The development of taxonomies at the University of Idaho introduced the idea of taxonomies at the industrial level.15 John K. Gershenson, Ph.D. and Larry A. Stauffer, Ph.D. (authors of “The Creation of a Taxonomy for Manufacturability Design Requirements”) defended that there is a taxonomy for each step in the product life cycle at a corporate level (i.e., marketing, business environment, strategic management, finance, accounting, manufacturing, shipping, support/service and retirement).


Simon Szykman, chief information officer at the National Institute of Standards and Technology (NIST), started researching the representation of functions.22The findings differentiated between flow and function and they also presented an extensive review of related work about flow- and function-based terminologies within the engineering context from 1976 until 1998.


A few years later, Julie Hirtz, one of the authors of “A Functional Basis for Engineering Design: Reconciling and Evolving Previous Efforts,” prepared a reconciliation of flow and function taxonomies for NIST, called the functional basis.17 This classification is used in the present work to search for correlations among the terminologies of TRIZ functions.


Other authors have also built their own function taxonomies or methods to deal with them in an attempt to optimize and facilitate the integration of functional analysis with computational models.13,16,20 Still other authors in a similar way have focused their work on behavioral, as well as structural, taxonomies as well as structural.4,5,6,14 The integration of TRIZ tool taxonomies with computer applications through a KBS of tools has been a topic for several authors.7,8,10,11,12


Materials and Methods


The structuring of any computer application (if it is to be useful and adaptable to the final user’s needs) requires a common framework. The framework used is B-FES (a variant of the FBS framework). The terms “function,” “behavior” and “structure” have been used since the 1990s in order to define a framework where the functionality of a system can be modeled and represented. Within this framework, function represents the functions of the system, structure represents the physical elements and behavior acts to relate them both (Figure 1).


The set of research algorithm solutions based on TRIZ for the model library must be built within this framework. In addition, it must be able to integrate with a KBS and a common unequivocal taxonomy must be chosen to work with all libraries in order to achieve this integration. The taxonomy selected for use in this work is the NIST’s functional basis.17




Figure 1: FBS Framework


The functional basis is a reconciliation of the most outstanding function taxonomies that have appeared within the field of mechanical engineering to date. One of the function taxonomies analyzed here consists of 30 mechanical design-related functions provided by the founder of TRIZ, Genrich Altshuller.1The functional basis appears as a result of a flow taxonomy and a function taxonomy. These are organized at three levels and a set of correspondences are provided as synonyms.


Despite the fact that the functional basis takes into account the function provided by Altshuller, there are some hindrances that prevent an individual from using this direct interaction between the functional basis and TRIZ in an easy, straightforward way.1 The first hindrance is that the TRIZ functions used by NIST are related only to mechanical design, while a broader application level is desired. The second hindrance is that these TRIZ functions date from 1984 and consequently, these terms have evolved considerably since then. This calls for a greater TRIZ-based functions analysis.


For example, consider the computer tool, CREAX Function Databaseas asource for TRIZ functions. This database consists of an encyclopedia of solutions linked to functions. It takes into account 37 different primary functions and each of them provides a set of solutions that perform a particular function when linked to a state: solid, liquid, gas or field. When examining the B-FES framework it can be seen that these solutions may be both sub-functions and behaviors. The same Web site allows the user access to the tool, CREAX Attribute Database. This tool provides a set of solutions to perform any of the functions: increment, decrement, stabilize, change and measure as they are related to each of the 26 attributes. Now the total number of primary functions is 42, which an individual can use to compare with those of the functional basis.


The function level was developed from these primary functions in the first step of the functional design process. The software, CREAX Innovation Suite 3.1, and its system model tool were used to analyze the functional level of an item.


Correlation Results


It was found that not all the items can achieve a direct one-to-one correlation when linking TRIZ functions (provided by CREAX databases) with the functional basis. In fact, four different cases appeared, as shown in Table 1.




TD>Indirect correlation












Table 1: Sort of Correlations
CaseDescriptionExample: TRIZExample: Functional Basis
Direct correlationThe function shares the same name in TRIZ and in the functional basisExtractExtract
The function has a different nomenclature in TRIZ than in the functional basis, but a one-to-one correlation can be performed Polish Remove
Correlation 1:2 The TRIZ function can be related to two different functional basis terms Move



Move
Remove



Transport
No correlation No term within the functional basis can fulfil the meaning of the TRIZ function Vibrate ø


Next, Table 2 shows the correspondences between the NIST functional basis and the TRIZ functions provided by the CREAX Function Database and the CREAX Attribute Database. It shows how the result is like a set of synonyms for each term. It is also possible to observe the gaps in some of these terms. The CREAX terms that correspond to more than one term in the functional basis appear in italics.




Table 2: Correspondence Among the NIST Functional Basis and TRIZ Functions in the CREAX Function Database


Example: Underwater Robotic Explorer


An underwater robotic explorer (a camera called the Bleeper by Praesentis) was used to develop an example of the system model tool. This camera allowed the seabed to be seen and recorded from the surface. It can be submerged 50 meters underwater and it is connected to the surface by an umbilical wire. The wire supplies energy and sends the image signal to the surface.


This modeled system can be seen in Figure 2. The structures are represented by yellow boxes and the arrows represent the functions, in accordance with the TRIZ taxonomy in Table 2. Two different environments appear in this image:one corresponding to the surface and the other to the seabed. Not all the structures shown belong to the explorer (solution). There are certain structures that belong to the environment like fish and saltwater. These special kinds of structures are called “restrictions” and they cannot be modified.




Figure 2: System Model of an Underwater Robotic Explorer


The same figure also shows an important detail. When applied between two structures (subject and object) the different functions take on different shapes and colors that give them a special meaning (Table 3). In this case, this can be seen as a basic behavior; this characterization is the way to establish a link among the functional design and other TRIZ tools including the contradiction matrix. The negative behavior of one function over one structure and the positive behavior of another function over the same structure can lead to a contradiction.




Table 3: Meanings and Interpretation of the Arrows Within the FBS Framework in the System Model Tool

Symbol

Kind of action

Effect

Effective action

This is the required behavior.

Insufficient action

The effect of the function is not enough to reach the desired level. This behavior can be improved.

Excessive action

The function performs too much action over the structure. This behavior can be improved.

Harmful action

This is an undesired behavior.

Missing action

A function is needed, but there is no behavior that carries it out.


Conclusion


There are two basic difficulties with trying to link the NIST functional basis and the TRIZ functions at a functional level on the CREAX Function Database:



  1. The large number of synonyms that appear in some cases, like the functional basis term remove, which corresponds to the TRIZ terms remove, polish, corrode and erode (Table 2). There is a lot of information missing in these cases when transferring a function between functional design models and TRIZ tools.
  2. The lack of TRIZ functions to fulfill all functional basis terms. A solution is needed to fill these gaps.

The final goal is the integration of TRIZ tools into a KBS that is focused on functional design. This requires an information exchange, which was assumed at the start of this work to be achieved through functions. The most adaptable TRIZ tools for that purpose are Su-Field models, the contradiction matrix and trends of evolution (a great example of this is evolution trees).21 But none of these tools works accurately with functions nor the solutions provided by the CREAX Function Database and CREAX Attribute Database. If trying to fit them into the proposed B-FES framework, it can be deduced that they act as sub-functions or as behaviors.


In Figure 2, the system model (or functional model) lacks TRIZ functions that are included in the functional basis and vice versa. Thus, the function signal (referring to signal capture and transmission) does not appear in the TRIZ taxonomy, but it is included in the functional basis. Furthermore, there is the need to use sub-functions and/or behaviors, as in the case of the relationship between light and the environment, where illuminate has to be used. When applied over two structures the same characterization of functions provides a first basic behavior. This fact can be compared with other similar methods, like functional clues.20 Here, structures are subject and object and applied functions are equivalent to actions when creating a generic application condition.


Consequently, the final conclusion from this work is that attempting to link functional design to TRIZ tools is possible, but using the function level does not seem to be the best way to do it. TRIZ tools need a level of specificity that functions cannot provide. Functions are too general and they seem better suited to state the general purpose of a device in an abstract way. The functions used in TRIZ are directly related to physical parts (structures). These statements fit in the definition of behavior. Behaviors are more specific and are directly related with structure (as shown in Figure 1). It seems best to use the behavior level for integrating TRIZ tools into a KBS focused on functional design through a FBS framework. Future work will be focused on the taxonomy of behavior in order to elaborate a model that allows the desired integration.5


Acknowledgements


This research was partially supported by the Ministry of Education and Science in Spain and the Spanish Rare Disease Federation (FEDER) from the European Union. They are also grateful for the help of the members of the Engineering Design Group (EDG).


References



  1. G.S. Altshuller, Creativity as an Exact Science: The Theory of the Solution of Inventive Problems, Gordon and Breach Science Publishers, Luxembourg, 1984.
  2. G.S. Altshuller, Introducción a la innovación sistemática: TRIZ: TRIZ,Translated by Lose M. Vicente Gomila Internet Global S.L., 1997.
  3. G.S. Altshuller and L. Shulyak, 40 Principles: TRIZ Keys to Technical Innovation, Technical Innovation Center, Worcester, MA, 1997.
  4. A. Bonaccorsi, Grammars of Creation. Mapping Search Strategies for Radical Innovation, Innovation Pressure Conference, Tampere, Finland, 2006.
  5. S. Borgo, M. Carrara, P.E. Vermaas and P. Garbacz, Behaviour of a Technical Artifact: An Ontological Perspective in Engineering, Frontiers in Artificial Intelligence and Applications, Vol. 150, 2006, p. 214–225.
  6. S. Borgo, A. Gangemi, N. Guarino, C. Masolo and A. Oltramari, Wonderweb Deliverable d15, 2002.
  7. G. Cascini, State-of-the-Art and Trends of Computer-aided Innovation Tools, IFIP 18th World Computer Congress, Toulouse, France, 2004, p. 461-470.
  8. G. Cascini and P. Rissone, Plastics Design: Integrating TRIZ Creativity and Semantic Knowledge Portals, Journal of Engineering Design, Vol. 15, 2004, p. 405-424.
  9. D. Cavallucci and N. León Towards “Inventiveness-oriented” CAI Tools, IFIP 18th World Computer Congress, Tolouse, France, p. 441-452, 2004.
  10. H.T. Chang and J.L. Chen, The Conflict-problem-solving Cad Software Integrating TRIZ into Eco- innovation, Advances in Engineering Software, Vol. 35, 2004, p. 553-566.
  11. H. T. Chang, J.L. Chen, T. Tainan, Eco-innovative Examples for 40 TRIZ Inventive Principles,The TRIZ Journal,August 2003.
  12. S. Dewulf, Directed Variation: Variation of Properties for New or Improved Function Product DNA, a Base for “Connect and Develop, ETRIA TRIZ Futures, Kortrijk, Belgium, 2006.
  13. N. Feygenson, Function Synthesis: New Methodological Tool and Case Studies, ETRIA TRIZ Futures, Kortrijk, Belgium, 2006.
  14. J.S. Gero, K.W. Tham and H.S. Lee, Behaviour: A Link Between Function and Structure in Design, IFIP Intelligent Computer Aided Design, North Holland, 1992, p.193-220.
  15. J.A. Gershenson and L.A. Stauffer, The Creation of a Taxonomy for Manufacturability Design Requirements, Design Engineering Technical Conferences, Vol. 2, 1995, p. 305-314.
  16. I.J. Golden, Function Based Archival and Retrieval: Developing a Repository of Biologically Inspired Product Concepts, Department of Mechanical Engineering, 2005.
  17. J. Hirtz, R. Stone, D. McAdams, S. Szykman and K. Wood, A Functional Basis for Engineering Design: Reconciling and Evolving Previous Efforts, Research in Engineering Design, Vol. 13, 2002, p. 65-82.
  18. S. Ikovenko, TRIZ and Computer-aided Inventing, IFIP 18th World Computer Congress, Toulouse, France, 2004, p. 475-485.
  19. E. Mulet and R. Vidal, Integration of Inventive Principles of TRIZ in Systems of Conceptual Design Attended by Computer (CACD), AEIPRO-International Congress on Project Engineering, Spain, 2005.
  20. A. Pinyayev, Functional Clues, ETRIA TRIZ Futures, Kortrijk, Belgium, 2006.
  21. N. Shpakovsky, Evolution Trees. Analysis of Technical Information and Generation of New Ideas, Processing of Technical and Patent Information by Using Evolution Trees, Japan, 2003.
  22. S. Szykman, J. Racz and R. Sriram, The Representation of Function in Computer-based Design, Design Engineering Technical Conferences, Las Vegas, Nevada, 1999.
  23. S.B. Tor, G.A. Britton, W.Y. Zhang and Y.M. Deng, Guiding Functional Design of Mechanical Products Through Rule-based Casual Behavioural Reasoning, International Journal of Production Research (IJPR), Vol. 40, 2002, p. 667-682.
  24. Y. Umeda, H. Takeda, T. Tomiyama and H. Yoshikawa, Function, Behaviour and Structure, Applications of Artificial Intelligence in Engineering, Springer, Berlin, 1990, p. 177-194.
  25. M. Verbitsky, Semantic TRIZ, TRIZ Journal–Español, 2004.
  26. K.L. Wood and J.L. Greer, Function-based Synthesis Methods in Engineering Design, 2001, p. 170-227.
  27. W.Y. Zhang, S.B. Tor and G.A. Britton, Automated Functional Design of Engineering Systems, Journal of Intelligence Manufacturing, Vol. 13, 2002, p. 119-133.