A Formal Model for the Representation of Problems Based on TRIZ

Sébastien Dubois (1) , Philippe Lutz (2) , François Rousselot (3, 4) , Emmanuel Caillaud (1)

1 LGeco - Laboratoire de Génie de la Conception
2 LAB - Laboratoire d'automatique de Besançon
3 LIIA - Laboratoire d' Informatique et d'Intelligence Artificielle
4 LGC - Laboratoire de Génie de la Conception.

Designing, and especially in inventive design, is mainly understanding and solving problems. Simon describes the designer activity as a problem forming, finding and solving activity. Nevertheless, if the importance of the problem solving process is admitted, a lot of methods exist to guide the inventive process without integrating this fact. TRIZ proposes to answer this lack by focusing the design on a problem framing and solving process. On a generic point of view, TRIZ can be described as a rule-based problem solving method. In fact it is a whole of methods and tools which aim is to identify problems and formulate them through generic frames enabling the use of patterns of solutions. One of the disadvantages of TRIZ is its lack of formalization which implies a lot of difficulties to implement it and even to understand it. To fulfill this lack a formalization, a process is described in this article and a resulted partial model is presented. This model is partial as it is only focused on the TRIZ problem formulation frames. The implementation of the model is presented to show the interest of the model and to validate its efficiency. This implementation is argued by the description of a problem formulated by the use of a prototype of software.

Mots clés: Inventive design; UML; ontology building

Domaines: Autre

Sébastien Dubois, Philippe Lutz, François Rousselot, Emmanuel Caillaud. A Formal Model for the Representation of Problems Based on TRIZ. International Conference on Engineering Design, ICED 05, Aug 2005, Melbourne, Australia. pp.NA. ⟨hal-00340989⟩