Creativity techniques for a computer aided inventing system

The development of a Computer Aided Inventing system, assisting product designers in the creative stage, is a lengthy process because of lack of any real systematization of design knowledge for software implementation. The existing development tools for knowledge-based systems offer limited support for intelligent design. A knowledge-based architecture for intelligent product implementation is described in this paper. It is aimed at facing those problems where radical implementation is needed. Intended solutions are not based on improvement of existing inventions, but are oriented towards a new technological jump. A modification of the function-behaviour-structure (FBS) ontology is proposed to analyze the product to be innovated. Then an innovative behaviour oriented search aids the designer to systematically conceive a set of alternative behaviours, by means of a combination of three different creativity methods. The identified function – alternative behaviour couple is then automatically translated into targets for an automatic patent digging activity. Modified FBS ontology is also exploited to classify all the information collected at functional, behavioural and structural level, and a specific algorithm provides both patent search and classification in the form of a tree-like diagram. The software can be implemented as a support for building patent technological surveys, a personal knowledge database, technological transfer and forecasting.