Drawing and sketching is an essential tool for generating and exploring ideas. Virtually any medium can be used to create sketches, including paper and pencil, cardboard boxes, clay, charcoal, stylus and tablet, in fact just about anything can be pressed into service. Philippe Stark famously used a table napkin to sketch the design for his iconic ‘Juicy Salif’ lemon squeezer. The sketch may be the starting point for a design or a blueprint for a new design. Sketching is also one of the main ways that designers communicate and collaborate with others. There is experimental evidence that supports the idea that the interpretation of previous sketches can be used as a source of inspiration for generating new design concepts . Studies of creativity in design interpretation often connect unexpected discoveries with sketching activities .
In addition to sketching, designers also carry out extensive visual searches to explore ideas, seek sources of inspiration, and find out what has been done before and what is trending at the moment. This form of visual inquiry is an essential part of the design process. Thanks to the web, designers now have access to vast repositories of visual data. The challenge is the sheer volume of visual information that is available. Finding the ‘right’ image has become a skilled process. So what is needed is a way of finding images that goes beyond the standard keyword search. Part of the difficulty is that the data being sought is often imprecise; it is a hunch, a vague feeling rather then a precise image. In addition, the metadata used to describe and store the image in the first place may be inaccurate or misleading. What is needed therefore, are algorithms that can identify images, generate metadata and support iterative searching and recommendation of somehow related content.
Researchers working on the COnCEPT project are developing a prototype application that will support this idea, with the aim of assisting designers to undertake visual research whilst sketching and collaborating with colleagues. The aim is to provide designers with a system that allows them to use initial sketches to actively search for visual content such as similar, pre-existing ideas, products, or images on the web or other repositories.
The Figure below shows the system as it starts to work by searching for relevant data in the form of other sketches and images based on the semantic similarity to the original sketch. The results of the search and recommendation will change according to the actions taken by the user. Designers may make modifications to their original sketch, discuss and collaboratively edit sketches while the system automatically generates semantic metadata, finding other sketches that may have been generated during previous projects. Various sources of associated content - such as Google databases, museum collections, personal collections and project databases - can provide new sources of inspiration to support the ideation process and stimulate creativity.
Figure. Web based sketching application supported by associated content recommendation
 Schön, DA and Wiggins, G.: Kinds of seeing and their functions in designing, Design Studies 13(2): 135-156 (1992)
 Suwa, M, Gero, JS and Purcell, T: 1999, Unexpected discoveries and s-inventions of design requirements: A key to creative designs, in JS Gero and ML Maher (eds) Computational Models of Creative Design IV, Key Centre of Design Computing and Cognition, University of Sydney, Sydney, Australia, pp. 297-320