Design work involves uncertainty that arises from, and influences, the progressive development of solutions. This paper analyses the influences of evolving uncertainty levels on the design process. We focus on uncertainties associated with choosing the values of design parameters, and do not consider in detail the issues that arise when parameters must first be identified. Aspects of uncertainty and its evolution are discussed, and a new task-based model is introduced to describe process behaviour in terms of changing uncertainty levels. The model is applied to study two process configuration problems based on aircraft wing design: one using an analytical solution and one using Monte-Carlo simulation. The applications show that modelling uncertainty levels during design can help assess management policies, such as how many concepts should be considered during design and to what level of accuracy.

This paper presents the results from a UK survey about the knowledge and information requirements of managers and engineers in design and service. The survey aimed to gather requirements regarding the needs for retrieval of previous designs and the needs to capture knowledge and information from current designs to support future engineering tasks. The survey nature, method and instruments are described with the information and knowledge categories identified. The findings are analysed and contrasted with the results of other empirical studies. Notable results include the variety of information needs, unexpected absences and low frequency categories, the expectation of practitioners that tomorrow's needs will be similar to today's, the apparent failure to satisfy those needs, and suggestions from respondents for improvements.

Methods and tools to support engineering design should be based on an understanding of how it is conducted in practice and how information is used in the design process. The development of better tools and methods could thus be facilitated by a better understanding of information use in design practice. This paper explores how information use in design could be captured in-situ by process participants using a diagrammatic node-link modelling tool. Through literature review we synthesise a 'language' of concepts which we propose can be used by designers to describe information use in their work. The language contains classes for describing concepts in the product, process, specification and rationale domains. It is implemented in a diagrammatic modelling tool, which we illustrate by application to describe a design process fragment as conducted in a UK manufacturing company. We argue that studying the models designers create using this tool could provide insights into information use in practice and thereby support development of better design process support approaches.