046 - Empirical Modelling Principles in Application Development for the Disabled
Abstract
Our ability to develop systems to meet special needs is influenced by several factors. The importance of emerging low cost computer technology cannot be denied, but this is not itself enough. In this paper, we are mainly concerned with the complementary role played by principles for computer programming and software development in exploiting technologies for the disabled. An abstract justification for our emphasis can be found in analyses of software engineering by Brooks [1] and Harel [2], which suggest an intimate connection between the challenges faced by modem software development and the broader problems of specifying and simulating complex systems of interacting human, electronic and mechanical components ("reactive systems"). From a more pragmatic point of view, the trend towards ever higher hardware specifications at ever lower cost has yet to be matched by comparable growth in the power and diversity of computer applications. Despite cheap computer technology, it remains difficult and costly to develop radically new software packages, or even to adapt existing software in comparatively modest ways. In practice, a new software product may be viable only where there is a mass market, because of the large investment in software development involved. Such issues are particularly relevant to the development of technology for the disabled, in view of the unusually high degree of innovation and customisation this typically demands.
An ongoing programme of research into computer-based modelling at the University of Warwick has led to the development of new principles and software tools that we believe can address some of the problems associated with more rapid and flexible exploitation of emerging computer technology. Many case studies in a wide variety of application areas have been investigated, both through sponsored research and through student projects at undergraduate and postgraduate level. The aim of this paper will be to highlight those aspects of our modelling method ("Empirical Modelling") that seem most relevant to developing software and hardware technology to meet special needs.