Abstract
The PhD work presented in this abstract is performed as part of the EMERGENCY project . The project develops decision support tools for mobile users in emergency response. A main difficulty when developing on-site end user solutions for emergency response is the need for flexibility. The solutions need to be flexible with regards to type of operation, special needs for the given operation, available and needed information sources, applications and services, available and needed sensors, available infrastructure, type of equipment to be used, work situation of the user, and modalities to exploit. ICT support for mobile workers in emergency response needs to integrate information and services from a number of existing ICT systems and sensors applied to the operational area [1]. Such ICT support must be highly adaptable both to the type of operation, characteristics of the operation at hand, and incidents happening during the operation, as well as the available types of equipment and the current user situation (e.g. degree of mobility). These requirements for flexibility have at least two implications. Firstly, that developing optimal solutions for all combination of needs will be utterly expensive. Secondly, that it is almost impossible to specify an optimal end user solution in advance. Thus, there is a need for development support that makes it easy to compose support tools, partly at design time and partly at run time. The PhD work will address methods and languages for specifying adaptable user interface components and the necessary framework for composing them at design- and run time. To solve the main problem outlined above, we will use model-based approaches for user interface development. The starting point will be a user interface modelling approach based on modelling patterns and compound user interface components [2]. In earlier work [3], we have showed how this approach may be used to facilitate user interface adaptation to changing context information at