Today's information technology is rapidly moving small computerised consumer devices and hi-tech personal appliances from the desks of research labs onto sales shelves and into our daily life. These include low performance PDAs, embedded computers in cameras, cars, or mobile phones, as well as high performance wearable computers and tablet PCs. Many of these devices are becoming essential tools that we rely on increasingly both in private and in professional settings. In addition, a growing number of locations are being outfitted with ubiquitous devices and networking access. This combination promises to enable new approaches to solve daily tasks and to open up new possibilities.

However, in order to use these systems new interaction metaphors and methods of control are required. Well-known interaction devices, such as mouse and keyboard are oftentimes unfeasible or even unavailable, thus rendering user interfaces that rely on them inappropriate. Other resources such as power or networking bandwidth may also be limited or unreliable depending on time and location. Moreover, the physical environment and context can change rapidly throughout the interaction with mobile systems and must be taken into account appropriately. Spatial and temporal relationships between devices and users are continuously changing as well and may be a key factor to be considered when interacting with a ubiquitous infrastructure.

In addition, we can expect a shift from single users towards groups, from single applications to multiple concurrent services, and from strictly personal to (semi-)public artefacts that can be configured dynamically. The resulting complexity needs to be addressed on all levels, from interface design to power issues. Within the field of Artificial Intelligence, several of these problems have been investigated for many years (such as how to make user interfaces more adaptive or how to deal with limited technical or cognitive resources). Therefore, AI methods are promising tools for building mobile and ubiquitous systems that are aware of the location and situation of their users, and that can unobtrusively adapt to these factors.

AIMS 2004 will be the fifth workshop in a series of AIMS workshops held at different conferences (ECAI, IJCAI, UbiComp). Due to the very positive response to last year's AIMS, this year's workshop will again be held in conjunction with UbiComp 2004 in order to further explore the benefits of combining research from artificial intelligence and ubiquitous computing.