The main idea of the Ambient Intelligence vision is a paradigm chance of traditional human-computer-interaction. Yet, humans have to handle technical devices manually and they have to be aware of the devices’ functions and their interrelationships and cooperation possibilities. In future, Ambient Intelligence will assist users in achieving their goals and will thus disburden humans from excessive demands and cognitive overload in technical or foreign environments. First of all, Ambient Intelligence puts the human being in the centre of future technical developments. Already at an early stage of Ambient Intelligence the IST Advisory Group identified a variety of key technologies that are important for the assembling of smart environments and for the implementation of personal user assistance. This thesis presents solutions for some key technologies for Ambient Intelligence environments. The first part of the thesis specifies a distributed software infrastructure that allows a dynamic orchestration of devices and realizes selforganized communication and cooperation of autonomous device entities by means of execution of conflict resolution strategies. There are some advantages of the presented software infrastructure compared to other software solutions. That is primarily an increased support of dynamic ensemble behaviour. This is achieved by abandonment of any central component and the application of domain-dependent conflict resolution strategies. Thus the presented software infrastructure could be used in a variety of self-organized Ambient Intelligence scenarios. The second part introduces a component topology for Ambient Intelligence that supports the implementation of multimodal interaction in smart environments. Furthermore the necessary strategies for self-organization for multimodal environments are defined. Based on these fundamentals multimodal and assistive applications in the home entertainment as well in the automobile domain are described and discussed. Directly following, based on visualization techniques, developing tools for information flow control within a component ensemble are presented. Additionally a rule interpreter that allows the configuration of smart environment behaviour as well as its application is described and discussed. The presented thesis comprises solutions for key technologies as defined by the ISTAG. The main focus of the presented work relies on enhanced infrastructure, functional co-ordination, intelligent scheduling, and user profiling respectively reflexive systems. After an in-depth discussion of the results current projects and their applications based on this thesis’ results are introduced. They are dealing with industrial, agricultural and home scenarios and demonstrate the large impact of Ambient Intelligence on today’s information technology. This thesis ends with an outlook on future scientific work.