The demographic shift causes great challenges for the society, mainly due to the increasing of people's cognitive and physical limitations due to the increase of age. Hence, the integration and participation of elderly can be limited, especially in work environments with complex work processes.

One of most active fields to tackle such problems is Ambient Assisted Living (AAL), where great effort is being paid to develop assistive systems and technologies. In the last 10 years, many developments and projects have covered different areas, especially tackling daily life problems and challenges. Nonetheless, the massive adoption of technologies impose new cognitive and physical demands, especially for autonomous handing of Human-Computer interactions (HCI). Most of the previous research efforts in this direction focus solving problems (e.g., adaptation) on a micro level, where the focus is mainly on a particular aspect (e.g., illness, disability, work process, activity, etc.) or a particular usage context (e.g., Smart Home, Smart Office, Industry 4.0, etc.). Moreover, most current interaction approaches are designed and implemented in isolation, and only adaptive according to the One-Size-Fits-All.

More importantly, new interaction paradigms have being proposed to employ more natural abilities for interactivity in smart enviroments. Recently, hundreds of natural interaction techniques and technologies (Natural User Interfaces - NUI) have been developed. Those techniques aim for multimodal interactions based on the human senses (speech recognition) and the motoric abilities (whole body interactions). However, the current design, implementation, marketing, and usage of those technologies are fundamentally challenged in real world application scenarios. The vast majority of approaches in the development of interaction techniques and technologies do not take into account the characteristics of the individual users and are not adequately adaptive in ambient environments.

In this project, we propose and study a radical change in the process of developing and deploying of interaction techniques. We aim to develop an interactive ecosystem from dynamically composed interaction components (Interaction Ensembles) according to users' physical context, abilities, and preferences. Thus, interaction techniques in ambient spaces are modeled and deployed as cooperative, self-reflective, and context-aware interaction objects (Interaction Plugins).