Functional assessments are designed to ascertain a person's ability to perform activities of daily living (ADL) and provide valuable diagnostic as well as care-planning information. Currently, the gold standard for the assessment of functional ability involves clinical rating scales. However, scales are often limited in their ability to provide objective and sensitive information. In contrast, information and communication technologies (ICT) may overcome these limitations by capturing more fully the functional, as well as behavioral and cognitive disturbances associated with Alzheimer disease (AD). In this context, the ANR Tec San 2009 SWEET-HOME project aims at building an innovative framework for modeling ADL. The first result of the SWEET-HOME project has been the installation in a classical consultation setting of a specific room equipped with audio and video sensors. This leads to the following results: (1) physical activity recognition done by patients, (e.g., balance test, repeated transfers between sitting and standing): the monitoring system is able to detect the full set of activities with a detection rate varying from 96.9% to 100% (true positive rate); (2) activity of daily living: the monitoring system had an average sensitivity of 90% and an average precision of 83.51%. Using a functional score it is possible to differentiate AD patients from healthy controls; (3) acceptability of the system: results of the survey (all the 64participants who accepted to be assessed using the system) indicated that the assessment has been perceived as pleasant (83%).