Several application domains could take advantage of the knowledge of the social signals. Among them, we address the emotional wellbeing for an aging population. This is a current issue emerged from the evidence that life expectancy keeps growing, augmenting the proportion of older people over the population, thus opened a social debate about how to face this phenomenon. To date, a vast majority of methods  proposed in this field addresses the problem of monitoring health status of people mainly considering physical attributes with a marginal consideration for the emotional domain although undeniably related to the overall wellbeing of an individual. The main goal of our research is to design methods for evaluating automatically the emotional wellbeing of elderly, including novel computational models and testing procedures to assess the emotional state in an unconstrained, unbiased, and ecological way. Together with the methods for an automatic evaluation, we will also investigate specific interventions, to be implemented automatically or manually by an operator, to influence positively the emotional state.