Involvement in sufficient physical activity declines rapidly during later adolescence and young adulthood. As social networks and interaction with friends is important for this age group (15-23 years), this should be included in future strategies aiming at motivating and empowering young people to engage in more PA. Since the majority of young people use the (mobile) internet for social networking activities, the current project aims to make use of web-based social networks, modern mobile technology and the social setting to provide cues to promote development of and engagement in PA and empower people to be more physically active. In this project three different technological solutions will be developed and combined. First, devices and reasoning mechanisms for capturing and understanding the context and physical activities will be developed. These form the basis for context-specific tailored feedback. Secondly, computational models of habit learning will be developed to make predictions of, and thus to guide, the effect of specific interventions (such giving a cue) to promote physical activity habits. Thirdly, a web-and mobile application based on social network sites of youngsters will be used to create and strengthen social support and descriptive norm groups of youngsters with similar motivations and contexts for further physical activity promotion. Thus, the current project aims to develop an innovative solution that is appealing to young people, makes use of existing social networks and mobile devices to automatically send coaching messages and feedback that empower people to become more physically active, based on an intelligent prediction of the effect of specific suggestions.