The shared mobility model of transportation services in cities has gained significant attention due to the proliferation of on-demand ride-sharing applications and the advancement of autonomous driving technologies. In this paper, a new shared mobility model is proposed accommodating the activity attribute of users' trip requests. Our key goal is to determine the minimum fleet size required to satisfy all on-demand requests while minimizing the total travel costs. Since this is an NP-hard problem, the model leverages a set of novel heuristic-based components including the clustering-based formation of ride-sharing groups, carpool-like schedule and ridesharing schedule generation, and clique-based trip integration. All work together to obtain the set of energy-efficient shared route schedules. The proposed model can also be extended for a heterogeneous vehicle fleet configuration (e.g. vehicles of various capacity and functionality) to work for different types of trip activities.