The idea of designing an integrated smart public transport (transit) shuttle service is stemming from the need to overcome the problem of using an excessive number of cars arriving and parking at a train station within the same time span. This problem results in high parking demand around the train station. Moreover some potential train riders will, instead, use their cars and hence become a party to increasing the traffic congestion. The purpose of this work is to examine an innovative transit shuttle system that will comply with (i) users' needs and desires, (ii) intelligent transportation technologies, and (iii) operator's viability. In order to arrive at a design of this innovative system, new stepwise multi-criteria and multi-strategy concepts are developed in this work. In the search for best solution, the minimum fleet size required for the shuttle service is determined followed by a development of an optimization model and a heuristic algorithm for optimal route-design. The findings show that the heuristic algorithm indeed provides good (optimal in the test case) results. The heuristic procedure then is implemented over medium-and large-sized networks, generated randomly, and found to provide good results within a reasonable running time. At the operational level 10 different routing strategies are examined, with all the combinations of fixed/flexible routes, fixed/ flexible schedules, a uni-or bi-directional concept, and short-cut (shortest path) and/or short-turn (turn-around) concepts. These strategies are investigated by employing a simulation model specifically developed and constructed for this purpose. This simulation model is used in a case study of Castro Valley in California in which the shuttle service is coordinated with the BART (the metro of San Francisco).