To manage the growing penetration of renewables and electric vehicles (EVs), a concept of clean energy on wheels (CEW) is presented for active distribution networks in smart cities. The CEW is an EV fully equipped with battery energy storage systems and EV chargers to improve the performance of modern distribution systems, especially in peak load hours. A multi-area power flow calculation method is employed to predict energy surplus and deficit feeders and locations at the same time. A bi-level optimisation framework is developed to determine the integration buses and capacities of PV systems and CEW vehicles in multiple distribution systems in smart cities. A Peer-to-Peer (P2P) energy trading scheme is adopted to obtain optimal scheduling and power dispatch of CEW in different areas of the city. An improved version of the genetic algorithm is adopted to determine optimisation variables of both the levels. To demonstrate the applicability of the proposed model, eight low-voltage feeders of different areas are selected from the city. Moreover, multiple scenarios are proposed and investigated. The simulation results obtained are found to be promising to accommodate a high penetration of EVs and renewables while increasing the operational flexibility of associated distribution systems.