With the development of connected vehicles and autonomous vehicles, various innovative sensing services are proposed to improve the driving experience based on the vehicle to vehicle (V2V) communication. However, due to the heterogeneous mobility pattern of different vehicular fleets and the dynamic nature of the vehicles, it is hard to systematically measure the feasibility of those services in large scale before implementation. In this work, we design a measurement framework called mChat to measure the interactions among heterogeneous vehicle fleets, which is quantified by the topology of the urban vehicular fleets. mChat utilizes two key metrics, i.e., static connectivity feature and dynamic coexistence feature, to explore the feasibility of real-time vehicle to vehicle service on inner-fleet and inter-fleets. We apply mChat on one month of real-world data of a Chinese city, Shenzhen, from (i) a regular vehicle fleet with 3 thousand vehicles; (ii) a taxi fleet with 14 thousand vehicles; (iii) a bus fleet with 11 thousand vehicles; (iv) a truck fleet with 3 thousand vehicles. Our measurement results reveal interesting results related to the feasibility of V2V-based sensing service, leading to utility implications for the upcoming applications for connected vehicles and autonomous vehicles.