Hybrid architecture of autonomous vehicle system was built based both on function and behavior so that requirements in autonomy, real-time and adaptability could be met. Functional decomposition architecture was used to automatically control the vehicle while behavioral decomposition architecture was used to handle the complexity and variability of environment which highly challenged responsiveness and adaptability of vehicle’s planning system. Experimental results showed that a hybrid architecture based system enabled the vehicle to drive autonomously while having sufficient adaptability and realtime as the environment changes.