Planning and controlling autonomous vehicles requires one to be aware of the state of the vehicles and their configurations, the state of environment (obstacles, weather conditions), account for possible changes in mission requirements as well as hardware failures. Operations with multiple autonomous vehicles pose advantages in many scenarios but complicate even further the job of the human operator. One approach for dealing with the added complexity is distributing the responsibilities among several operators but this is not scalable as it requires constant communication among the operators. In this paper we discuss the design, implementation and field results of a framework aimed at planning and coordination of multiple autonomous vehicles, as well as some results and possible improvements.