Service-oriented agent architecture for unmanned air vehicles

• This paper has presented the structural aspects of service-oriented agents that form the core of an Intelligent Vehicle Control Architecture (IVCA) for autonomous air vehicles. • The architectural IVCA foundation lies on the basic service features of service-oriented computing, i.e. discovery of system capabilities, dynamic system reconfiguration, and decoupled interaction among applications. • The approach also improves the SA of the above vehicles by relying on OODA loops performed by agents that combine their knowledge and skills with the information acquired during missions. • The above architectural elements make it possible to achieve adaptive and reflective mission planning. • The autonomous nature of the IVCA paves the way for a reduction in the expensive deployment and operation of remotely-piloted vehicles, and brings within reach complex multi-vehicle collaborative missions that were previously too costly or logistically infeasible. • There is still significant investigation to be made into this promising IVCA approach. Future next steps of this research will be: • To implement the IVCA by means the ROS middleware and develop a SOA based on such a middleware that supports agent technology. • To test functional aspects of the IVCA through a computer simulation according to the test procedures and cases presented in this paper. • To take into account more complex evaluation scenarios. Initial scenarios involve just two single vehicles. Subsequent scenarios will include multi-vehicle configurations. • To extend the tests to on-air trials with real-scale prototypes.