Combining temporal planning with probabilistic reasoning for autonomous surveillance missions.

Area of Science:

• Robotics
• Artificial Intelligence
• Autonomous Systems

Background:

• Autonomous vehicles face challenges in surveillance due to unpredictable environments and limited resources.
• Current probabilistic methods for autonomous vehicle behavior have high computational costs in real-world scenarios.

Purpose of the Study:

• To develop a hybrid approach for autonomous vehicle surveillance missions.
• To enhance the capabilities of autonomous vehicles in search-and-tracking missions.

Main Methods:

• A hybrid approach integrating Monte Carlo simulation for probabilistic reasoning and automated task planning for strategic capabilities.
• Demonstration using a fixed-wing UAV in simulation and a Parrot AR.Drone2.0 quadcopter in a physical environment for search-and-tracking missions.

Main Results:

• The hybrid approach effectively addresses the limitations of purely probabilistic methods.
• Experimental results demonstrate the viability of integrating probabilistic and deterministic reasoning for realistic surveillance missions.

Conclusions:

• The proposed hybrid approach offers a more efficient and effective solution for autonomous vehicle surveillance.
• Integrating probabilistic and deterministic reasoning is beneficial for tackling complex, real-world missions.