Autonomous Reconnaissance Capability for Unmanned Aerial Systems

Important results from the project

To automate reconnaissance tasks through the use of cooperative, autonomous drones in missions that were previously solved manually. This includes developing techniques for planning and simulation of a reconnaissance mission with three drones that carry out a joint ISR mission. The capability has been realized using open-source software and simulated and evaluated in two separate environments. In the first, extensive simulations and evaluations have been done and in the second we have demonstrated the possibility of integrating this capability in a system-of-system context.

Expected long term effects

The result shows how to develop a cost-effective and efficient drone swarm-based reconnaissance capability by exploiting the potential of open-source software and AI-based planning algorithms. The capability has been evaluated in two environments. Focus in the first environment has been to perform extensive simulations to identify the impact of different scenario configurations on the effectiveness of the search capability. In the other environment, focus has instead been on demonstrating the possibility of integrating the capability in a system-of-system context.

Approach and implementation

Focus has been on problem definition, planning/execution and evaluation of a collaborative search capability. As problem definition, a scenario has been defined. Which is realized by three drones that iteratively, search terrain areas via joint, continuous re-planning, triggered by the acquisition of new information about the area. Evaluation has been done through simulations to identify the impact of different scenario configurations on the effectiveness of the search capability. The possibility of integrating this capability with external systems has also been explored.