To develop control strategies for automating the flight of PAVs in various cases such as take-off amongst buildings, airborne navigation in a dynamic environment, selection and validation of a landing place and landing in the vicinity of obstacles.

For enhanced safety, and in order not to be reliant on centralised control of the PAVs in flight, every PAV should have on-board capabilities to perform the control and navigation of the vehicle. In this WP, the data from different sensors to perform control and navigation of a single PAV will be combined. Consideration will be given to the sensors that are particularly suited for use in this decentralised environment, including vision sensors (on-board cameras), inertial sensors and GPS receivers.

The main task in this WP is to develop the required algorithms to combine the sensory data such that the control of PAVs in cluttered environments can be performed. Furthermore, we will develop algorithms for the selection and validation of landing places, including the detection of obstacles.

A step-by-step approach will be used to bring a small-scale unmanned helicopter to autonomous operation. This flying platform will serve as a demonstrator for the different algorithms that will be developed in a safe and cost-effective environment.