Ardea is a Micro Aerial Vehicle (MAV) that has been developed from the ground up at the German Aerospace Center's Institute of Robotics and Mechatronics. Two major research directions for MAVs are autonomous exploration of unknown environments and cooperation with emergency teams in disaster situations.
autonomous mapping and navigation in unknown, GPS-denied environments
System description
Especially desirable is the development of systems which can operate in diverse environments, for example in collapsed buildings or on alien planets, where GPS positioning is not available and self-contained localization, navigation and mapping using on-board sensors is crucial. Ardea is well suited to operating in such foreign environments and difficult terrain. An important research goal for Ardea is cooperative action with the LRU as a heterogeneous robot team where the various skills of each robot can be leveraged to achieve common goals more efficiently. Continuing development of Ardea is vital to well-rounded research of useful and autonomous systems — to that end our in-house engineering allows direct access to all critical on-board hardware, such as the motor currents, which are used to estimate relative wind speeds and other external forces.
Ardea has four Ultra-wide Angle Lens cameras which enable 240° view in the vertical direction. Thus Ardea has simultaneous coverage of the ground and directly above, facilitating navigation and mapping in narrow spaces such as caves. Additionally Ardea's FPGA runs Semi-Global-Matching (SGM) stereo image processing of the four cameras to produce depth maps which enable it to perceive the three dimensional world.
Ardea estimating wind velocities in wind tunnel
In order to validate the developed wind estimation method, Ardea was flown in different wind conditions in a 3D wind tunnel (WindEEE in Canada).
The heterogeneous robot team, consisting of the rover LRU and multicopter Ardea, is used to autonomously explore unknown environments. As a ground-based vehicle, the rovers' locomotion is energy-efficient and with its landing platform it serves as a mobile base to the multicopter. Due to the compact size and the ability to fly the multicopter is used in difficult terrain and even caves. Both robots have the same capability to use their optical and inertial sensors to simultaneously sense the 3D environment to avoid collisions and to create 3D Maps. Furthermore, resulting 3D maps can be exchanged between the robots and merged together.
