sFly

Despite the many impressive examples of human inventiveness, our technological solutions for micro air vehicles are still pale in comparison with the elegance, effectiveness, power efficiency and supreme functionality with which nature has solved so many problems. This is particularly true for the flight of birds: No comparable technical solution of a small, autonomously flying vehicle has yet been demonstrated, despite the many potential applications. It is typical for these applications that they require complete autonomy in navigation, small size of the air vehicle to access also confined out- and indoor spaces and ultra-low power consumption for extended operation times.

The small scale Autonomous Micro Air Vehicles requires new approaches to various aspects of navigation and control because existing solutions typically make use of complex and heavy sensors systems that also consume too much power to micro air vehicles. We propose therefore to investigate the use of digital cameras for all key aspects of autonomous operation from vision based flight stabilization support to global localization. For full autonomy all the computations will be carried out on an on-board embedded computer. Energy and weight constraints and limited computing power are major challenges which require novel and efficient computer vision approaches. Furthermore, we propose to use the local communication not only for exchanging data between the micro helicopters, but to estimate also their relative arrangement. This will enable us to realize small swarms of micro flying robots capable of fully autonomous and coordinated operation in cities and other dense environments.

Our S&T endeavor proposed in sFly project will therefore focus on 1) micro helicopter design, 2) visual 3D mapping and navigation, 3) low power communication including range estimation and 4) multi-robot control under environmental constraints. It shall lead to novel micro flying robots that are:

• Inherently safe due to very low weight (< 500g) and appropriate propeller design;
• capable of vision-based fully autonomous navigation and mapping;
• capable of coordinated flight in small swarms (3-5 robots) in constrained and dense environments.

In order to achieve these challenging overall goals of the project, major progress is required in the following fields:

• Advance micro helicopter technology in order to build fully autonomous flying robots that are below 500 g, have long flight autonomy and are inherently safe.
• Development and evaluation of flight stabilization and fully autonomous navigation using monocular vision as the only exteroceptive sensor.
• Obstacle detection and avoidance using optical flow and structure from motion
• Vision based SLAM with limited calculation power and memory
• Low power communication using GSM and local wireless system
• Distance estimation between air vehicles using local wireless system
• Optimal control of groups of micro-helicopters under various constraints.

All these challenges need to be addressed under the hardware constraints imposed by a micro helicopter, i.e. total weight of around 500g, low computation power and low energy. In the following we continue by re-visiting the current state-of-the-art and by presenting detailed outlines of the sub-goals.

Full Title: Swarm of micro FLYing robots
Funded by: EU-FP7, ICT-2007-3
Start-End Date: 01/01/2009 – 31/12/2011