Fast Transitions of a Quadrocopter Fleet Using Convex Optimization

Fast, safe transitions of multiple quadrocopters are often required in the Flying Machine Arena. In this video, we use an algorithm based on convex optimization to plan collision-free trajectories.

In the first part of the video, the destination points are selected ahead of time and collision-free trajectories are pre-computed. All the trajectories are stored before execution. In the second part of the video, however, the next set of destination points is picked at random while the vehicles are still en-route, demonstrating that the algorithm is fast enough to be used in real-time.

* Project by
Federico Augugliaro, Angela Schoellig and Raffaello D'Andrea
Institute for Dynamic Systems and Control, ETH Zurich, Switzerland

* Based on the work by
Yang Wang, Ekine Akuiyibo, Stephen Boyd
Information Systems Laboratory, Stanford University, USA

* Filmed at ETH Flying Machine Arena
http://www.FlyingMachineArena.org