Rhythmic flight of multiple quadrocopters
We aim for a novel visual musical experience: a dance performance featuring multiple flying vehicles, which move in synchrony with the music and perform compelling maneuvers. The Flying Machine Arena forms the stage, and small autonomous quadrocopters are the actors of this performance. We envision an expressive flight performance that is automatically composed and controlled, given any arbitrary piece of music.
The research carried out within the Music in Motion project includes various topics: characterization of periodic motion primitives, assessment of motion feasibility, generation of collision-free trajectories for multiple quadrocopters, and development of algorithms and software tools that enable the creation of quadrocopter choreographies. Music in Motion incorporates many results from the Flying Machine Arena research team, such as trajectory following controllers or aerobatic maneuvers.
On the road towards the automatic generation of multi-quadrocopter choreographies, many challenges must be solved. They include:
In order to create a choreography that reflects the music's character, we require information about the selected piece of music. Using publicly available beat extraction software, we analyze the music to determine beat locations. Periodic quadrocopter motions are then synchronized to the music beat.
To make the quadrocopters dance to music, it is essential to time the robot movement with the music beat precisely. However, experimental results have shown that, at steady-state, the quadrocopter response to a periodic sinusoidal input exhibits a constant error in amplitude and phase, resulting in asynchrony. Therefore, an appropriate synchronization algorithm is indispensable ,.
Given the body and dynamics of a quadrocopter, translating music into suitable motion patterns is a significant challenge. Unlike humanoid dance robots, which may imitate human dance motion, it is not straight-forward for quadrocopters to produce expressive gestures. In addition, motor and sensor limitations constrain the range of motions achievable with these flying vehicles . These limitations must be taken into account at the design stage.
We envision a system that automatically generates a dance performance with multiple quadrocopters, where the quadrocopters' motion reflects the character of the music. To this end, motion primitives are first classified and parametrized (forming a motion library), and then concatenated in meaningful sequences. The concatenation step is currently manually done during the design stage ,.
A choreography is created by concatenating motion primitives from a motion library. Smooth, collision-free transitions for multiple robots between consecutive motion primitives are essential for a proper execution. We developed a method for generating collision-free trajectories for multiple vehicles from a set of initial states (specified by position, velocity, and acceleration) to given final states . The trajectories satisfy various constraints, including the physical limits of the vehicle, space boundaries, and a minimum distance between vehicles.
A list of current student projects can be found on the IDSC Theses and Projects page. Motivated students who are interested in doing a custom project related to the Music in Motion project should contact Federico Augugliaro.
The following students contributed to this project:
An archived version of this page can be found here.
Diese Website wird in älteren Versionen von Netscape ohne graphische Elemente dargestellt. Die Funktionalität der Website ist aber trotzdem gewährleistet. Wenn Sie diese Website regelmässig benutzen, empfehlen wir Ihnen, auf Ihrem Computer einen aktuellen Browser zu installieren. Weitere Informationen finden Sie auf
The content in this site is accessible to any browser or Internet device, however, some graphics will display correctly only in the newer versions of Netscape. To get the most out of our site we suggest you upgrade to a newer browser.