Scientists in Europe recently announced a breakthrough in the development of drones capable of swarming autonomously much like how birds fly in flock formation.
Funded by the European Research Council, the Eötvös Loránd University, and the Hungarian Academy of Science, the EU ERC COLLMOT project, led by Professor Tamás Vicsek, is a five year study on the complex structure and dynamics of collective motion.
The research team initially studied the flight patterns of pigeons fitted with miniature GPS tracking devices. Scientists then used these results to generate computer animated models showing the individual data points of the pigeons in motion.
The COLLMOT robotic research center now boasts a fleet of self-adjusting and self-stabilizing quadcopters, aircraft that have four rotors. These quadcopters are also equipped with GPS receivers, and can navigate and communicate their flight plans and positions among themselves via radio signals.
According to Tamás Vicsek, the team’s quadcopters are unique in that previous attempts at replicating collective flight behavior required direct input, predefined routes, and controlled environments. In contrast, the COLLMOT quadcopters are the first drones to fly as a coordinated flock without the need for continuous commands from a central computer. They can also reorder and rearrange their formations autonomously in response to updated external inputs.
Vicsek’s team based their algorithms in part on a computer program created by Craig Reynolds in 1986 called Boids. Currently a computer graphics expert at the University of California in Santa Cruz, Reynold’s Boids program simulates virtual flying objects that move according to three rules: alignment, attraction, and repulsion. Put simply, these three rules help Vicsek’s quadcopters maintain an average distance from each other as they assume a particular formation toward a decided direction without crowding and colliding into each other.
In time, Vicsek hopes to equip cameras on these quadcopters so they can visualize each other, thereby reducing their reliance on radio and GPS for communication.
Although it’s frightening to predict how these fleets of flocking drones will be used in various military and surveillance settings, Viscek’s team remains hopeful (and yeah, so do we) that these drones will be used to help emergency rescue teams find patients, and farmers to help irrigate their lands and monitor their animals.
Interview with Tamás Vicsek:
TGM: What key insight has your team learned about collective bird behavior?
TV: The key new insight is that birds are very much likely to use information beyond the instantaneous position and velocities of their neighbors during highly synchronized group flights. It is likely that they make use of a number of “pre-programmed” flight patterns, quite like dancers.
TGM: Besides agriculture, emergency rescue, and monitoring, what other applications can drone swarming technology be used for?
TV: Monitoring seems to be the most important. It has many facets, including the collection of detailed data on the status of the fields. We also think that video broadcasting of extended sporting events (e.g., bicycle races) could be an application.
Here’s another video providing more information about the project’s development: