Video Shows Progress of Swarm Spacecraft Research
In the latest video Soon-Jo Chung’s group has produced from its research into swarms of autonomous flying spacecraft and unmanned aerial vehicles, aerial drones react to a student’s arm movements like an orchestra to a director’s prompts.
In the video, reminiscent of the famous scene in the movie, Minority Report, the student wears gloves and/or a hat equipped with reflective markers the small quadcopters sense and to which they respond. The student moves one gloved arm and all 10 drones react, moving up and down or side to side. When the student moves his left and right arms separately, two lines of five drones respond to their corresponding arm’s direction, independently of the other line. In another option, the automatically controlled quadrotors move as a unit to avoid the student as he walks through the spacecraft collection.
The testbed employs a Vicon Motion Capture system to detect the locations of each individual quadrotor through the reflective markers, which work as an indoor GPS system. Additionally, the gloves and hat used in the video are similarly marked to determine the person’s location. Each quadrotor uses only the location of the leader (glove or cap), itself, and neighboring quadrotors to decide where and how to move.
The research is aimed at developing intelligent formation flying of swarms of autonomous flying robots such as aerial drones and spacecraft, according to Chung, assistant professor in Aerospace Engineering at Illinois. Researchers in Chung’s Aerospace Robotics and Control Laboratory have been working with a team at the NASA Jet Propulsion Laboratory, led by Dr. Fred Hadaegh, to develop spacecraft swarms.
Chung’s students, Giri Prashant Subramanian, Samip Shah, Daniel J. Morgan and Saptarshi Bandyopadhyay, produced the video. Subramanian developed the human avoider algorithm, Shah developed the formation avoider algorithm, and, together, they implemented and tested all remaining algorithms. Morgan and Bandyopadhyay set up the testbed and assisted with algorithm development.
Spacecraft swarms will eventually be made up of hundreds to thousands of very small spacecraft that are about the size of an iPhone. A system engineering paper that details the design of swarms of femtosatellites can be found here. Challenges that exist when flying swarms of drones, such as collision avoidance with neighboring agents, obstacles or even humans; and limited capabilities of a small drone, also exist in spacecraft swarms. The video also addresses the challenge of how a single human operator can control a large number of drones by implementing an intelligent cooperative control algorithm.
There are many benefits using swarms as opposed to traditional monolithic systems. It can improve the safety, adaptability, scalability, evolvability, maintainability and robustness of the entire system. By using so many identical drones, the cost of each drone will greatly decrease and the cost of the mission will decrease. Another advantage of flying so many drones is that the loss of a single drone is much less significant.
According to Chung, the purpose of the formation flying testbed is to demonstrate the effectiveness of the cooperative control and swarm guidance algorithms (example) the group has developed for future autonomous swarms of spacecraft or aerial drones. Even though the drones’ motions are different from a spacecraft’s motions, the testbed is a means to verify the group’s algorithms in a real-world setting, providing information that cannot be derived from computer simulations.
The testbed also allows Chung’s group to demonstrate potential uses of the swarms of drones, including avoidance. It is a particular challenge for spacecraft to avoid colliding with asteroids or space debris while maintaining a formation design.