Chung receives CAREER Award to develop robotic birds of prey
Avian and other wildlife strikes annually cause more than $715 million in damage to aircraft each year, estimated the Federal Aviation Administration in a 2011 Wall Street Journal article. Consider the dramatic 2009 water landing of U.S. Airways Flight 1549 in the Hudson River, after a flock of geese collided with the plane during its climb out.
But CSL Assistant Professor Soon-Jo Chung has cried fowl, so to speak. He is working to develop robotic birds of prey that could chase flocks away from airfields, where birds are most likely to cause damage. The National Science Foundation will fund the research at nearly $500,000 for five years through its CAREER Award program.
“Robotic falcons could be an efficient and cost-effective solution, but will require significant advancements in control and sensing,” said Chung, an assistant professor of aerospace engineering at Illinois.
Real birds of prey and guns have proved to be the most successful methods for removing flocks. But both come with significant challenges. While falcons were successfully deployed at J.F. Kennedy Airport and McGuire Air Force Base, for example, real birds are difficult to control and train. As they require human handlers in such cases, they are also expensive to maintain. Further, the most effective performers -- peregrine falcons -- are an endangered species.
Meanwhile, the use of guns has outraged animal rights groups, which successfully petitioned Kennedy – situated near a bird sanctuary – to quit using arms in 1993.
By creating a robotic falcon that can sense flocks and outfly them, Chung believes he can introduce a viable, though certainly not simple, solution.
“This is basically a grand control challenge problem,” said Chung, who is a senior member of the American Institute of Aeronautics and Astronautics (AIAA) and the Institute of Electrical and Electronics Engineers (IEEE). “The dynamics are so complicated due to the complex nonlinear flapping flight dynamics with many articulated wing joints. Then we should take into account the wing flexibility, which is difficult to model and control; flexible wing models are written in the Partial Diffierential Equations (PDEs).”
Chung, with his students, has been working on both the dynamic modeling and control challenges of bird-scale flapping flight. The team derived a limit-cycle-based control formulation for flapping flight while establishing PDE boundary control strategies for flexible, articulated-winged aircraft. “There are still some significant issues in flight control that must be resolved for them to work in the real world,” he said.
Researchers also must develop algorithms that enable the robotic falcons to identify targets, and then navigate and herd the birds away from the airfield. A novel aspect of the project will focus on multi-agent pursuit-evasion algorithms that will help enable the robotic falcons to chase and navigate the birds away from the airfields.
“Birds are smart and can distinguish real falcons from robots,” Chung said. “Our robots must fly like real falcons, look like real falcons and even sound like real falcons.” Chung intends to leverage his prior work on distributed control, real-time optimization and synchronization of multi-vehicle systems, as well as game-theoretic or geometric formulations of pursuit-evasion.
In addition to the potential of solving an expensive and dangerous real-world problem, Chung is also excited about the opportunity to contribute to the fundamental understanding of avian flight. He said: “At its core, this is a scientific exploration of how birds fly so well.”