Houssam Abbas, Assistant Professor
Electrical Engineering and Computer Science
Oregon State University
Safe planning and control of multi-drone fleets is a challenging problem due to the complexity of the missions and the size of the fleets. Methods that offer guarantees on mission satisfaction generally do not scale well, while more computationally tractable approaches do not offer satisfaction guarantees. In this talk, I will present Fly-by-Logic, a method that overcomes these limitations for a wide variety of multi-drone missions that consist of a combination of the following objectives:
1) spatial objectives, e.g. geofenced no-fly zones, or delivery zones,
2) temporal objectives, e.g. a time window in which to monitor wireless signal strengths in an area,
3) reactive objectives, e.g. in case of a drone failure, another drone picks up its mission.
Fly-by-Logic uses temporal logic to specify a wide variety of missions. Starting from the given mission specification, mission satisfaction is formulated as a non-convex optimization problem, solved to local optimality in a centralized manner. Additional constraints result in trajectories that can be tracked near perfectly by off-the-shelf lower-level controllers. I show the performance, scalability and real-time applicability of Fly-by-Logic method through simulations and experiments on real quadrotor drones. Finally, I will also present on-going work on an extension of this method for solving the mission planning problem in a distributed manner.
Wednesday, January 16, 2019 at 3:00pm to 4:00pm
Kelley Engineering Center, 1005
110 SW Park Terrace, Corvallis, OR 97331