Networked Quad-Rotor Flying Robots in Multi-Agent Systems

From DRLWiki

We wish to develop distributed algorithms for networked quad-rotor flying robots in multi-agent systems. Such robot teams are useful in a broad range of surveillance, security, and telecommunication applications. We explore high-dimensional (3+) coverage controllers, network update schemes, and vision-based localization methods.

Approach

Our quad-rotor flying robot fleet is composed of six AscTec Hummingbirds. Pitch, roll, and yaw are fully stabilized by the onboard controller described in [1]. Three of the robots are also equipped with an AscTec AutoPilot module, which allows for GPS and altitude position control. We developed an onboard ARM microprocessor module that acquires state estimations from the robot and wireless communicates them to other robots using an xBee module. The same ARM modules run our control algorithms to self-organize in both indoor (motion capture system) and outdoor (GPS, vision localization) environments. Our goal is to realize our researched algorithms on this hardware in fully distributed fashion.

Figure 1. A robot autonomously flying indoors using a motion capture system (photo courtesy of Jason Dorfman, CSAIL Photographer)

Optimal Camera Placement

We developed a coverage algorithm for hovering agents that move in three dimension, but monitor a two dimensional environment. In [2] we focused on the specific scenario of flying or floating robots with downward-facing cameras. We address the question of how to deploy multiple robots so that together their cameras produce the most informative image of the environment. Surveillance and environmental monitoring applications are numerous for such an algorithm. We propose a cost function describing the aggregate information per pixel for the group of robots. Taking the gradient with respect to each robot's position yields a distributed controller for driving the robots to locally optimal positions. Figure 2 shows numerical simulations and robots experiments, respectively, for the algorithm. Refer to the Distributed Control Algorithms for Networked Mobile Robots page for more work in this area.

File:ICRA09Cameras.mp4

Extensions

We have developed an extension of the method to control heterogeneous groups of robotic cameras. In this setting, mixed groups of cameras with arbitrary degrees of freedom can be controlled. For example some cameras may be mounted to a wall at a fixed position on rotating joints, some may be mounted by rotating joints (i.e. gimbals) on flying robots, and some may be fixed on flying robots. The movies below illustrate the control strategy for such heterogeneous groups.

File:FlyingCamerasMovie4down2FixedRotating.mpg

File:FlyingCamerasMovie4atOrigin.mpg

File:FlyingCamerasMovie4atCorners.mpg

The control strategy can also be used off-line to determine optimal positions for cameras that are to mounted statically in an environment.