A quadcopter swarm for active monitoring of smog propagation

Cyber-physical system (CPS) features the interaction between embedded systems and the physical world. In this paper, we focus on a CPS using unmanned quadcopters to interact with the environment to monitor smog propagation. Our quadcopters shall be equipped with various sensors including smog sensor, wind speed sensor, and GPS; and shall be able to fly omni-directionally in open air space. Compared to ground sensing, our approach shall enable 3D monitoring of smog propagation. Compared to conventional balloon based aerial sensing, the vast number of cheap quadcopters and high mobile speed shall enable better monitoring granularity and adaptability. In our preliminary study [1], we start with modeling and optimizing the control of a single quadcopter, taking into consideration partial differential equations governing the smog propagation, and nonlinear dynamic model of the quadcopter. The optimized single quadcopter control strategy is then extended for multiple quadcopters, under an additional formation control constraint. The strategy is validated with simulation. Effort is now underway to implement the preliminary study strategy in Arduino based quadcopters. Through this study, we aim not only to realize a novel CPS application, but also to build a experiment test-bed, to explore insights on how to exploit prior-knowledge to optimize interactions between embedded robotic systems and physical world in CPS.