Reactive Control of Autonomous Drones

Abstract

Aerial drones, ground robots, and aquatic rovers enable mobile applications that no other technology can realize with comparable flexibility and costs. In existing platforms, the low-level control enabling a drone's autonomous movement is currently realized in a time-triggered fashion, which simplifies implementations. In contrast, we conceive a notion of reactive control that supersedes the time-triggered approach by leveraging the characteristics of existing control logic and of the hardware it runs on. Using reactive control, control decisions are taken only upon recognizing the need to, based on observed changes in the navigation sensors. As a result, the rate of execution dynamically adapts to the circumstances. Compared to time-triggered control, this allows us to: i) attain more timely control decisions, ii) improve hardware utilization, iii) lessen the need to overprovision control rates. Based on 260+ hours of real-world experiments using three aerial drones, three different control logic, and three hardware platforms, we demonstrate, for example, up to 41% improvements in control accuracy and up to 22% improvements in flight time.

End-to-End Analysis and Design of a Drone Flight Controller

“…In Figure 6, Step 5 is guaranteed to happen before 3 . Thus, the worst case freshness time is essentially the worst case reaction time, shown in Equation 5.…”

“…Kirsch et al [18] use Giotto to reimplement a helicopter control system. Others have developed data-triggered rather than time-triggered drone flight control using reactive programming languages [5].…”

Fundamental concepts of reactive control for autonomous drones

“…The sched- uler operates in a best-effort manner based on programmerprovided priorities. Many autopilots share similar designs [7].…”

Mobile Systems Research with Drones

“…Our experience tells a different story. Several seemingly-solved problems must be revisited in the new context set by drones and the applications they enable, including control [4]. Moreover, it is at this level that one of the most severe issues still plaguing the operation of drones, that is, their dependable behavior, needs to be tackled [11], as we discuss next.…”

“…It made us understand more precisely how the drone reacts to external commands, which facilitated debugging the behavior during autonomous flight; how the environment may affect its operation, for example, in the case of wind; and what signs indicate that physical breakage is imminent, which turns out to be fundamental for a dependable operation in the wild. Some of our research in this area [4] was only possible because of the experience gained in flying drones manually. The more sensors, the better.…”