Table 4.5 has the links for the visualizations of all simulations and experiments per- formed in this work.
Table 4.5: Links for the videos for simulations and experiments. Here, Sim. stands for Matlab simulations, CF2.0 for experiments on the Crazyflies. Stop-go and vel. free are the two modes of operation of the trajectory generator, and B (R) is the Boolean (Robust) mode of solving the control problem. Shr. Hrz. stands for the shrinking horizon mode for online control. The reader is advised to make sure while copying the link that special characters are not ignored when pasted in the browser.
Platform Mode Specification Drones (D) Link
Sim. One-shot (B), stop-go ϕmRA 1,2,4,5 http://bit.ly/RABstopgo
Sim. One-shot (R), stop-go ϕmRA 1,2,4,5 http://bit.ly/RARstopgo
Sim. One-shot (B), vel. free ϕmRA 1,2,4,5,6 http://bit.ly/RAB1to6varvel
Sim. One-shot (B), vel. free ϕmRA 8,10,12,16 http://bit.ly/RAB8to16varvel
Sim. One-shot (R), vel. free ϕmRA 1,2,4,5,6 http://bit.ly/RAR1to6varvel
Sim. One-shot (R), vel. free ϕmRA 8,10,12,16 http://bit.ly/RAR8to16varvel
Sim. One-shot (R), vel. free ϕx−mission 2 http://bit.ly/multi2mission
Sim. One-shot (R), vel. free ϕx−mission 4 http://bit.ly/multi4mission
Sim. One-shot (R), vel. free ϕx−mission 6 http://bit.ly/multi6mission
Sim. One-shot (R), vel. free ϕx−mission 8 http://bit.ly/multi8mission
CF2.0 Shr.Hrz (B), vel. free ϕsRA 1 http://bit.ly/varvel1
CF2.0 Shr.Hrz (B), vel. free ϕmRA 2 http://bit.ly/varvel2
CF2.0 Shr.Hrz (B), stop-go ϕsRA 1 http://bit.ly/stopgo1
CF2.0 One-shot (R), vel. free ϕmRA 4 http://bit.ly/varvel4
CF2.0 One-shot (R), vel. free ϕmRA 6 http://bit.ly/varvel6
Chapter 5
The Fly-by-Logic toolchain for
UAV fleet planning
5.1
Introduction
Safe planning for fleets of Unmaned Aircraft Systems (UAS) performing complex mis- sions in urban environments has typically been a challenging problem. In the United States of America, the National Aeronautics and Space Administration (NASA) and the Federal Aviation Administration (FAA) have been studying the regulation of the airspace when multiple such fleets of autonomous UAS share the same airspace, outlined in the Concept of Operations document (ConOps). While the focus is on the infrastructure and management of the airspace, the Unmanned Aircraft System (UAS) Traffic Management (UTM) ConOps also outline a potential airspace reser- vation based system for operation where operators reserve a volume of the airspace for a given time interval to operate in, but it makes clear that the safety (separation from other aircraft, terrain, and other hazards) is a responsibility of the drone fleet operators. This chapter presents a tool that allows an operator to plan out missions for fleets of multi-rotor UAS, performing complex time-bound missions. The tool builds upon the correct-by-construction planning method of chapter 4 by translating missions to Signal Temporal Logic (STL). Along with a simple user interface, it also has fast and scalable mission planning abilities. We demonstrate our tool for one such mission later in the chapter.
It is inevitable that autonomous UAS will be operating in urban airspaces Federal Aviation Authority [2018]. In the near future, operators will increasingly rely on fleets of multiple UAS to perform a wide variety of complicated missions which could consist of a combination of: 1) spatial objectives, e.g. geofenced no fly zones, or delivery zones, 2) temporal objectives, e.g. a time window to deliver a package, 3) reactive objectives, e.g. action when battery is low.
In this chapter, we present a tool1 that allows an operator to specify such require- 1
Graphical User Interface
(MATLAB)
Fly-by-Logic: Library for
maximization of smooth
robustness of STL (C++)
Mission Parameters (YAML) UAS Trajectories (YAML) User Inputs CasADi Optimization formulation IPOPT Optimization solverROS planning and control stack
To UAS
The Fly-by-Logic tool
Figure 5.1: The Fly-by-Logic tool-chain. Through a MATLAB-based graphical interface (figure 5.2), the user defines the workspace and the multi UAS mission. This mission is interpreted as an STL specification (of the form in equation 5.1), the parameters of which are passed from the interface to the Fly-by-Logic C++ library. Through interfacing with off-the-shelf optimization tools, trajectories that satisfy the mission are generated for each UAS and visualized through the user interface. The way-points that generate these trajectories can also be sent to a Robot Operating Systems (ROS) implementation of trajectory following control to be deployed on board actual robots
(e.g. bit.ly/varvel8).
ments over a fleet of UAS operating in a bounded workspace and generates trajectories for all UAS such that they all satisfy their given mission in a safe manner. In order to generate these flights paths, or trajectories, our tool relies on interpreting the mission objectives as Signal Temporal Logic (STL) specifications Maler and Nickovic [2004]. We then formulate the problem of mission satisfaction as that of maximizing a notion ofrobustness of STL specifications Fainekos [2008]. Using the approach of Pant et al. [2018] (see chapter 4), we generate trajectories for all the UAS involved such that they satisfy the given mission objectives.