OpenUAV Playbook example (leader-follower)

The video shows an example with 6 UAVs flying in a leader-follower configuration.

The folder structure for the template simulation payload is shown below. Any simulation needs six inputs listed below.

• drone models
• simulation world
• setup (initialization)
• measure (of performance)
• controllers
• parameters

leader-follower/
└── simulation
    ├── inputs
    │   ├── controllers
    │   │   ├── test_1_Loop.py
    │   │   └── test_2_Follow.py
    │   ├── measures
    │   │   └── measureInterRobotDistance.py
    │   ├── models
    │   │   └── f450-1
    │   │       ├── f450-1.sdf
    │   │       └── model.config
    │   ├── parameters
    │   │   └── swarm.sh
    │   ├── setup
    │   │   ├── testArmAll.py
    │   │   └── testCreateUAVSwarm.py
    │   └── world
    │       └── empty.world
    └── run_this.sh

Shown below is an example simulation from remote Ubuntu box with the SSH keys to talk to the OpenUAV Playground server. The username is 'jdas', replace with your own username when trying to run on your machine.

:~$ name=openuav-${3:-`date +%s`}-$(uuidgen) 
:~$ scp -r /home/nsf/PycharmProjects/Open-UAV/examples/leader-follower/simulation/ [email protected]:~/"simulation-"$name
:~$ ssh [email protected] "nvidia-docker run -dit -v ~/\"simulation-\"$name/:/simulation --name $name openuav-swarm-functional"

You can combine the above three commands.

:~$ name=openuav-$(uuidgen)-${3:-`date +%s`} && echo $name && scp -r /home/nsf/PycharmProjects/Open-UAV/examples/leader-follower/simulation/ [email protected]:~/"simulation-"$name && ssh [email protected] "nvidia-docker run -dit -v ~/\"simulation-\"$name/:/simulation --name $name openuav-swarm-functional"

The logs are stored in the simulation/outputs folder as a list of CSV files.

jdas@airborne:~$ ls  ~/simulation/outputs/
measure.csv  uav1.csv  uav2.csv  uav3.csv

Running your first simulation

1. Talk to the project administrators for an account on the OpenUAV server and set up SSH keys