Robotics and Autonomous Systems (RAS)

RAS Team project 2022 spring semester, University of Turku, Finland.

Check out these cool videos to see autonomous drones flying our code:

• Sumit Gore sumit.n.gore@utu.fi,
• Marius Gurgu marius-mihail.m.gurgu@utu.fi,
• Myrthe Tilleman myrthe.e.tilleman@utu.fi,
• Bowen Tan bowen.b.tan@utu.fi

This repository includes two projects for our RAS course. The first project is a drone project, in which a Tello drone flies through different types of gates. The second project entails an open project. For this project, we use a Tello drone and a Jetbot to navigate the Jetbot to a predefined goal. We use the drone to detect obstacles and the Optitrack system to locate both robots.

Instructions on how to run the code for the drone project

Run move_drone.py.

Instructions on how to run the code for the open project

Notes

We ran this code on two separate machines, one that ran the Tello code and one that ran Jetbot code and path tracking algorithm. They were both connected to the same network.

Pre-requisites

There are many things that are required before being able to run our code. We assume in these instructions that if they are not described, you know how to find, install and run the following:

• ROS Noetic
• ROS Foxy
• Optitrack System
• ROS2-to-ROS1 bridge for the Optitrack System
• Jetbot running a motor driver to control the wheels + 4 Optitrack markers installed
• DJI Tello running a Tello Driver to be able to fly the drone, with the camera facing downwards + 4 Optitrack markers installed
• A obstacle course with ArUco Markers (101, 102, 0 and 3 for bottom left corner, top right corner, goal and obstacles respectively)

Step 1

Before being able to run the code, make sure that the Optitrack system is published correctly on ROS2 topics.
See ros_bridge_instructions.txt

Step 2

ssh into the Jetbot, make sure that there is a script / driver / code available on the Jetbot that subscribes to /cmd_vel and can move the robot. Run that code.

Step 3

Run path_tracker.py

Step 4

Place the drone in the middle of the obstacle course and run the Tello Driver and launch tello_mapping.py

Step 5

If everything ran accordingly, you should see the drone fly up to about 2m, hover, take a picture and publishing it to /tello_map_image. Then the Jetbot should move (it might need a tiny push at the beginning due to the weakness of the motors) according to the path that has been calculated. You can find the plotted path saved in Path.png