Taking the Plunge: Intro to Drones & AI

Welcome to the official course repository for IDEAL Fall 2026 – Taking the Plunge: Intro to Drones & AI.

This course provides an engaging introduction to drone technology, foundational programming, and how artificial intelligence enables drones to interact intelligently with their environment. Through hands-on projects and guided modules, students will explore how drones can be programmed to navigate, recognize objects, and perform simple autonomous tasks using AI techniques.

Course Goals

• Understand the basic principles of drone flight and operation.
• Learn fundamental programming concepts using Python.
• Explore how AI enables drones to detect and respond to their surroundings.
• Apply computer vision techniques for gesture control, object detection, marker tracking, and other applications.

Repository Structure

The folders are numbered and listed in the order we will use them in class.

Foundations

Folder	What you will do
00_Getting_Started	Learn how to write Python programs and use Python to communicate with a DJI Tello drone.
01_Getting_Started_AI	Learn to use an AI assistant to explain, modify, debug, and review your drone programs.

Computer Vision Projects

Every project follows the same idea: the drone's camera sees something → the program recognizes it → the drone reacts. Each one uses a smarter way of recognizing.

Folder	What the drone recognizes	How it recognizes it
Tello_Aruco	Printed ArUco markers (flip or land)	Pattern matching with OpenCV
Tello_QRCode	Printed QR codes with text commands	QR decoding with OpenCV
Tello_YOLO	Everyday objects (people, chairs, bottles...) while flying a square	YOLO neural network (AI)
Tello_GestureCTRL	Your hand gestures, to fly the drone	MediaPipe hand tracking (AI)
Tello_PoseCTRL	Your arm poses, to fly the drone	MediaPipe body tracking (AI)

Each folder has its own README with setup steps, the commands it understands, an explanation of the code, "Try This" exercises, and troubleshooting tips.

Every project also has a no-fly test (test_..._no_fly.py). It checks the packages, the AI model or detector, the drone connection, the battery, and the video, then runs a live rehearsal with the same threads as the flying program while printing the commands it would send. The drone never takes off. Run it before every flight; it ends with READY TO FLY or tells you exactly what to fix.

Project	No-fly test	Flying program
Tello_Aruco	test_Aruco_no_fly.py	test_Aruco.py
Tello_QRCode	test_QRCode_no_fly.py	test_QRCode.py
Tello_YOLO	test_YOLO_no_fly.py	test_sim_cam_sqv1_YOLO.py
Tello_GestureCTRL	test_gesture_no_fly.py	tello_gesture_control.py
Tello_PoseCTRL	test_pose_no_fly.py	tello_pose_control.py

All flying programs use the same structure: one thread reads the camera, one thread sends drone commands, and the main program runs the detector and shows the video. Drone commands such as flip_forward() or move_forward(100) take a few seconds to finish, so keeping them in their own thread means the video never freezes while the drone moves.

The file test_camera.py is a quick check that your computer can connect to the Tello and show its video. Run it whenever you are not sure the connection works.

Development Environment (Recommended Setup)

• Python 3.12 (Python 3.10 and 3.11 also work). Very new Python versions may not be supported yet by MediaPipe, which the gesture and pose projects need.
• One virtual environment (.venv) for the whole course, created in the repository folder. 00_Getting_Started walks you through creating it in PyCharm.
• PyCharm Community Edition (optional but recommended for easier debugging and project management).

💡 Students can use any code editor, but PyCharm Community Edition provides built-in support for virtual environments and is recommended for this course. Download here: https://www.jetbrains.com/pycharm/download

Hardware Requirements

• DJI Tello (Combo) – This is the primary drone platform used throughout the course.
• Optional: Webcam – The gesture and pose projects can use your laptop webcam for practice without flying. Other modules do not need it.

Getting Started

To participate in the projects, you will need:

• A laptop with Python 3.12 installed
• PyCharm Community Edition IDE
• A DJI Tello drone
• Optional: Webcam (for practicing gestures and poses)

Each project folder has a requirements.txt listing the packages it needs. With your course .venv active, you can install the packages for all projects at once from the repository root:

pip install -r 00_Getting_Started/requirements.txt -r Tello_Aruco/requirements.txt -r Tello_QRCode/requirements.txt -r Tello_YOLO/requirements.txt -r Tello_GestureCTRL/requirements.txt -r Tello_PoseCTRL/requirements.txt

This is a large download (mostly PyTorch, which YOLO uses), so do it at home or on a fast connection, before class.

Important: The Tello Wi‑Fi Has No Internet

When your laptop is connected to the drone's Wi‑Fi (TELLO-XXXXXX), it cannot reach the internet. Anything that needs to be downloaded must be downloaded before you connect to the drone:

Project	Download before class	How
All	Python packages	The pip install command above
Tello_YOLO	yolov8n.pt model	See Tello_YOLO → Setup
Tello_GestureCTRL	hand_landmarker.task model	Run python gesture_controller.py once in that folder
Tello_PoseCTRL	pose_landmarker_lite.task model	Run python pose_controller.py once in that folder

Flight Safety Rules

1. Run the project's no-fly test first, and fly only after it says READY TO FLY.
2. Fly only in the area your instructor has cleared. Keep at least 2 m of empty space around the drone.
3. Many programs take off automatically when they start. Look at the drone, not the screen, when you press Run.
4. Keep one hand near the Q key. In every project, Q lands the drone.
5. Never try to catch a flying drone.
6. Check the battery before flying. Flips need more than 50%.
7. If something goes wrong and Q does not work, press the power button on the drone.

Troubleshooting (All Projects)

Problem	Likely cause and fix
ModuleNotFoundError: No module named ...	The package is not installed in the environment PyCharm is using. Check that the interpreter is your .venv (see 00_Getting_Started), then pip install -r <project>/requirements.txt.
Program cannot connect to the Tello	Make sure your laptop is on the TELLO-XXXXXX Wi‑Fi, the drone is on, and no other program (or the Tello phone app) is connected to it.
Video colors look wrong (blue and red swapped)	The Tello library delivers frames in RGB order, while OpenCV expects BGR. The project code converts each frame with cv2.cvtColor(frame, cv2.COLOR_RGB2BGR).
import cv2 fails after installing MediaPipe or YOLO	Two OpenCV packages are fighting. Run pip uninstall -y opencv-python opencv-contrib-python, then pip install opencv-contrib-python.
The drone lands by itself	The battery is low, or the drone received no command for 15 seconds (a built-in Tello safety feature).

License

This project is intended for educational use and classroom exploration. Feel free to fork and adapt under the terms of the MIT License.

---

Stay tuned for updates and new modules throughout the semester!