Aerial Robotics IITK Hackathon 2026
Team Aerial Robotics IITK | Y25 Recruitment Hackathon 2026
Hey Y25s! Hope you are enjoying your time on the campus. These are the problem statements of the Aerial Robotics Hackathon 2026 . They are divided into three levels and a Bonus level . The submission guidelines for each level are mentioned clearly — please follow them correctly. Any doubts can be raised on GitHub in the Discussions tab of this repository. You can find the problem statements and required starter kits in this repo.
Level
Title
Description
Level 1
Operation SKYE-X: Search & Pursuit
Explore the map, find an evasive target, and track it under sensor limitations
Level 2
Operation Touchdown: Precision Landing
A harder variant with new constraints — unlocked after Level 1
Level 3
Project MIRAGE: Hybrid Marker Steganography
Create a single image that shows a decoy at high altitude and a secret ArUco marker at low altitude
Bonus
Operation FREEFALL: Autonomous Precision Landing
Use a 3D physics simulator, ArUco detection, and PID control to land a drone on static and moving targets
Quick Install (All Levels)
pip install -r requirements.txt
Level
Packages
Level 1 & 2
pygame
Level 3
opencv-contrib-python, numpy
Bonus
opencv-contrib-python, numpy, requests, simple-pid, pyautogui
All submissions must be combined and submitted in a .zip format.
The official submission form will be floated soon.
Please read all the Problem Statements (PS) carefully and ensure you understand the deliverables for each level before submitting.
Expected Folder Structure
Your final submission ZIP file should extract strictly to the following folder structure:
Submission_TeamName/
├── Level1/
│ ├── skye_controller.py
│ ├── score_proof.png
│ └── writeup.pdf (or .md)
├── Level2/
│ └── solver.py
├── Level3/
│ ├── generator.py
│ ├── hybrid_marker.png
│ ├── viewer.py
│ ├── visual_check.py
│ ├── drone_view_2.0m.png
│ ├── drone_view_30.0m.png
│ └── report.pdf
└── Bonus/
├── auto_land.py
├── landing_video.mp4
└── report.pdf
Only modify skye_controller.py — do not modify skye_env.py
You may adjust config.py parameters (e.g. MAX_TIMESTEPS) for tuning/testing
Submit your final skye_controller.py, score_proof.png, and a short write-up explaining your approach
-Submit solver.py only. Must run with the original unmodified simulator_level2.py in the same directory:
Submit a GitHub repository with: generator.py, hybrid_marker.png, viewer.py, visual_check.py, drone_view_2.0m.png, drone_view_30.0m.png, and report.pdf
viewer.py and visual_check.py must be built from scratch — they are not provided
report.pdf (max 2 pages) must include σ values, mathematical reasoning, and GSD calculation
Submit auto_land.py, landing_video.mp4, and report.pdf
auto_land.py must connect to the simulator (ports 8080/8081) and use ArUco detection — no hardcoded coordinates
report.pdf (max 2 pages) must include PID tuning approach, control loop diagram, and ArUco pipeline explanation
Level 1 — Search & Pursuit
Metric
How It's Measured
Tracking Score
+1 per timestep the target is within TRACKING_RADIUS (70px)
Survival
Score is forfeited on crash
Mission Duration
Max 3000 timesteps
Level 2 — Precision Landing
Metric
How It's Measured
Landing Accuracy
Distance (m) from platform center at touchdown
Success Threshold
≤ 0.35 m from center = SUCCESS
Precision Bonus
< 0.1 m = exceptional
Criterion
Weightage
How It's Measured
ArUco ID_X detected at ≤ 5m
40%
Image shrunk to 5m pixel size, ArUco detection run
ArUco ID_X hidden at ≥ 30m
30%
Image shrunk to 30m pixel size, ArUco detection run
Visual quality — no ghosting
15%
Judging script at 30m; visual inspection for ArUco bleed-through
Math justification in report
15%
σ values backed by GSD/Nyquist calculations, not trial and error
Bonus — Operation FREEFALL
Metric
Points
Phase 1 Completion (static landing)
30%
Phase 2 Completion (moving platform)
50%
Landing Smoothness
Up to 20% bonus (lower touchdown velocity → higher score)
Time to Land
Fastest landing breaks ties