Capstone Net Launcher

Capstone Net Launcher

Senior Capstone | Georgia Tech | 2026

← Back to Projects
Mechanical Design Electronics & Wiring System Integration

Overview

For Interdisciplinary Senior Design at Georgia Tech, I collaborated with a team of six to develop a low cost, kinetic, counter UAS system for the Defense Innovation Unit (DIU). Our team created an autonomous, vehicle mounted turret that uses computer vision to detect and track a drone, then a pneumatic net launcher to neutralize it.

The Problem

Small, high speed, and highly manuverable kamikaze drones are increasingly targeting armored vehicles, risking lives and military missions. Current counter UAS solutions often rely on RF jamming, which is ineffective against drones using fiber optic communicaiton, or are expensive to deploy at a large scale such as missile systems. Lower cost solutions often require manual operation which is not feasible for personnel inside an armored vehicle.

Design

Our final design is a vehicle-mounted, autonomous, net-launching turret which can track and intercept a rogue drone before it reaches an armored vehicle. The design uses computer vision to identify and track a drone up to 22 feet away. After the target is identified, the turret aims towards it and launches a net using compressed air, which has an effective range of over 7 feet away. The net entangles itself around the propellers of the enemy drone and halts its forward momentum, grounding it before it can get within the effective range of its explosive payload.

Detection:

  • Trained a YOLOv8n computer vision model on ~25,000 images
  • Model runs onboard a Jetson Orin Nano
  • Optimization of video resolution and use of multithreading to minimize latency and maximize range

Aiming:

  • Jetson sends the position of the drone to an STM32 through UART communication
  • STM32 controls pitch through 90° and roll through 300° using a PID controller
  • Implemented limit switches to prevent motor over-torquing

Firing:

  • Solenoid valve releases compressed air from a pressurized tank
  • Air redirected through a custom 3D-printed barrel adapter
  • Weights attached to the net are propelled forward to intercept drone

Results

The project successfully demonstrated that it could be a robust and reliable counter UAS solution with successful field testing. The field testing yielded 100% capture rate for hovering drones, 75% capture rate for targets in lateral or approaching motion, and 100% successful net deployment. The field testing proved a 22ft detection range and a 10-24ft horizontal grounding range.

Key technical takeaways included:

  • Latency Management: Minimizing frame to motor delay is more critical than maximizing image resolution with a fast moving target.
  • Environmental Robustness: Testing revealed that varying light and background conditions required a more diverse training dataset for the YOLOv8n model, leading us to expand our dataset.
  • Control Optimization: Refining the PID controller must account for the physical inertia of the loaded turret to prevent oscillation while maintaining speed.

Project Poster

Capstone Net Launcher Project Poster

Demo Video

Next Project →