Autonomous Search & Rescue with AR.Drone Swarms

Multi-Robot Coordination & Target Detection
Supélec Smart Room, Metz, France | Aug 2011 – Dec 2011

The Challenge

Design and implement a coordinated multi-drone system capable of autonomously searching large areas, detecting lost objects, and managing real-world operational constraints like battery depletion—laying the groundwork for search and rescue operations.

What I Built

Working with Parrot AR.Drones, I developed a complete autonomous search system with the following capabilities:

Network & Communication:

• Established TCP/UDP connections for motion commands , navigation data , and video streaming

• Created GUI for real-time drone control and status monitoring

• Implemented multi-drone coordination across up to three robots

Search Operations:

• Designed alternating linear search patterns with user-defined lengths and speeds

• Programmed autonomous target detection based on image analysis

• Built interrupt logic to halt search when target was detected and report location data

Battery Management:

• Developed autonomous return-to-base functionality when battery levels dropped

• Created swap methodology to replace low-battery drones with fresh ones

• Enabled continuous search operations despite individual drone power constraints

Technical Implementation:

• Predefined flight path system for coordinated area coverage

• Sensor feedback integration from navigation data

• Real-time state monitoring across multiple drone units

Outcomes & Challenges

Successfully demonstrated autonomous multi-drone coordination at . Navigation data streaming worked at Supélec but encountered connectivity issues at Georgia Tech Lorraine. Video data integration ran out of time but framework was established.

Team: Laura Vogelaar, Jasmine Lawrence
Location: Georgia Tech Lorraine x Supélec Exchange Program