SafeSkies: Development of an Autonomous System for Detecting and Tracking Unmanned Hot-Air Balloons and Other Aviation Hazards in High-Risk Areas Using Computer Vision
ISEF · 2026 Embedded Systems
Overview
The release of unmanned hot-air balloons is a Brazilian cultural tradition that, despite being illegal, poses environmental and socioeconomic threats to the country. Balloons are a major cause of wildfires in protected areas, as they use fire for propulsion and can ignite fires upon landing. Additionally, these balloons pose a risk to national airspace since they cannot be detected by radar and may collide with aircraft. In light of this scenario, this project aims to detect and track balloons in high-risk areas, such as airports and environmental conservation units, by installing on-site cameras integrated with a YOLOv10 computer vision model. The system was trained on 6,073 images of balloons, birds, and airplanes, identifying these objects in images and videos with 94% precision and 92% recall. Next, a prototype consisting of two cameras mounted on pan-tilt units was designed, and a PID-based tracking system was developed to calculate optimal adjustments for each camera. Triangulation was then used to derive an equation for the object's position from the distances between the two cameras and their horizontal and vertical angles. Finally, balloon trajectory predictions were implemented using real-time meteorological databases, the Lyapunov exponent was used to estimate predictability horizons, and an automatic alert system was built to notify authorities upon detection. Current results demonstrate accurate detection and real-time tracking in controlled tests, and indicate that the system could be expanded to identify and track other aviation hazards, such as birds and drones. However, further testing will be conducted to evaluate on-site performance.
Awards (2)
- Fourth Award of $600 $600
- Association for Computing Machinery: Fourth Award of $500 $500
Competition history
- ISEF 2026
Resources
Related projects
ISEF · 2018
Improving Aviation Safety Using Low-Cost Low-Fidelity Sensors Augmented with Extended Kalman Filters to Develop an Accurate 3D Dynamic Sense-and-Track System
ISEF · 2017
Drone Defense System: Detection, Tracking, Classification and Targeting of Flight Objects in 3D and Real Time
ISEF · 2016
Using Optical Flow Modeling Methods and Sensor Fusion to Create a Low-Cost, Competent Autonomous Emergency First Responder
ISEF · 2016
Safecopter: Developing a Collision Avoidance System Based on an Array of Time-of-Flight 3D Cameras
Closest projects by meaning, across every fair and year in the corpus.
Source: Regeneron International Science and Engineering Fair