Demos
Simulations, hardware testbeds, and demonstrations from across the project. Videos play right here on the page.
UC Irvine’s demonstration of FlyTrap, the first security analysis of camera-based autonomous target tracking in drones. A specially patterned umbrella draws a tracking drone dangerously close. Evaluated on commercial drones and presented at NDSS 2026.
Videos
Overhead drone video of a delivery pad. GWU’s model tracks the people and objects around it, forecasts how long the zone will stay clear, and decides whether to lower the package, wait, or treat the zone as occupied.
A drone descends to a residential drop-off in TSU’s IRIS simulator, one of the scenarios behind the IRIS D³ drone-delivery dataset.
UC Irvine’s demonstration of FlyTrap, the first security analysis of camera-based autonomous target tracking in drones. A specially patterned umbrella draws a tracking drone dangerously close. Evaluated on commercial drones and presented at NDSS 2026.
A real-world capture of marker-based pose estimation, comparing the estimate from a clean marker with one under the MirageMarker perturbation.
Subtle geometric changes to a fiducial landing-marker board shift a PX4 drone’s precision landing 8–12 m off the pad in NVIDIA Isaac Sim, while clean boards land within centimeters. Presented at IROS 2026.
BlueSky simulation of drones from two delivery companies over Frisco, Texas, each fleet flying its own attention-based multi-agent reinforcement-learning deconfliction policy.
A fine-tuned large language model issues tactical deconfliction actions to drones in closed-loop BlueSky simulation.
Three simulated drones in Gazebo maneuver around one another using their neighbors' broadcast positions. This is the cooperative behavior that Remote ID spoofing targets, and that spoofing-aware planning protects.
A walkthrough of GridEye, the read-only traffic-management dashboard built by NVCC interns. It tracks a 12-drone delivery fleet over simulated Frisco, Texas, and flags flight-boundary, no-fly-zone, and altitude violations.
Dozens of simulated vehicles launched in a single experiment, showing how the simulator scales for cyberattack studies.
Simulated drones, each linked to its own operator’s ground control station, as in the multi-operator setting of the FAA UTM concept of operations.
Vanderbilt’s networked simulator flies many drones from multiple operators over a downtown map, each running its own ArduPilot autopilot. It’s the platform used to generate attack datasets and evaluate detectors.
A tour of TSU’s IRIS physics-based simulation, with downtown, suburban, and airport scenes and several eVTOL aircraft flying coordinated routes.
IRIS scenes rendered as RGB, thermal-infrared, night-vision, and LiDAR imagery, and under fog and snow. These are the conditions our perception methods are tested in.
Candidate safe landing zones are detected in aerial imagery and compared by size and clearance before one is selected.
TSU’s emergency landing system locks onto the selected landing zone and keeps it centered as the aircraft descends, with live tracking plots.
A UAV adjusts its trajectory and antenna beam pattern to jam an eavesdropper while keeping a communication null toward a friendly client (ACC 2026).
Spring 2027
A project-wide showcase of integrated demonstrations across our aviation use cases, for NASA, industry, government, and the research community.
Tools & datasets
Simulators, datasets, and software that power our experiments and demonstrations.
Read-only traffic-management dashboard that tracks a simulated delivery fleet and enforces flight boundaries, no-fly zones, and altitude limits.
Physics-based urban air mobility simulation with downtown, suburban, vertiport, and airport scenes, eVTOL and drone models, and simulated RGB, infrared, LiDAR, GPS, and inertial sensors under degraded conditions.
More than 20,000 synthetic RGB images with matching LiDAR point clouds across delivery scenarios, from clean pads to intrusions, occlusion, and degraded visibility.
Distributed multi-UAV simulator integrating ArduPilot SITL, Gazebo, QGroundControl, and OMNeT++, modeling multi-operator operations under the FAA UTM concept of operations, with a library of adversarial scenarios and demonstration notebooks.
Decentralized planning that localizes a suspected Remote ID spoofer from signal-strength measurements and avoids a chance-constrained unsafe region.