Advanced GPS-denied navigation with real-time object detection powered by Pixhawk, Intel RealSense, and edge AI processing.
Explore TechnologyPrecise indoor/outdoor navigation using Intel RealSense D435i depth vision and PX4 VIO with visual-inertial odometry.
VINS-Fusion ORB-SLAM3Robust positioning through sensor fusion of UWB, TF-Mini-S LiDAR, IMU, and optical flow data.
MAVROS EKF2Real-time obstacle identification with YOLOv8 running on Raspberry Pi 5 + Coral TPU accelerator.
TensorFlow Lite OpenCV| Component | Model | Key Features |
|---|---|---|
| Flight Controller | Pixhawk 2.4.8 | PX4 firmware, EKF2 sensor fusion, MAVLink protocol |
| Depth Camera | Intel RealSense D435i | 1280×720 @30fps, integrated IMU, global shutter |
| LiDAR Altimeter | TF-Mini-S | 12m range, 100Hz update rate, ±10cm accuracy |
| Edge Computer | Raspberry Pi 5 (8GB) | Coral USB TPU for AI acceleration, ROS 2 Humble |
| Power System | 3S LiPo 5000mAh | High-C rating (50C), smart voltage monitoring |
Autonomous exploration of collapsed buildings with victim detection.
GPS-denied infrastructure assessment with crack/defect identification.
Indoor greenhouse monitoring with plant health analysis.
Our SLAM drone platform is available for research partnerships and commercial deployments.
Request Technical DetailsPX4 Flight Stack with MAVROS bridge to ROS 2
PX4 v1.14 ROS 2 HumbleVINS-Fusion with UWB integration
VIO EKF2YOLOv8s with TensorFlow Lite for edge deployment
OpenCV PyTorch