Sensor Architecture

At FlyScope, we develop modular sensor systems designed specifically for autonomous drones operating in urban and industrial environments. Our sensors combine precision optics, AI acceleration, and real-time connectivity — all optimized for infrastructure inspection, cleaning, and servicing missions.

Modular Design

Each sensor module is built as a self-contained unit, combining:

• RGB and IR cameras for visual and thermal imagery
• Jetson Nano/Xavier or Qualcomm RB5 modules for on-board AI inference
• IMU/GPS synchronization for geotagged data
• On-device SSD storage and real-time streaming (RTSP/USB3)

This allows deployment on a range of drone platforms (DJI, PX4, custom UAVs).

Sensor Stack Layers

1. Optical Layer
   High-resolution RGB sensors (up to 4K) and optional IR/thermal imaging. Calibrated for outdoor lighting and anti-glare conditions.
2. Processing Layer
   Edge AI module (e.g., NVIDIA Jetson) runs deep learning models (YOLOv8, SegFormer) directly on-device. Enables fast segmentation, classification, and anomaly detection.
3. Synchronization Layer
   Integrates GNSS, IMU, barometer, and altimeter sensors to enable precise location tagging and pose estimation.
4. Interface Layer
   Supports USB3, RTSP, UART, and LTE/5G uplinks. Also includes GPIO for actuator control (e.g., sprayers, nozzles, paint units).
5. Thermal & Structural Enclosure
   Lightweight, water-resistant, shock-absorbing housing suitable for outdoor flight and extended exposure to vibration and heat.

Connectivity & Control

Sensor data is:

• Processed on-board in real time
• Streamed to a base station or cloud backend via LTE/5G or Wi-Fi
• Integrated with FlyScope Cloud via MQTT or REST APIs
• Stored locally for redundancy and post-flight analytics

Customization

Our modular architecture allows:

• Swapping camera types (RGB/IR/Multispectral)
• Upgrading compute units (Nano ↔ Xavier ↔ Orin)
• Adding extra ports for cleaning modules or visual lighting systems