Research and Development of Unmanned Systems, AI Models, Counter-UAS and Jet Turbines

Research & Development

Research Areas: System Architecture, Artificial Intelligence, Counter-UAS and Jet Turbines

At UXT Systems, platform, algorithms, sensors and engines are developed as one integrated system – from the concept phase to flight testing.

• 01

  Architecture and Integration of Unmanned Aerial and Ground Systems

  The performance of an unmanned system is determined at the interfaces between its components. We therefore design airframe, powertrain, power supply, flight control, sensors and data link together from the outset, implement new concepts on test vehicles and verify them step by step in laboratory, ground and flight tests. A key research focus is robust navigation in environments where GNSS is degraded or unavailable.

  System Architecture · Test Vehicles · PX4 · ArduPilot · GNSS-Denied Navigation · Sensor Fusion

• 02

  Onboard Artificial Intelligence for Unmanned Systems

  Autonomy requires a system to perceive and interpret its environment independently – even when the link to the ground control station is degraded or lost entirely. We develop neural network models for object detection, classification and scene understanding and optimize them for deployment on energy-efficient edge hardware. They are evaluated against defined metrics for detection performance, latency and robustness to real-world, noisy sensor data.

  Computer Vision · Edge Inference · Model Optimization · Sensor Fusion · Autonomy

• 03

  Counter-UAS: Detection, Tracking and Classification of Non-Cooperative Drones

  Small unmanned aircraft operate at low altitude and low speed with a minimal signature – characteristics that make them considerably harder to detect. We investigate how complementary sensing principles – RF detection, radar, electro-optical and infrared imaging, and acoustics – can be fused into a reliable common operating picture. AI-based classification methods help to reliably distinguish drones from birds and other clutter.

  Counter-UAS · RF · Radar · EO/IR · Sensor Fusion · Common Operating Picture

• 04

  Jet Turbines for High-Speed Unmanned Aircraft

  Electric motors can achieve high flight speeds and long ranges only to a limited extent. Compact jet turbines give unmanned platforms a significantly extended performance envelope. Our work covers the aerodynamic and structural design of the airframe, the design of the fuel system, the integration of the engine control unit with autopilot and telemetry, and thermal management. Validation proceeds in stages, from the engine test stand to flight testing.

  Jet Turbines · ECU Integration · Fuel System · Thermal Management · Test Stand

Technologies

Technologies and Components for Unmanned Systems

• Flight Control

  Open autopilot platforms such as PX4 and ArduPilot as the basis for custom control and mission functions.

• Edge Computing

  High-performance companion computers for real-time image processing and AI inference on board.

• Sensor Fusion

  State estimation from IMU, GNSS, LiDAR and camera data for precise and robust navigation.

• LiDAR & Camera

  Three-dimensional environment perception as the basis for GNSS-independent navigation.

• EO/IR & Radar

  Complementary sensors for detecting small targets with a low signature.

• Data Link

  Telemetry and video transmission via radio, cellular networks or optical fiber.

• Jet Turbines

  Compact turbine engines with an electronic engine control unit (ECU).

• Test Stand

  Instrumented measurement of thrust, exhaust gas temperature and fuel consumption prior to flight testing.