Skip links
Flyward
Contact Us
飞沃智航
Flyward
联系我们
01 · AI for Aircraft

AI Visual Navigation and LocalizationKnows where it is, even when GPS is denied.

Flyward delivers cutting-edge AI-based visual navigation and positioning technology for UAVs, eVTOLs, and aircraft, enabling intelligent flight with enhanced safety, precision, and efficiency.

Talk to our team See it in action
ONBOARD CAMERA REFERENCE MAP GNSS · DENIED POSE FEATURE MATCHING · ILLUSTRATIVE
  • 01A third navigation layerAlongside GNSS and inertial navigation
  • 02Camera onlyWorks from onboard cameras or electro-optical pods
  • 03No retrofitNo additional avionics or airframe changes
  • 04Any airframeFor eVTOLs and drones of all sizes and types
Product introduction

Seamless, reliable navigation in the air.

This civilian-grade AI visual navigation and localization software integrates multiple advanced modules powered by artificial intelligence and machine vision algorithms. Designed as a third navigation solution alongside GNSS and IMU systems, it enables real-time guidance for eVTOL aircraft and UAVs of various types and sizes, even in extreme environments where GNSS and inertial navigation fail and only onboard cameras remain operational.

This product delivers the most accurate pure visual real-time navigation and positioning capabilities available in the Asian market. Without requiring any additional avionics or airframe modifications, our software enables reliable visual navigation using only onboard cameras or electro-optical pods, ensuring maximum flight safety and seamless mission execution.

Why visual navigation

Satellites can be jammed. Inertial sensors drift.

Aircraft rely on GNSS and inertial navigation today, and each has weaknesses that cannot be engineered away.

GNSS · Satellite navigation

Signals get lost or jammed

  1. GNSS signals are susceptible to malicious interference from geopolitical factors and electronic warfare, as well as shielding in urban environments such as high-rise buildings, leading to signal loss or instability.
  2. In urban canyons or mountainous areas, GNSS signals reflect off buildings or terrain, causing multipath effects and significant positioning errors.
  3. Changes in the ionosphere and troposphere, especially under solar activity and adverse weather, delay GNSS signal propagation and reduce positioning accuracy.
  4. Failures, maintenance, or orbital drift of GNSS satellites affect signal stability and reliability, leaving receivers without enough satellite signals for accurate positioning.
IMU · Inertial navigation

Errors build up over time

  1. Accelerometers and gyroscopes in an IMU drift over time and errors accumulate. This is especially noticeable during extended use and requires regular calibration.
  2. IMUs are highly sensitive to vibration, temperature changes, and electromagnetic interference. Strong vibration can damage mechanical parts, and extreme temperatures can make sensors unstable or cause them to fail.
  3. Over time, mechanical parts in an IMU wear and fatigue with long-term use, reducing measurement accuracy.
AI visual navigation · The third layer

Real-time positioning from the camera

  • Navigates using only onboard cameras or electro-optical pods
  • Keeps positioning in real time when GNSS and IMU both fail
  • Can also fuse GNSS, IMU, altimeters, and other sensors for more precise multi-source navigation
  • No additional avionics or airframe changes
See it in action
Demo · frame-by-frame matching
Algorithm demo

Camera frames matched to satellite imagery become a position fix.

This is the algorithm’s output on real aerial footage.

  • Green tileThe current onboard camera frame, projected onto the reference map
  • Colored linesFeatures matched between the camera frame and the map
  • Blue dotThe computed aircraft position, output as latitude and longitude in real time
Demo · track over a full flight

The green line is the flight path computed by visual localization; the onboard camera view runs in sync at the top right.

Key features

One software stack for navigation, avoidance, and safety.

Pure Visual Navigation

Autonomous real-time navigation and positioning when GNSS and IMU are unavailable.

Multi-sensor Fusion

Integrates GNSS, IMU, altimeters, and other sensors for more precise multi-source navigation.

Visual Obstacle Avoidance

Real-time detection and avoidance of obstacles in the flight path to ensure flight safety.

Real-time Monitoring

Automatically triggers visual navigation when flight control anomalies occur, ensuring continuous safe flight.

Data Recording

Complete recording of flight and attitude data for verification, analysis, and debugging.

Safety Assurance

Multiple fault detection and emergency measures make the algorithms and software more robust.

Product advantages

Designed for life on board.

From changing light and weather to limited onboard compute and civil aviation certification.

  1. 01

    Environmental Adaptability

    Robust against fluctuating light and varied weather. Deep learning and image-processing techniques keep the flight path assessed and tracked accurately in real time.

  2. 02

    Sensor Compatibility

    Works with many sensor combinations, including cameras, GPS, and IMU, providing stable navigation and autonomous return-to-home across hardware and signal conditions.

  3. 03

    Modular Design

    Each functional module runs independently and can be maintained or upgraded on its own, making the system more flexible and easier to maintain.

  4. 04

    Strict Airworthiness Certification

    We are working with airworthiness certification centers and teams to ensure the software meets civil aviation certification standards such as DO-178C.

  5. 05

    Operating System and Processor Compatibility

    Runs on a wide range of platforms, including Linux and real-time operating systems (RTOS), and supports multiple processor architectures, leaving hardware choices open.

  6. 06

    Customization and Configurability

    From software functions to system architecture, the system can be optimized and adjusted to each customer’s specific requirements.

  7. 07

    Computational Resource Optimization

    Designed around the limited computing resources of eVTOL onboard platforms; algorithm optimizations cut computational demand and improve efficiency.

Advanced feature expansion

One onboard vision system, more to come.

More advanced AI vision functions can be added later, including:

3D Perception and Obstacle Avoidance

Multi-view visual perception that detects and avoids surrounding obstacles in real time.

Learn more

Ground Environment Recognition

Image analysis that recognizes landmarks, buildings, roads, runway markings, airports, and helipads.

Learn more

Aerial Target Detection

Multi-target detection that separates sudden objects from the usual scene, flags hazards such as birds, drones, or other moving objects, and tracks specific targets.

Learn more

Situational Awareness

Recognizes severe weather and sudden hazards in real time, improving navigation, scene understanding, and decisions in complex environments.

Learn more
Application scenarios

Built for the places GNSS can’t be trusted.

A helicopter over a city at dusk

Urban Air Mobility

Keeps aircraft navigating precisely and positioned in real time in dense cities, even where GNSS signals are limited.

A delivery drone carrying a parcel

Logistics Delivery

In urban logistics, and last-mile delivery in particular, helps aircraft move through complex city environments, avoid obstacles, and complete deliveries efficiently.

A drone light show forming a galaxy in the night sky

Aerial Display

Keeps formation shows stable in complex environments, executing precisely even when GNSS and IMU signals are disrupted.

A drone flying beside a cable-stayed bridge

Intelligent City Management

Records key data in real time during infrastructure inspection and monitoring, even when GNSS is unstable.

A helicopter dropping water on a forest fire

Emergency Rescue

In disaster response, keeps aircraft accurately positioned and responsive when signals drop, so the mission gets done.

Work with us

Give your aircraft a way to navigate when GPS can’t.

Tell us about your aircraft, cameras, and missions, and we will work out how AI visual navigation fits in.

Contact us [email protected]