Intelligent Autopilot for flying boats

Austral offers you a wind and steering solution at the highest level based on Quantum and Esprit products.
Quantum’s agile infrastructure makes it easy to carry out improvements, adaptations and performance or safety supervisions, on a service basis or by your team, and even to enhance your work.

In a continuous improvement approach, we are preparing the future of ocean racing by working on a major breakthrough for the foilers :

  • For 30 years, as the state of the art has progressed, 14 control-command methods, robust, adaptive, AI and a dozen ‘overlays’ have been experimented by the creator of Austral in real life or in simulation and on different supports (foilers, racing sailboats, archimedean drones, underwater drones..).
  • Various commercial products use these advances today (autopilots, drones, flight controls…), they have won many Vendée Globe and dozens of transats.
  • This work has led to the invention of a new innovative method of the ULC category ‘Unsupervised Learning Control‘ which is currently being protected by intellectual property. It has been validated in simulation and is at the stage of sea qualifications in yaw control, in order to verify all its theoretical promises :
    – Robust : Converges necessarily towards stability.
    – Optimal : The response to a step is ideal, minimizing the delay of the actuator and eliminating any overshoot of the set point, thus providing a gain in performance and energy.
    – Ultra adaptive : With a learning time that can go down to less than 2s for 50% of model change and less than 5s without any knowledge of the boat, the plasticity of the solution largely rivals that of living neurons. Some sensors become optional such as the speedometer and even the steering angle (to drive a car, you don’t need to know either the speed or the angle of the wheels).
    – Predictable : The cognitive functions are humanly understandable and ‘replayable”.
  • The solution is particularly suitable for offshore racing and flying :
    – No adjustment
    – Learning is continuously reinforced during navigation
    – Adapts almost instantly to frequent changes in support and behavior
    (sudden changes in speed / heel / submerged surface, Achimedian<->Skimming<->Flying transitions, sail or appendage adjustments, appendage damage…).
  • Moreover, the chosen method is adaptable for 3D altitude/immersion and attitude controls (UAV, AUV, Foiler…).