AIRSHIP

Project Overview

The AIRSHIP project focuses on developing cutting-edge autonomous systems for Unmanned WIG (Wing-in-Ground) Vehicles. These innovative aircraft operate in the ground effect zone, offering unique advantages in terms of fuel efficiency and payload capacity for maritime and coastal applications.

As Lead Research Scientist for RL-based autonomous navigation systems, I'm developing a novel modular flight dynamics simulator, used to derive advanced Guidance, Navigation, and Control (GNC) strategies that leverage both optimal control and reinforcement learning policies to enable fully autonomous operation in complex flight control scenarios.

Technical Innovation

🛩️ Ground-Effect Aerodynamics Simulation

Modular flight-dynamics simulator in MATLAB and Python, capturing the highly nonlinear ground-effect aerodynamics of WIG vehicles flying close to the water surface.

🤖 RL-based Autonomous Control

Deep reinforcement learning policies for autonomous low-altitude flight, achieving enhanced speed, flexibility, and energy efficiency compared to traditional approaches.

📊 Classical GNC Benchmarking

Systematic benchmarking of RL policies against classical GNC baselines — PID, LQR, and MPPI — providing rigorous performance comparison across flight regimes.

My Research Contributions

Simulation Framework

Modular flight dynamics simulator: Built a novel modular flight dynamics simulator operating in both MATLAB and Python, capturing ground-effect aerodynamics and parametrizable airframe configurations.
Realistic disturbance modeling: Integrated wind disturbances, sensor noise, and surface proximity effects to enable training under realistic maritime operating conditions.

Autonomous GNC Policies

RL-based control strategies: Designed and trained deep RL controllers for autonomous low-altitude flight in ground effect, targeting robust operation for maritime and coastal missions.
Classical baseline comparisons: Benchmarked RL policies against PID, LQR, and MPPI controllers to rigorously evaluate the advantages of learning-based approaches.
Goal: Enable fully autonomous maritime and coastal operations for unmanned WIG vehicles, a class of vehicles with no existing autonomous GNC solution.

Project Resources

Official Project Website Project Objectives

Autonomous Guidance, Navigation, and Control for Unmanned WIG Vehicles