Hybrid Transformer based Multi-agent Reinforcement Learning (HTransRL) is for drone coordination in air corridors, addressing the challenges of dynamic dimensions and types of state inputs, which cannot addressed by the traditional MARL.

An eVTOL Aircraft Design Toolbox

A modular, simulator-agnostic Python library for applying physically accurate sensor perturbations (lens distortion, chromatic aberration, Poisson-Gaussian noise) to synthetic imagery, designed to bridge the sim-to-real gap in Advanced Air Mobility computer vision task.