Welcome to Vertiport Autonomy’s documentation!

Vertiport Autonomy is a Deep Reinforcement Learning platform for autonomous vertiport coordination. This project implements a sophisticated simulation environment for training and evaluating autonomous agents that manage drone traffic at vertiports (vertical take-off and landing ports).

Contents:

• vertiport_autonomy
  • vertiport_autonomy package
• Tutorials
  • Planned Tutorials

Features

• Deep Reinforcement Learning Environment: Built on Gymnasium API with PPO agent support

• Realistic Simulation: Discrete-time vertiport simulation with formal state machines

• Curriculum Learning: Progressive training strategy for complex scenarios

• Comprehensive Evaluation: Built-in metrics and performance analysis framework

• Configurable Scenarios: YAML-based configuration system with Pydantic validation

Quick Start

Install the package:

pip install -e .

Run a basic evaluation:

python scripts/evaluate.py --agent heuristic --scenario scenarios/easy_world.yaml --episodes 10

Train a new agent:

python scripts/train.py --scenario scenarios/easy_world.yaml --timesteps 10000

Indices and tables

• Index

• Module Index

• Search Page