Agipix (Autonomy) — Quick Guide

Use this page to navigate the AgiAUTO stack used identically in simulation and real flights.

Autonomy overview

Perception

  • LiDAR-inertial state estimation and mapping for GPS-denied indoor operation.
  • Dynamic obstacle handling and occupancy map generation for planning.
  • Designed for robust indoor inspection with onboard sensing.

Read: Perception

Planning & Control

  • Local planner for dynamically feasible trajectories in clutter.
  • On-manifold MPC tracking with PX4 setpoint streaming.
  • Supports teleop, goal, navigation, and exploration control modes.

Read: Planning & Control

Data‑Driven Autonomy

  • Integration path for learned navigation policies with zero-shot transfer tests.
  • Includes deployment notes and experimental examples.

Read: Learning-Based Planning & Control

Next steps

← Back to Home View on GitHub