mit-acl/mighty

MIGHTY: Hermite spline-based real-time trajectory planner for single- and multi-agent UAVs and ground robots (ROS 2). Accepted to IEEE RA-L.

MIGHTY – Hermite‑spline based trajectory planner

What it is – MIGHTY is a ROS 2‑based open‑source library for fast, smooth trajectory planning of aerial and ground robots. It uses Hermite splines to generate collision‑free paths in real‑time, supporting single‑agent, multi‑agent, and exploration scenarios.

Key capabilities

  • Multi‑UAV swapping – dozens of quadrotors start on a circle and exchange positions while avoiding each other.
  • Ground‑robot exploration – a Pioneer 3‑AT autonomously maps an unknown office and returns to its start.
  • Dynamic‑obstacle handling – the planner re‑plans on‑the‑fly when obstacles move.
  • Simulation & hardware – runs in Gazebo, RViz2, or on real robots (UAVs, Pioneer 3‑AT). A Jetson‑compatible build is provided for embedded deployment.

How to try it

  1. Docker (Linux/macOS) – make build creates a reproducible image that pulls all pinned dependencies. Then run a scenario with one of the make run‑* targets, e.g. make run-multiagent for the UAV swap demo or make run-ground-exploration for the Pioneer test.
  2. Native (Ubuntu 22.04, ROS 2 Humble) – clone the repo, run ./setup.sh (installs ROS 2, pulls the mighty.repos dependencies, builds the C++ code, and adds sourcing to ~/.bashrc). Afterwards launch any scenario with the unified launcher src/mighty/scripts/run_sim.py.
  3. Jetson – the same setup.sh script detects Jetson hardware, skips Gazebo, and builds a lightweight ARM64 version.

Typical command line (native example for a single UAV with interactive RViz goals):

python3 src/mighty/scripts/run_sim.py \
    --mode interactive \
    -s ~/code/mighty_ws/install/setup.bash

Replace --mode with multiagent, gazebo, exploration-singleagent-ground, etc., and use flags such as --num-agents 5, --env easy_forest, or --no-rviz for head‑less runs.

Paper & videos – The method is described in the IEEE Robotics & Automation Letters article MIGHTY: Hermite Spline‑based Efficient Trajectory Planning (2026) and the accompanying arXiv pre‑print. A full demo video is hosted on YouTube (https://youtu.be/Pvb‑VPUdLvg) and the README includes many GIFs showing the planner in complex forests, dynamic environments, and long‑duration flights.

Who might use it – robotics researchers, developers of UAV swarms, or anyone needing a lightweight, real‑time planner that works out‑of‑the‑box on both simulation and real hardware.

Related

  • Project
    AndrejOrsula/pymoveit2

    pymoveit2 is a Python wrapper for ROS 2 MoveIt 2, providing a high‑level API to plan and execute robot motions, run kinematics, control grippers, and edit the planning scene from simple Python scripts.

  • Project
    cosmicstack-labs/mercury-agent

    A soul-driven AI agent with permission-hardened tools and a SQLite-backed Second Brain memory, accessible via CLI, Web, and Telegram.

  • Project
    e-sensing/sits

    SITS is an open‑source R package (with a Python wrapper) for building, regularising, and classifying satellite‑image time‑series data cubes. It connects to many public EO collections, offers a suite of ML/DL models (including GPU‑accelerated CNNs and attention encoders), and provides end‑to‑end tools for training, classification, smoothing, uncertainty estimation, and active learning.

  • Project
    JacopoPan/aerial-autonomy-stack

    A flight-proven, open software stack for developing, simulating, and deploying perception-based multi-drone autonomy using ROS2, PX4, and ArduPilot.

  • Project
    dusty-nv/jetson-containers

    jetson‑containers is a modular Docker‑based build system that supplies ready‑made, CUDA‑compatible AI/ML, robotics, and simulation packages for NVIDIA Jetson edge devices. It lets you pick frameworks (PyTorch, TensorFlow, LLM toolkits, ROS, etc.), build a single container with a single command, and run it with NVIDIA runtime defaults. Ideal for developers and researchers who need reproducible, high‑performance AI environments on Jetson hardware.