rpng/MINS

An efficient and robust multisensor-aided inertial navigation system with online calibration that is capable of fusing IMU, camera, LiDAR, GPS/GNSS, and wheel sensors. Use cases: VINS/VIO, GPS-INS, LINS/LIO, multi-sensor fusion for localization and mapping (SLAM). This repository also provides multi-sensor simulation and data.

What it solves

MINS addresses the difficulty of fusing data from multiple asynchronous sensors—such as cameras, LiDAR, GNSS, wheel encoders, and IMUs—into a single, robust navigation system. It overcomes common challenges like high computational complexity and the need for precise intra-sensor calibration to provide accurate localization for robots.

How it works

It uses a filtering-based approach to fuse five different sensing modalities. The system employs a dynamic cloning strategy and high-order state manifold interpolation to maintain lightweight estimation performance while ensuring consistency. It also performs online calibration of all onboard sensors to ensure they work together accurately in real-time.

Who it’s for

This project is designed for robotics researchers and engineers developing autonomous systems that require high-precision localization across various sensor combinations (e.g., VINS, LiDAR-IMU, or full multi-sensor suites).

Highlights

  • Flexible Sensor Fusion: Supports arbitrary combinations of IMU, wheel encoders, cameras, LiDARs, and GNSS.
  • Online Calibration: Automatically calibrates onboard sensors during operation.
  • Lightweight Performance: Uses dynamic cloning and manifold interpolation to reduce computational load.
  • Comprehensive Tooling: Includes a multi-sensor simulation toolbox and an evaluation suite for analyzing accuracy, timing, and consistency.

Related

  • Project
  • Project
  • Project
  • Project
  • Project