Coming soon: Marine surface vehicle example using radar for localization

One of our immediate aims is to demonstrate the NavAbility technology in real-world applications, and in this vein we’d like to give a preview of results from an exciting project.

This example is applicability to a variety of use-cases, from docking marine vehicles through to navigation of industrial indoor robots… And yes, self-driving too.

In the coming weeks we will release this as a complete working example. In the meantime, here is a short overview and preliminary results.

Seagrant REx at the MIT pavilion with Njord submersible hanging from center

REx and Radar Data

The MITSea Grant Remote Explorer (REx), shown above, is an autonomous marine surface vehicle used for diverse research applications . The vehicle is equipped with a variety of sensors, however the focus of this project is to use only the radar data for localization as it entered the Boston harbor. 

Dehann Fourie (CEO) constructed a GPS-denied navigation solution using the NavAbility stack, processing raw ROS radar payloads (shown below) into a complete factor graph to calculate a SLAM solution.

The example demonstrates how specialized data types can be created to capture and process radar information, which is simple to do using the Caesar.jl stack. A preliminary result of the solved trajectory is shown below.

Raw radar sweep extracted from ROS dataset
The solved trajectory (light-blue) of the vehicle as it travels into the harbor (poses highlighted).

Complete application example to follow soon

Keep an eye out for the complete example in the coming weeks!

Intrigued? Imagine what we can do with your robotics application! Please reach out to us at if you’re curious or have any questions.

NavAbility News

Welcome! We’re going to post regular updates on the NavAbility technology, cutting-edge applications, and innovative projects here.

Please feel free to reach out for more information!

About SlamInDB

  1. Unified Data Management – In this tutorial we that SlamInDb is NOT just a localization solution, it’s also a way recording history and managing robotic memory. We show how easy it is to ingest sensory data and extract video, poses, LIDAR scans for multiple sessions across multiple robots. This tutorial is for the other guys in the room (not the SLAM guy) who want to use the sensory data for image recognition, dashboarding, and other neat applications. Props to ya’ll folks, we love you too!
  2. Non-Gaussian Localization – SlamInDb has unparalled robustness using multimodal SLAM solving. In this tutorial we demonstrate a robust solution that cannot be solved by GTSAM and other competitors. Yep, we’re that good!
  3. Distributed Solving – Real-time factor graph fragment solving for enhanced localization. Here we show how you can keep a conventional SLAM fragment for real-time performance
  4. Minimal Bandwidth Solving – A SlamInDb cloud solution using a 100kb/min channel in an mobile robotics (here’s a relevant use-case)
  5. Global Access – to a 3D multi user environment where multiple stakeholders, live and historic robot data come together — all in the convience of your browser, without having to install any fancy software on your computer. 
  6. API and/or Turn-Key Solution – Want access to our flexible API? Sure! Want to just get a box with our best in it? Of course! you can buy a well integrated piece of robot navigation hardware if just want to solve your mobile robot navigation problem overnight.  
  7. Open Source Community – Our solution draws on the strength of an open source code base (in much the same way Linux does) to offer cutting edge and transparent solutions for your navigation needs. 
  8. Sensor-Agnostic Solution – Whether you’re a intrepid startup, an enterprise looking to build out a solution, a PhD completing a thesis, or a hobbyist with a shiny new Raspberry PI – we are your solution. Take a look at our Getting Started examples here for more details! 

cutting edge research in SLAM maintains our competative advantage

Curious? We provide SLAM as a service free during for the first four hours of infiinitely many navigation sessions.

We provide legal assurance and contracted ensured availability of the Transactional Navigation system (TM).