Back to the blog
Introducing Autonomous Navigation Modules – a toolkit for autonomous off-highway vehicles

Introducing Autonomous Navigation Modules – a toolkit for autonomous off-highway vehicles

At Polymath, we make it easy to automate off-highway vehicles. Today, we're launching Autonomous Navigation Modules, offering you a better way to build your autonomous navigation capabilities.

Stefan Seltz-Axmacher
November 2, 2023

Polymath Robotics was created to radically simplify autonomous navigation for off-highway vehicles. 

Autonomous vehicles demand a broad and deep technical approach. Our autonomous navigation system is tailor-made to handle the wide range and in-depth aspects of this challenge.

Today we’re excited to announce the first of 40 Polymath autonomy modules. Each module is available for individual use, or as part of our full autonomous navigation solution.

It’s way too hard to build autonomous off-highway vehicles

Autonomous vehicles are a boon for off-highway industrial applications. They automate dull, dirty, and dangerous tasks – increasing efficiency, lowering costs, and attracting new types of talent to heavy industry.

Headline from TechCrunch: "The future of autonomous vehicles runs off roads and on to farms, construction sites, and mines"
Source: TechCrunch

Where are all the robots?

Despite demand, there are still fewer than 100 autonomous off-highway retrofits in the US today (via Frost & Sullivan). Even industry leader Caterpillar only automates 4 of its 300 machine models.

Building an autonomous off-highway vehicle remains a massive undertaking, with complex components from low-level hardware abstraction to fancy vehicle maneuvers. Each of these elements is a steep challenge in itself, requiring multi-disciplinary engineers with seemingly godlike intelligence. And that's all before you even get to validate the product!

This is because robotics today is still in its infancy.

Autonomous vehicle technology only became “a thing” in the last 10-15 years, following the DARPA Grand Challenge (2005), the introduction of high-level robotics programming languages (like ROS in 2010) and cheaper sensors (Lidar was only invented in 2004 … and cost $75k as recently as 2010!).

Remember what it was like to build payments Stripe, for example. Sure, developers could build their own payments infrastructure, but it was a huge pain and needed constant maintenance.

Similarly, building an autonomous vehicle today takes a huge amount of effort (and therefore time & money) before you even “pass go” with a working prototype.

That’s why we built Polymath Robotics – the autonomous navigation toolkit for off-highway vehicles.

Our mission is clear: to make automating off-highway vehicles radically simple.

On average, Polymath accelerates autonomy projects by more than a year, and in many cases, we've replaced the need for 5-10 engineers on robotics teams. Polymath helps you build robots without drowning in complexity.

Introducing our Autonomous Navigation Modules

40 autonomy modules: Today, we’re excited to share that Polymath has developed 40 distinct autonomy modules to make autonomous off-highway navigation easier than ever.

Modularity for maximum flexibility: Each autonomy module can be installed individually, or combined as part of our comprehensive autonomous navigation system. 

Automate faster with our building blocks: Polymath’s autonomy modules are the building blocks for your autonomous navigation system, helping you put your off-highway equipment to work faster than ever.

We’ll deep dive into all 40 modules over the next 20 weeks. Every module is a piece of the robotics puzzle, helping you build your autonomous off-highway vehicle faster. 

So without further ado, hare the first two modules:

Module 1: GPS-Free Trailer Localization Module

Why we built this: Many off-highway vehicles pull trailers. One of Polymath’s early customers, for example, is automating a custom vehicle that pulls 2 tandem trailers through farm fields.

A Polymath customer vehicle that tows trailers. Localizing the trailers with GPS is not cost-effective, since each GPS unit costs ~$5k.

The problem: When a vehicle tows trailers, it must know where those trailers are to avoid considering them as obstacles. This is called "trailer localization," usually done using GPS.

However, putting GPS on trailers is expensive, around $5,000 per trailer just for the GPS unit. If you have 100 trailers that get swapped out, that's $500,000 to put GPS on all of them. Plus, the trailer connections can be tough and may break the GPS. 

Solving trailer localization with Lidar: Polymath’s GPS-denied trailer localization module  localizes trailers without GPS, using a Lidar placed on the base vehicle. The Lidar takes in the shape of the trailer, and our module intelligently estimates its relative position and orientation.

Simplifying trailer localization: With Polymath, you can localize trailers easily without having to wade through those complex calculations on your own.

Here’s a video of the GPS-denied trailer localization module in action:

GPS-denied trailer localization used on the customer's vehicle. The module intelligently estimates the position and orientation of the trailers without GPS.

How to use this module: the GPS-denied trailer localization module can be used on its own to enhance your existing autonomy stack, or as part of our full autonomous navigation system. 

Module 2: Lidar Fusion

Why we built this: Useable Lidar data is a prerequisite of almost every autonomous vehicle project.

The problem: The raw output from Lidars is effectively useless. It takes months of dedicated engineering time to wrangle Lidar data into a useable form for robotics. The average robotics team spends ~2-3 months of engineering time denoising, filtering, fusing, clustering, and tracking points. All this before you even start on the real work of your robotics stack!

That’s why we built our Lidar Fusion module – the Lidar data whisperer for your autonomous vehicle.

Skip the noise: Polymath’s Lidar Fusion turns your raw, noisy Lidar data into clean, ready-to-use point clouds.

Powerful and flexible: Lidar Fusion works with any type of Lidar, including multi-beam.

Easy to incorporate: like all our modules, you can use Lidar Fusion on its own, or as part of our full autonomous navigation system.

Save months of development: Lidar Fusion works out of the box, with intelligent algorithms that can be tuned to your vehicle’s unique operational environment.

Our Lidar Fusion module fuses and filters your Lidar data, resulting in a ready-to-use pointcloud

What’s Next: All 40 Modules

Over the next 20 weeks, we are thrilled to unveil all 40 modules in the Polymath autonomy stack, each designed to simplify the path from concept vehicle to reality. 

Each module can integrate seamlessly with existing autonomy projects, or combine with other Polymath modules to form a full autonomous navigation system.

We're excited to be a part of the transformation in the autonomous off-highway industry, from demos and dreams to practical, real-world solutions. With Polymath Robotics, the road to autonomy is not just clearer; it's faster and more accessible than ever before. 

Stefan Seltz-Axmacher
Stefan is CEO & Co-Founder of Polymath. He formerly co-founded Starsky Robotics, the first company to drive an unmanned truck on a public highway. He's been featured in CBS Good Morning, 60 Minutes, WSJ, and Forbes 30 Under 30.

Stefan is CEO & Co-Founder of Polymath. He formerly co-founded Starsky Robotics, the first company to drive an unmanned truck on a public highway. He's been featured in CBS Good Morning, 60 Minutes, WSJ, and Forbes 30 Under 30.

Want to stay in the loop?

Get updates & robotics insights from Polymath when you sign up for emails.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.