Why Industrial Autonomy Will Be Open Autonomy
- vpeng2
- Jul 30
- 3 min read
Updated: Sep 5
By: Stefan Seltz-Axmacher, CEO & Co-Founder, Polymath Robotics
Autonomy is the lynchpin of the future of industrial operations, especially in mining. But for autonomy to reach its full potential, it must be open. That is: built on interoperable, modular technologies rather than vertically integrated, closed systems. This vision, Open Autonomy, is why we at Polymath are excited to contribute to OpenAutonomy.com alongside leading companies like Wenco and Scania.
Closed Systems Dominated the First Wave
Today, most autonomy deployments—whether on public roads, military theaters, or remote mine sites—are built as closed programs. A single vendor or tightly coupled consortium handles every layer: vehicle platform, onboard autonomy stack, fleet management tools, and operational oversight. The result is full-stack vertical integration.
This approach, while occasionally necessary in early development, has been staggeringly expensive and often commercially unsustainable. Decades of work and billions of dollars have gone into creating systems that are narrowly scoped and brittle, with many programs operating at a loss or depending on government support.
I’ve seen this firsthand. My first autonomy company, Starsky Robotics, was an unapologetically closed system. We built our own hardware, designed and deployed all autonomy and safety stacks in-house, developed our own fleet management tools, and even ran our own trucking company.
Back then, this made sense. Truck OEMs weren’t ready to collaborate, GPU compute was limited, and industry infrastructure was primitive—trucks were still dispatched via fax. It was a time when autonomy had to be built like early smartphones or PCs: tightly controlled and hardware-bound.
The Walls Are Coming Down
But the landscape has changed. In mining and other industrial domains, operators now use cutting-edge fleet management platforms like those built by Wenco. The billions invested in automotive autonomy have created a wave of ruggedized, affordable, and high-performance sensors and compute platforms capable of running large AI models at the edge. And forward-thinking OEMs—like Scania—are embracing open, collaborative approaches to autonomy integration.
If the autonomous haulage systems (AHS) of the 2010s were walled gardens, the walls have now come down.

Why Polymath Believes in Open Autonomy
At Polymath Robotics, we’re building onboard software designed from the ground up to support open industrial autonomy. Our platform can integrate with vehicles pre-equipped by the OEM or retrofitted by a systems integrator. We can own the entire onboard autonomy stack—or collaborate, deploying only our certifiable safety layer alongside another company’s teleop or AI system. Our architecture supports flexible modularity: perception only, planning only, or the full stack.
This open architecture makes it easy to support a wide variety of vehicles and use cases. Improvements in one deployment immediately benefit others. It enables a faster feedback loop, easier innovation, and more scalable, sustainable autonomy systems. Most importantly, it mirrors how the best software in the world is built today: with a modular, interoperable philosophy.
Open Autonomy Is the Path Forward
Open Autonomy isn’t just a technical framework—it’s a strategic imperative. If industrial autonomy is going to deliver on its promise—driving productivity, safety, and scalability—it must be built using open principles. That’s why we’re proud to contribute to OpenAutonomy.com, a new industry blog and initiative dedicated to making this vision a reality.
We invite others across the ecosystem—OEMs, system integrators, autonomy developers, operators, and regulators—to join us in shaping the open, modular future of autonomy.



