From Autonomous Machines to Software-Defined Machines: The Next Battle in Off-Highway

Over the last few months, much of the discussion around Off-Highway autonomy has focused on deployment.

Does the technology work? Where does it work? What infrastructure is required to support it?

These are still important questions, but when I look at recent developments across Mining, Agriculture and Construction, I think another theme is beginning to emerge.

The conversation is shifting from autonomous machines to something much broader - Software-Defined Machines. Machines whose capabilities are no longer fixed at the point of sale, but continue to evolve through software, data and AI throughout their operating life.

Agriculture Reaches a New Stage

In recent months, I have deliberately spent more time looking at Agriculture.

Mining often dominates autonomy discussions because of the scale of deployments and the visibility of autonomous haulage fleets. However, recent developments in Agriculture suggest the sector has now reached a similar inflection point.

Autonomous tractors have now surpassed more than one million acres farmed in North America, representing significant growth in deployed autonomy and moving the conversation firmly beyond experimentation and pilot projects.

Just as Mining moved from asking Does autonomy work? To Where does autonomy work? Agriculture appears to be making the same transition. The challenge is no longer proving the technology. The challenge is deploying it across the huge variety of environments agriculture demands.

The Agricultural Challenge Is Different

Unlike Mining, Agriculture operates in highly variable and often unpredictable conditions – terrain and weather changes, unexpected obstacles. Combining this with the realities of rural operations – connectivity limitations, GPS reliability issues, aging legacy equipment – creates a very different challenge.

Mining autonomy was accelerated by relatively controlled operating environments. Agriculture does not have that luxury. As a result, agricultural autonomy is increasingly moving beyond GPS-only solutions and towards systems that combine computer vision, LiDAR, advanced sensing and real-time mapping.

A Pattern Is Emerging

What caught my attention recently wasn't just the technology but the acquisition activity surrounding it.

John Deere's acquisitions of Sentera and GUSS Automation are particularly interesting. Together they add aerial perception, agronomic intelligence and autonomous execution capability into Deere's broader technology stack.

They resemble Caterpillar's acquisition of Monarch Tractor earlier this year and its more recent acquisitions of RPMGlobal and Skycatch. Skycatch adds spatial intelligence, digital twins and near real-time operational visibility, while RPMGlobal brought mine planning and optimisation expertise.

These acquisitions are deliberately strategic - none of these businesses manufacture core equipment. Instead, they provide Autonomy, AI, Planning & Optimisation, Spatial Intelligence, Software and Analytics.

These are capabilities that sit above the machine itself. This suggests OEMs are no longer just building machines, they are building ecosystems around them. Which increasingly points towards a different conclusion: The machine is no longer the product.

For decades, the Off-Highway sector competed on familiar metrics:

·      Horsepower

·      Reliability

·      Durability

·      Fuel efficiency

·      Manufacturing excellence

Those things still matter. But they are no longer the whole story.

Recent announcements from Komatsu and Caterpillar illustrate this clearly.

Komatsu's partnership with Applied Intuition centres on a software-defined vehicle platform designed to continuously evolve throughout the life of the machine, incorporating machine learning, autonomous capabilities and ongoing software updates.

Caterpillar's partnership with NVIDIA points towards embedded AI and increasingly intelligent machine-level decision making.

The common theme is not autonomy itself, but what happens after deployment.

Historically, a machine left the factory as a largely finished product. Today, machines are increasingly being designed as platforms with capabilities that can be added, performance that can be improved and functionalities that can continuously evolve.  The software becomes the product.

Mining, Agriculture and Automotive Are Starting to Converge

Interestingly, this mirrors developments that we have been witnessing occurring in Automotive for years. The term "Software-Defined Vehicle" is now commonplace within automotive discussions.

Vehicles increasingly receive new features, safety enhancements, performance improvements and autonomous capability updates long after they leave the factory. We are beginning to see the same principle emerge across the Off-Highway sector.

The environments and challenges remain very different, but the strategic direction appears to be increasingly similar.

·      Mining is moving toward software-defined haulage systems.

·      Agriculture is moving toward software-defined farming systems.

·      Construction appears to be heading in the same direction.

Technology ecosystems are becoming the key.

The Next Competitive Advantage

This raises an interesting question, if autonomy is becoming established and if electrification is gradually becoming viable - then what creates competitive advantage next?

I increasingly think the answer is:

Rate of improvement.

Two identical machines may no longer deliver identical outcomes.

The difference could be determined by:

·      The software platform

·      The quality of the data

·      The ability to learn from operation

·      The speed at which improvements are delivered

In that world, competitive advantage no longer comes purely from building a better machine. It comes from building a machine that gets better over time. Which requires an ecosystem to support that.

Conclusion

The last few years have largely been about proving autonomy works, but I feel the next few years may be about something entirely different.

John Deere's and Caterpillar’s acquisitions, Komatsu’s software-defined strategy all point in the same direction – moving beyond autonomous machines, to software-defined machines.

The next battleground in Off-Highway is not autonomy itself. It’s the software, data and AI ecosystem surrounding the machine.