When I first started following autonomous trucking, I believed pure-play startups would lead the way. They moved faster, attracted top AI talent, and captured investor attention.
But over time, it became clear that speed alone isn’t enough and the companies making real progress today look very different.
Spending time with startups, OEMs, fleets, and being on the Tier-1 supplier side, it became clear that the real progress happens where these worlds intersect. The strongest projects today are not defined by who builds them, but by how well teams integrate across traditional industry boundaries.
Startups bring incredible AI innovation but often underestimate the complexity of commercial vehicle engineering and industrialization.
OEMs and suppliers have deep manufacturing expertise but still adapt to faster, software-driven development cycles.
Fleet operators know operations inside out but often lack the technical depth to assess competing solutions.
The winning approach:
▪ Integrated teams - engineers from startups, OEMs, suppliers, and fleets collaborating from day one.
▪ Shared risk - each partner having skin in the game beyond formal contracts.
▪ Operational feedback - real-world data constantly shaping development priorities.
The most successful programs today aren’t just partnerships - they’re true collaborations, where roles blur and everyone contributes to solving the same technical and business challenges.
That’s why I now see the future of autonomy not as a battle between startups and incumbents, but as a convergence of their strengths.
In autonomy and beyond, progress doesn’t come from competition - it comes from collaboration.
Do you see similar patterns of convergence in other emerging tech fields?
Disclaimer: All views are my own.
It's fascinating to see how technology is revolutionizing personal injury law, especially with the impact of AI on case management. Exploring the intersection of legal practices and emerging tech trends could lead to innovative solutions that benefit both firms and clients alike.