September 19, 2025 The West Is Forgetting How to Build Things

As Western automakers slash costs, they’re outsourcing the knowledge that built their industries. China is watching—and waiting.

By Ville Lehtonen, Vice President of Product, Realtime Robotics

Western automakers are in a bind. Chinese EV makers have cracked the code on building low-cost vehicles at scale and are designing them 30-40% faster than Western equivalents. Legacy manufacturers are struggling to respond. With pressure mounting to compete, many are pressuring their integrators to go faster and cost less, with the consequence being their local partners cutting internal engineering teams and outsourcing production expertise to suppliers in Vietnam, India, and China.

On paper, this makes sense. Labour is cheaper. Integration firms promise turnkey lines. And boards demand greater margins. But in practice, Western OEMs are walking into a strategic trap: after losing the ability to build their own factories, they now risk losing the ability for the collective west to build factories at all—and with it, the autonomy to adapt and compete. The manufacturing expert is at an advantage in any industry where production quality and price matter, and that’s practically all of them.

The automotive industry has always prided itself on scale and complexity. But those same traits are now turning into liabilities. Building a modern, robot-heavy assembly line involves everything from part logistics, to weld planning, to robot motion sequencing. When those tasks are outsourced—piecemeal and globally—the company outsourcing them is no longer the manufacturer. It becomes a design-and-marketing firm, dependent on others for execution.

That would be concerning, even in peacetime. But this is a moment of geopolitical and technological upheaval. China’s industrial advantage isn’t just cheap labour, it’s end-to-end control of manufacturing infrastructure. Meanwhile, Western firms are hollowing themselves out, laying off the engineers who once gave them operational leverage. The result? Each new vehicle platform takes longer and costs more to launch, even as demand shifts and product cycles compress.

This trend isn’t unique to automotive. It’s visible in aerospace, consumer goods, and industrial machinery. But auto will be the next industry to hit the wall—solar panels and batteries already did—largely because of the massive investment by the Chinese government. Others are sure to follow as manufacturing skills accumulate more and more in China.

Some are betting that digital tools will close the gap. After all, if software does most of the work, the skilled workforce that China has will be of far less value. NVIDIA’s Omniverse, Dassault’s 3DEXPERIENCE, and Siemens’ Xcelerator all promise to bring the factory into the cloud—to make it programmable, virtual, and adaptive. And in theory, they can.

But one bottleneck remains: motion planning. The ability to simulate how hundreds of robots will behave—efficiently, safely, and without collisions—is still a computational nightmare. And without it, all other simulations are speculative. You can model material flow, logistics layouts, and workstation placement. But until you prove that every robot in a line can execute its tasks at speed, across product variants, without collisions or deadlocks, your entire factory plan remains a hypothesis—and a precarious one.

The issue is that if you’re wrong, the consequence isn’t a missed KPI—it’s an expensive and time consuming rebuild. That risk demands costly, time-consuming engineering trials at every stage. It slows the pace of innovation, locks in rigidity, and undermines confidence in every decision. Without fast, scalable motion planning, the dream of concurrent design and flexible manufacturing remains just that: a dream, with manufacturers deterred from ever touching a line once it’s “good enough.”

But recent advances are changing that. New software platforms using cloud-scale motion planning—think of it as AI for robots—can simulate entire workcells, or even whole production lines, in hours. They can run hundreds of variants through feasibility checks, generate cycle time estimates, and suggest optimal robot sequences. It’s the industrial analog of going from writing code in assembly to using a modern compiler.

What this unlocks is not just efficiency, but agility. A manufacturer with this capability can test a new product, simulate its manufacturability, and launch production weeks—not months—later. They don’t need to rebuild lines. They just need to revalidate the ones they already have. The implications are enormous: better line reuse, lower CapEx, faster pivots, and fewer costly errors.

This is what China is already investing in. Western manufacturers still talk about automation, but too often they treat it as an end-state, not a discipline. They build fixed lines, run them for 10 years, and hope the market doesn’t shift. That model is dying.

If Western manufacturers want to remain manufacturers—and not just tech-branded marketing agencies—they need to stop bleeding industrial knowledge. To maximize profitability and competitiveness, organizations must embrace rapid technological advancements and seek out the partners and integrators that can make that happen. That means reinvesting in line engineering, but more importantly, in the tools that make line engineering fast, precise, and reusable.

Digital twins are necessary. But without real motion planning behind them, they’re just expensive PowerPoint slides.

About the Author:
Ville Lehtonen is the VP Product at Realtime Robotics. He has worked in robotics for over 17 years, with experience spanning Life Sciences, Logistics, and Manufacturing. In Life Sciences, Ville contributed to automation innovation at both LabMinds and HighRes Biosolutions. In Logistics, he led product at Pickle Robot, helping scale AI-powered unloading systems into commercial operation. Most recently, he has focused on real-time motion planning and adaptive control for high-mix manufacturing environments. Ville holds graduate degrees from Oxford University and Aalto University, with a background in both business and computer science. Originally from Finland, he currently resides in Boston, MA.

Subscribe to Industry Today