Classical computing gave us automation. What's coming next is something closer to a factory that thinks. But getting there requires two technologies converging in a way that many might not fully understand. And the implications for industrial operations are bigger than most people in the industry currently assume.

Quantum computing offers a fundamentally different mental model, one that reframes the questions that industrial companies should be asking.

If classical computing is working with zeroes and ones, one value at a time, then quantum computing represents a zero and a one simultaneously. Stack enough of these quantum bits (qubits) together, and the system can explore many solutions in parallel. Armed with this capability, quantum computing stands to transform industrial operations.

The ideal quantum problem is easy to understand and requires few data inputs, yet demands a staggering amount of computation to solve completely. 

Consider the Traveling Salesman Problem, a classic example in computer science. The setup is simple: A salesperson needs to visit a set of cities exactly once and return home via the shortest route possible. With five cities, there are 120 possible routes to evaluate. With 10, the number exceeds 3.6 million. Increase the amount of cities to 100, and it would take billions of years for existing computers to check the number of possible routes.

The problem fits on a napkin, but the computation required to solve it is beyond anything a classical machine can do at a meaningful scale. Quantum computing could evaluate these solutions far more efficiently, and researchers are making progress on this exact problem.

Quantum computing’s potential raises questions about which solutions are possible on the factory floor, if only we weren’t limited by the classical computers of today.

Say that a manufacturer runs 12 machines across 40 jobs, each requiring a specific sequence of operations with varying setup times between them. What is the optimal schedule to maximize throughput and minimize idle time? 

The inputs are clean and well understood. The solution space is enormous. Every planner in every plant in the world is solving this problem every day, and every one of them is settling for an approximation instead of an answer.

The same is true in supply chain network design. Which combination of suppliers, distribution centers and transportation lanes minimizes both cost and risk across a global network? 

Add in disruptions like a port closure, supplier failure or a demand spike, and the combinations multiply into territory that classical optimization software handles through heuristics (rules of thumb) and compromise.

When industrial companies say their operations are optimized, what they usually mean is that they found the best answer their systems could compute within a practical timeframe. That is meaningfully different from the best answer that exists. 

We have built entire operational philosophies around the limits of our calculators. For the most part, we don't even realize we've done it.

So why isn’t quantum computing a widespread industrial solution today? There are a few reasons: 

• High error rates. Current computers can run small experiments, but lack the stability and size to solve industrial-scale problems.
• Scaling challenges. Qubits are fragile, requiring isolation and sub-zero cooling. If they decay and lose their quantum state, they produce errors. The right environment is difficult to maintain at scale.
• High costs. Quantum computers require large investments in specialized infrastructure.

Quantum hardware today is a bit like early transistor-based computers: the physics works, the potential is real, but the engineering still has a long way to go. It’s not yet able to outperform classical systems on the problems that matter most at industrial scale. 

But artificial intelligence today is contributing to the advancement of quantum computing by developing new quantum algorithms, optimizing how quantum programs are written, and reducing the error rates that make current hardware unreliable.

Picture a modern facility with hundreds of robots, assembly stations and computer numerical control (CNC) machines, each equipped with a local AI model that reads work orders, adapts to variation on the line, flags anomalies, and makes real-time decisions within its immediate context. That intelligence layer is where large language models (LLMs) and task-specific AI models excel.

Next, consider the simultaneous coordination of all those nodes. How does the system optimize and synchronize the behavior of hundreds of discrete automation points? How does it account for every interdependency, constraint and ripple effect across the network in real time?

Classical computing manages this reasonably well when factories run fixed automation on deterministic programs — that is, programs where randomness isn’t involved. But as the nodes themselves become intelligent and adaptive, the coordination problem grows. That is precisely the problem that quantum is capable of solving .

In the future, manufacturers could take advantage of two distinct layers of intelligence operating in complementary roles. AI would live at the edge, with every machine, robot and station making smart local decisions. Quantum would be the orchestration layer, coordinating all of those nodes simultaneously and optimizing the whole system in ways no classical computer can match. 

Fault-tolerant quantum hardware is still years away from broad commercial availability. But that doesn’t mean industrial leaders shouldn’t be thinking about quantum now. 

Following are three ways to start laying the groundwork for quantum at your company:

Map your hardest optimization problems. Quantum won’t be useful for everything. Identify the scheduling, routing and network-design problems in your operations where you know you are currently settling for approximations. These are the problems worth having ready when the hardware catches up.

Invest in the connective tissue. The middleware layer between edge AI and system-level orchestration — the operating system of the intelligent factory — is still largely unbuilt. A practical first step is auditing how your current automation systems communicate: where data flows freely, where it hits walls, and where decisions made at one node never reach the next.

Companies that develop fluency in how those layers communicate, and what data needs to flow between them, will have a significant head start.

Build cross-disciplinary capacity now. Quantum algorithm development requires people who can think across physics, computer science and applied domain knowledge simultaneously. That is a rare combination, and it takes time to develop. Start building those capabilities through hiring and professional development to position your organization to move when the technology is ready.

Thanks to AI, quantum computing in industrial organizations is closer than you might think. The infrastructure is being built; the research is advancing, and the problems waiting to be solved are sitting right there in your operations. 

Quantum computing will change the industrial landscape. By the time it’s practical, knowing how to apply quantum will be a greater advantage than access to the technology.

Jason Hehman is industrials vertical lead at TXI.