Quantum computing in manufacturing

Factories run on two kinds of hard problems: deciding what to build in which order, and understanding what things are made of. Quantum computing in manufacturing touches both, at very different speeds. There is at least one scheduling application in production today. Materials simulation is a research bet measured in years. And there is a third topic most plant managers haven’t heard about yet, which is the cryptography inside the machines they buy and the products they ship.

What does quantum scheduling in a real factory look like?

Quantum scheduling in a real factory today means a hybrid solver, where classical computing does much of the work alongside a quantum annealer, sequencing vehicles or jobs through a line. The best-known case is Ford Otosan in Turkey, which put such an application into production in 2025 and reported much faster scheduling.

Ford Otosan, Ford’s joint venture in Turkey, put a sequencing application into production in 2025 for its Ford Transit line, built on D-Wave’s hybrid solvers. According to D-Wave’s announcement, the line handles a very large number of vehicle variants, and scheduling time fell sharply. The company said it plans to extend the approach to other body shops, paint shops and buffer zones.

Two details matter when you read results like this. The solver is hybrid, meaning classical computing does much of the work alongside the annealer. And the comparison is against the process the plant had before, not necessarily against the best classical solver money can buy. That doesn’t make the result less useful to Ford Otosan. It does mean your own test needs a strong classical baseline, or you won’t know what you are paying for.

Problems with a similar shape show up in most plants: job-shop scheduling, mixed-model sequencing, changeover minimization, maintenance windows, and line balancing. If you have one that costs real money and that your current planning tool handles badly, it is a reasonable candidate for a hybrid quantum proof of concept.

How far off is quantum materials simulation for manufacturers?

Quantum materials simulation is still years away from giving manufacturers useful answers. Today’s machines are used to develop methods and train people, and claims of advantage on materials problems are contested. Industrially useful simulations of catalysts, polymers or battery materials are expected to need error-corrected hardware.

Quantum computers were first proposed for simulating nature, and materials are where many researchers expect the clearest long-term advantage. Large chemical and automotive groups, Covestro and Mitsubishi Chemical among them, have run multi-year collaborations with quantum hardware and software companies on catalysts and materials, and the BMW Group and Airbus quantum computing challenge included a category on simulating material behavior.

The contested part is worth knowing. In March 2025 D-Wave announced a Science paper claiming quantum supremacy on a useful simulation of magnetic materials. The same week, EPFL researchers Mauron and Carleo posted an arXiv preprint arguing that classical methods could simulate the same kind of dynamics. A manufacturer reading vendor claims should expect that kind of back and forth.

Look at who is doing this: large chemical and automotive groups with R&D budgets and long horizons. Most of these projects use today’s small machines to build methods and skills, on the expectation that the useful answers come with error-corrected hardware. If you are a mid-sized manufacturer, the practical move is to keep one or two people close to this work, not to fund your own simulation program.

Does quantum computing help with quality control or predictive maintenance?

Quantum computing has not been shown to beat classical machine learning for quality control or predictive maintenance. Quantum machine learning is an active research field with some industrial proofs of concept, including failure prediction for heavy machinery, but none has shown an advantage a plant should budget around this year.

Results from those pilots are worth following. For a plant’s quality or maintenance program this year, better sensors, labeled data and standard models will do far more.

Which factory systems need post-quantum cryptography first?

The factory systems that need post-quantum cryptography first are the ones that are hardest to change later: firmware signing on the equipment you build or buy, remote access to plants, and design files shared with suppliers. A PLC or connected product that only verifies classical signatures may never accept new keys in the field.

This is the part with dates. NIST published post-quantum standards in August 2024 and, in draft guidance from November 2024, plans to deprecate RSA and elliptic-curve algorithms in 2030 and disallow them in 2035. For an IT system, that means a software upgrade. For a PLC, a robot controller or a connected product, it can mean hardware that can’t be updated at all.

Three places deserve attention. Firmware signing on the equipment you build or buy, because a device that only verifies classical signatures can’t be moved to new keys later. Remote access to plants, often through VPNs and vendor portals. And the design files and process know-how you send to suppliers, which have a long useful life and are exactly what someone would record now to read later.

If you supply the US defense sector, the NSA’s CNSA 2.0 suite adds concrete targets: new national security system acquisitions are expected to support quantum-resistant algorithms from January 1, 2027, software and firmware signing is among the first categories to move, and networking equipment should use CNSA 2.0 algorithms exclusively by 2030. A post-quantum cryptography migration plan for a manufacturer usually starts with the products you ship, because those are the hardest to fix after the fact.

A realistic plan for quantum computing in manufacturing this year

1. Pick one scheduling problem with a clear cost and good data, and benchmark it against a modern classical solver before anyone mentions qubits.
2. Ask your automation and robotics vendors how their current products will support post-quantum firmware signing, and put that question into new purchase specs.
3. If you design connected products, make sure the next hardware revision can verify ML-DSA or SLH-DSA signatures, or at least can accept a new root of trust in the field.
4. Train two or three engineers from planning or R&D so they can judge vendor claims. Our quantum training for teams is built for that.

Where AndesQubit fits

We help manufacturers separate the scheduling problems worth testing from the ones a classical solver already handles, through quantum use case discovery, and then run a scoped hybrid pilot if one is worth it. On the security side we plan the cryptographic migration for plants and products. If your operations also involve heavy logistics, the logistics and transport page covers routing in more depth. We are opening engagements in stages; contact us to join the early-access list.