Export control is a design requirement, not a check at the shipping dock

For most makers of precision machine tools, export control happens at the end. A machine is designed, built and sold; then, before it leaves the factory, a compliance office checks the destination, the end user and the classification, and either applies for a licence or releases the shipment. I believe that sequence is a mistake for companies like mine. Whether a machine needs a licence, whether its location can be known after delivery and how its service will confirm who is using it are design decisions. They should be settled at the concept review, with the export officer in the room, by the same engineers who choose the spindle and the scales.

I want to be careful about what I am and am not saying. I am an engineer, not a trade lawyer. Japan, like the other members of the Wassenaar Arrangement, controls the export of certain machine tools according to their capability, and licences under the Foreign Exchange and Foreign Trade Act are granted by the authorities on applications that our export trade control office prepares. I will not describe the thresholds here, because they are revised and because the detail belongs to specialists. My argument concerns how a manufacturer organises itself, not what the law requires.

A parameter is a poor foundation for a classification

In 2023 we were preparing a five-axis machining centre for aerospace and medical parts. Its accuracy placed it within a controlled category, so shipments to many destinations would need a licence and take weeks longer to deliver. Our overseas sales leadership proposed a variant whose accuracy would be capped by a controller parameter: identical hardware, a lower stated specification and no licence.

The proposal was not cynical. Competitors were quoting shorter deliveries, our sales teams were losing orders, and the people who proposed it believed it was accepted practice. But our export trade control office took the view that a cap any service engineer could lift by editing a parameter would not be a sound basis for treating the machine as uncontrolled, and I agreed. A machine's real capability lives in its castings, its feedback devices and its compensation. A number stored in memory does not change what the hardware can do.

We built a hardware-distinct variant instead, with different feedback devices and a separately qualified accuracy, for the markets where licences took longest. It took nine months and ¥380 million. During that time we lost or delayed orders we estimate at about ¥5.1 billion. That is a real cost and I do not pretend otherwise. What it bought was a classification we could explain to anyone, including the authorities, without relying on the hope that nobody would ever change a setting.

Designing for the machine that moves

The second design question is what happens after delivery. A precision machine tool can work for twenty years and change hands several times. Some Japanese makers, ourselves included, now fit controlled machines with relocation detection: sensors that notice when a machine has been moved and stop it from running until the maker's service centre confirms the new location and user.

Relocation detection is easy to describe and hard to do well. The first version we deployed treated every move as a suspicion. A customer who shifted a machine across his own factory to make room for a new line waited, on average, nine days for re-authorisation in the early months, and two customers complained to their sales contacts before they complained to us. That taught me that a control measure designed without the legitimate user in mind will be resented and, eventually, circumvented.

We redesigned the process with a group of customers. A customer can now notify a planned move in advance, a service engineer confirms the new position on site or through an agreed video inspection, and re-authorisation takes three and a half days on average. Between April 2024 and March 2026 the system was triggered 38 times. Thirty-six were ordinary moves. Two were referred to the export trade control office, which handled them under our compliance procedures. I do not know what followed in those cases, and I should not.

Service procedures belong to the same design. A software update, a replacement scale or an engineer's visit can each restore or extend a machine's capability, and each is an occasion to confirm where the machine is and who is using it. Technical assistance can raise export questions of its own, so our service engineers are trained to recognise when a request must go to the export office before they act on it.

A seat for the export officer at the concept review

Since 2024 our export trade control office has attended every concept review for a new machine, and it has a vote at the design freeze. At concept, the engineers state the expected classification of each configuration and their reasons. At the freeze, the export office confirms it or holds the design until the question is resolved.

Our engineers did not welcome this at first. Several felt a compliance function had been handed authority over engineering. The export office was equally uneasy, because it was being asked for opinions on designs that did not yet exist. What made it work was a shared document: a short classification note that travels with the design from concept to release and records every decision about capability. When a sales team later asks why a configuration is not offered in a particular market, the answer is already written down.

There are costs beyond the lost orders. Concept reviews take longer. Some features customers asked for have never been offered in certain markets. And the discipline is only as strong as the weakest export practice among our distributors, which we audit but cannot fully control.

None of this replaces the export office's judgement on a particular shipment, and none of it would satisfy the authorities on its own. It changes something narrower and, I think, more important: the moment at which the company discovers that a design has an export consequence. Discovered at the shipping dock, that consequence becomes an argument between sales and compliance about one order. Discovered at the concept review, it becomes an engineering choice that can still be made well.

I would put one question to the board of any precision-equipment maker. When did your export officer last see a product before it was ready to ship? If the answer is never, the company is making its classification decisions in the design office without knowing it, and asking its compliance team to defend them afterwards.