Tadao Kashio and the Machine That Changed Calculations

I stood in front of it last year at the Kashio family home in Western Tokyo, watching it work.

Clicking away noisily, relays firing in sequence, a desk-sized machine doing arithmetic with a calm certainty that must have felt extraordinary at the time. This was Tadao Kashio’s 14-A relay calculator, developed in the 1950s from solenoids, telephone relays and salvaged parts, and still performing arithmetic nearly 70 years later.

Kashio san did not just build a calculator, he changed the practical boundary of calculation.

Before machines like the 14-A, calculation speed was part of professional skill. Slide-rule proficiency mattered. Accuracy depended not only on understanding the problem, but on the trained hand and eye of the person doing the calculation. That ability was hard-won, unevenly distributed, and easy to take for granted.

The 14-A began to remove that, it made calculation less personal, less manual, and more repeatable.

I never used a slide rule, I entered the workforce in the mid-1980s, when electronic calculators were already normal and desktop computers were beginning to reshape engineering offices. I inherited the world Kashio san helped build, without ever experiencing what it had replaced.

Standing watching those relays, I felt the gap in my own professional history. I had benefited enormously from this shift and had barely thought about it.

That is the thing about genuinely transformative engineering. It removes itself from view.

Nobody thinks about relay calculators now because they succeeded completely. We complain instead about software, BIM coordination, model performance and data exchange, because that is where the friction now lives. The friction that Kashio san and others removed is gone so thoroughly that we can no longer feel where it was.

The 14-A was not elegant in the modern sense. It was large, heavy, power-hungry, and limited to arithmetic operations. But it was a systems shift, not just a product. It changed the relationship between the engineer and the calculation, and that changed what engineering practice could become.

That is the question worth carrying forward into AI-assisted design: not “is the tool impressive?” but “which professional skill is it about to commoditise, and are we ready for what comes after?”

Kashio san was. He helped build the future before most people had named the problem.