Developing the OSP control in-house, and designing machine and control as one (1963)

Fifteen years of not buying the control

What was chosen in 1963 was not a supplier for control units but a state of affairs: holding inside the company a division that would keep losing money. Buying control from outside required neither development spending nor staff, and it fitted the division of labour of the time. President Okuma Koichi took the opposite road, and as a consequence the company carried an unprofitable division right through to the loss-making years of the mid-1970s. The danger of engineers’ enthusiasm drifting away from the market comes out frankly in the phrase president Okuma Takeo later used for it — the wastrel son.

That said, doing it in-house was not in itself the right answer. That Takeo rebuilt the development process in 1978 from engineer-led to product-that-sells; that Japanese manufacturers began adopting NC machine tools in earnest that same year; that the company had enough strength as one of the machine-tool big five to carry fifteen years of losses without cutting the division — these three together produced the ¥1.5 billion of net profit in the year to March 1979, and if any one had been missing, OSP would have ended as the wastrel son. The self-definition restated from “machine and electronics as one” to “machine, electronics, information and intelligence as one” stands, in every version, on what was taken in-house in 1963. The value of a decision about what not to buy from outside cannot be measured in fifteen years or so.