AT&T telephone network mechanical switch failure risk

The 1 in 10,000 per 10 years FIT rate that drove the projection of system collapse. Origin of the "FITs" (failures in time) reliability terminology now used across semiconductors.

In about 1925, AT&T, which was called Bell Laboratories back then, the Bell System, was looking at all the mechanical switches they had. People always say that the transistor came out of basic research. No — AT&T in about 1925 was looking at how many mechanical switches they had in their system, and they started something, a reliability group, a bunch of statisticians who were looking at how to make the system more efficient. They projected that sometime by about 1960, 35 years hence, the whole system would come to a screeching halt because the reliability of mechanical switches was about 1 in 10,000 per 10 years. They call it FITs, failures in time. If you're in the semiconductor business, they talk about FITs, failures in time.

And they say — I've been listening to this all my professional career — that we need to do more science, because doing more science will bring us great new things like the transistor. Well, it turns out the guys who won the Nobel Prize, Bardeen, Shockley, and Brattain at Bell Labs, developed the transistor. But they were doing it because they wanted a highly reliable switch. Everybody says, oh, this came out of basic research. No. It came out of the fact that in the old days — you see it on the old movies, you see the switchboard and operators making the connections. Then they got to the point where they had switches, and just down the street from me is a great big AT&T building — it just be full of switches in the old days. Today they could put it all on a little laptop to do all that same switching.

Tom's debunking of the basic-research transistor myth. Bell Labs' 1925 statistical projection showed mechanical switch failure rates would collapse the phone system by the 1950s; transistor development was applied research justified within the 6% regulated-profit framework.

Regulatory. Utilities and the rate base. Anybody know what a rate base is for a utility? A regulated monopoly. We don't have two companies that deliver electricity to your home, we have one. In the old days we had something called American Telephone and Telegraph — we actually still do, but it was a regulated monopoly in the old days. In a regulated monopoly the regulators in the government will actually set the profitability of the company. People used to think, wasn't it foresightful for AT&T to run Bell Laboratories and spend all this money on basic research. Well that's a load of bull. AT&T had a promise from the regulators that they could make six percent profit on whatever they could justify spending, and they could justify spending on a huge research facility. They did a lot of great basic research, but almost all of it was aimed at improving the productivity and the reliability of the telephone system.

The failure-rate calculation (one in 10,000 hours) that drove the Bell Labs search for an electronic switch. Tom references the AT&T Reliability Handbook as a primary data source.

A lot of what we know about statistical process control was developed at Bell Labs in 1925. They wrote the book originally on statistical process control and statistics and manufacturing, because they had to make telephones. And they were looking ahead, because they could afford all these people to look ahead at six percent profit. They asked, what's the growth of the phone business? They had switchboard operators — people sitting there moving plugs around. Those are all mechanical switches. They looked at the failure rates. There's something called the AT&T Reliability Handbook, and it has all the statistics for failure of mechanical switches and bearings and motors and electrical switches.