In July and August, I made two business trips to Australia and Austin, Texas. While there, I did the usual bookend keynotes, but by far most attendees were interested in my “fireside chats.”
The fireside chat is an interactive discussion stimulated by the audience with unfettered comments by me.
They were very interested in the history of the PLC that’s not in the history books. I told about my experiences as a machinist on the factory floor helping to pay my academic tuition. I was a member of the machinists’ union. This was helpful because I understood the culture, the society, the problem and the needs of factory floor personnel — from the operator to the production engineer.
The actual beginning of the PLC was done circa 1968. I was president of Bedford Associates doing customer automation applications. On New Year’s Day I had a terrible hangover and was, as usual, late for the next proposal. What I wanted was a generic application box that would solve 80 per cent of my automation needs. Since I had been working on these problems for several years I put them all together into a controller. I knew it could not be a computer filled with spaghetti software. The people I know on the floor wanted to make product, not lines of code. I guess we were making the first “app.”
We were very aggressive in not calling it a computer. The mean old president of Bedford Associates (me) would erase a blackboard that had standard software approaches for the user or the word computer. We had to deal with the problem and find a solution — not take a solution and find a problem.
The attributes had to be easy to use, deal with time to market and appeal to the culture of the right-brain individuals on the factory floor. It could not appeal to the academics or the classical software spaghetti crowd. We developed a specialized logic configuration called “ladder logic.” Details of this will appear further on in this column if you’re still interested. We built some units and went to a show.
The very early approach was robust hardware and classical software, but we knew that would not work. At that time, relays were the champion of automation. They use a system called ladder logic and there were immense sheets of paper covered with a ladder of symbology indicating a connection to a single coil and multiple outputs — just like on the back of every washing machine at the time. Although I resisted somewhat, I finally bought into ladder logic. I had used it often for automation myself but felt we could never sell it to classical engineering staff. We wanted ladder logic to be simplified and not solve every problem. So how many nodes should we allow the line to have? One is silly, two has no expansion capability and three, although mathematically sufficient for languages, seems stupid. So we settled on four contacts and no more.
Ladder logic has interesting characteristics. It has no cross-modulation products. It is a “come from” philosophy and not a “go to” approach. Essentially each line of code can reference the data generated by another line without interacting.
The reason that ladder logic is powerful is because it allows for a shorter time-to-market product. When we used classical computers and relays, a model change in the automotive automation applications took six to nine months. With PLC and ladder logic it took two to three months. We could charge for value, not costs. We could barely keep up with production in the early days. Over time, our revenue increased and our market share slipped as others got into the act.
Some errors are in Wikipedia and other historical descriptions. Why the unit’s approach was successful is not covered on a valuation basis. We did not respond to a GE proposal. There’s a lot more to this story, but I’ll stop now except for a closing note.
I guess I am the oldest geek in town. My hearing, my teeth, my eyes and my legs all have limited value. I was a wee bit late getting out of bed to make the airplane out of Manchester for my Houston talk and could not find my hearing aids. Egads. I had to go anyway. I got a second banana volunteer to help me handle questions from the audience, especially for the fireside chat. I did find an app for a hearing aid on my iPad but it was too big to put in my pocket. But since it was an app, my banana could load the aid app into Siri with some wiring and some coaching since I don’t have an Apple smart phone. We got the smart phone to act as a pocket hearing aid with reasonable success.
Walt Boyes stood on the podium with me to help with audience interaction. He forgot to turn off the phone. Midway through the session we heard some wonderful rock music, which was his incoming ring. It was loud, and I didn’t know how to shut it off. We fixed it and got applause from the audience for being the ultimate geeks.