How Programmable Logic Controllers Emerged from Industry Needs
Programmable logic controllers (PLCs) were more of a technology evolution than a startling discovery. The earliest robust, easily re-programmed industrial controller—which we know as the PLC—evolved nearly concurrently along three independent paths and involved five companies: Bedford Associates; General Motors Hydra-matic Division, Ypsilanti, MI plant; International Instruments Inc.
Programmable logic controllers (PLCs) were more of a technology evolution than a startling discovery. The earliest robust, easily re-programmed industrial controller—which we know as the PLC—evolved nearly concurrently along three independent paths and involved five companies: Bedford Associates; General Motors Hydra-matic Division, Ypsilanti, MI plant; International Instruments Inc. (3I); Digital Equipment Corporation (DEC); and Struthers-Dunn Systems Division in Bettendorf, IA.
Identifying the needs and early ideas first began to take shape in 1967. Documentation and actual building of prototype devices started in 1968, with early model deliveries and factory tests taking place in 1969 and 1970.
Three early paths
Probably the most publicized early PLC development took place at GM’s Hydra-matic Division plant. Several engineers there collaborated on a concept for what they called a “standard machine controller,” sharing their ideas with a number of GM technology vendors such as DEC and 3I. The GM engineers envisioned a controller to replace troublesome relay panels and provide a simpler interface between computers and machines.
Meanwhile, a second path was being pursued by Bedford Associates, a small New England company that today would be categorized as a control systems integrator. Bedford Engineers developed a controller to replace costly minicomputers and reduce programming time for various machine tool applications.
The third approach underway at the time occurred at the Struthers-Dunn Systems Division. (S-D was a relay manufacturer that was probably best known at the time for its pocket “Relay Engineering Guide,” which taught engineers how to do logic and other clever things with relays).
Struthers-Dunn Systems also had strong automotive ties and was well aware of the shortcomings of relay and timer panels in the automotive mass production environment. The company’s intent was similar, in many ways, to that of the GM group in Ypsilanti—to develop a product that could replace costly, troublesome relay and timer panels.
Don’t call it a computer
Applying industrial computers during the mid- to late 1960s required a great deal of vendor support due to the cost and size of the associated projects. In the mid 1960s, for example, IBM 1800s or GE PAC 4000s were being budgeted at about a half million dollars—about evenly split between the hardware and programming/debugging (all dollar figures mentioned in this article reflect the time period noted).
As a result of these costs, minicomputer competition became intense and costs dropped appreciably. However, most installations were still a $200,000 headache. These high level expenditures made computer installations the last thing most plant managers wanted to deal with at their facilities. Therefore, the persons incorporating digital monitoring and control capabilities into fledgling programmable controllers dared not mention the word “computer” to potential customers.
Dick Morley, founder and president of Bedford Associated, recognized this issue and, along with the other engineers in the company, decided to name their product Modicon, which was an acronym of MOdular DIgital CONtroller. This “don’t mention computer” strategy carried over to other PLC suppliers.
In 1967, a large problem for Morley’s operations was the 6-month or so programming time required to get each minicomputer installation up and running at a client site. Many of Bedford’s projects were similar, thereby leading to frustration with the costs of programming and debugging that was required to be done over and over. This led Morley to begin having thoughts of how to build a simpler but rugged, computer-like control unit that could replace minicomputers for machine tool control and related parts handling.
According to Morley, while walking off the effects of a New Years Eve party on January 1, 1968, he composed a 12-page concept memo which has become known as the earliest documentation of PLC development activity.
Bedford had a prototype up and running by March that they nicknamed “Stupid.” This was the company’s 84th project and so the prototype controller officially became the 084—in accordance with Bedford co-founder George Schwenk’s project numbering system. Early demonstrations of the 084 were made at Bryant Grinders and Landis Machines.
In May, 1968, Bedford Associates sales manager, Lee Rousseau, attended the Westinghouse Annual Machine Tool Forum in Philadelphia. There, a speaker from GM’s Hydra-matic Division presented a wish list of needs and specifications for a standard machine controller. Rousseau immediately made plans to head to Detroit.
Visits to the Detroit area, however, completely revised Rousseau’s market perceptions and a separate manufacturing operation was proposed for Bedford. The company would remain an engineering services provider and a manufacturing company would also be formed.
Four vendors—3I, DEC, Allen-Bradley, and Century Detroit—were initially given copies of the preliminary specifications from GM. Cutler-Hammer, Cincinnati Milling Machine, and Bedford Associates were also included in the solicitation shortly thereafter. Of the group, only 3I, DEC, and Bedford Associates responded with product offerings.
First with a product to GM was DEC, with its PDP-14 unit delivered in June 1969 and installed on a gear grinding machine. Later that summer, 3I delivered its PDQ-II to control a segment of an assembly machine (PDQ stood for “Program Data Quantizer”). Bedford Associates, having organized its new manufacturing company—Modicon—delivered its 084 unit in November of 1969 to replace relay panels for a gear grinding machine.
After initial factory evaluations, Square D negotiated with DEC and began selling the PDP-14 and 14L, GE private-labeled Modicon’s 084 as the GE PC-45, and Struthers-Dunn Systems began marketing its VIP line.
Prices for these early PLCs ran in the $3,500 to $7,000 range. Because it didn’t take much of a production operation to justify the cost of a PLC, word of the new programmable controllers spread rapidly.
Working with Hydra-matic, 3I President John Dute appraised the PLC market potential and realized he needed financing to compete. He brought in Allen-Bradley, who bought 25% of 3I, rights to market the PDQ product, and an option to acquire 3I. Allen-Bradley exercised that option in 1969, but also began an in-house PLC R&D program. The PDQ went through several upgrades but eventually was discontinued.
Production line engineering companies such as Ingersoll, Lamb, and Jervis-Webb looked to PLCs as a way to offer their customers an upgraded system for transfer line control. Meanwhile, Struthers-Dunn regained time lost in the PLC race to legal issues associated with hiring Pete Bartlett and two other Eagle Signal engineers by working through Ingersoll, which was automating some Ford Company lines (S-D Systems later became Uticor).
Amid all this activity, seven companies displayed PLCs at the 1970 Machine Tool Show (see “1970 Machine Tool Show PLC Vendors” graphic).
By 1972, there were about a dozen PLC suppliers offering some 20 models. Both GE (Logitrol) and Square D (SY/MAX) developed their own PLC products and became major suppliers. Industrial Solid State Controls (ISSC) introduced its IPC-4000 and Honeywell became the sales agent. Reliance Electric introduced the Automate 33. Allen-Bradley brought out the PMC (programmable matrix controller) and the PLC-1. Cutler Hammer began marketing an Entrekin-developed CON-64, and Datrak became a PLC supplier.
Control Engineering publisher Byron Ledgerwood estimated at the time that 2,000 PLCs had been installed.
The period between 1972 and 1975 saw further growth and solidification in the use of PLCs. By 1975, a PLC base was established which led to exponential growth throughout the remainder of the decade and beyond. A Modicon sales history reflects the exponential increases in sales throughout the industry.
Fifteen companies exhibited at the 1975 Controls Show. And by that year, a number of European and Japanese controls manufacturers began offering PLC products (see “1975 European and Japanese PLC Manufacturers” graphic).
Maturation to an industry standard
Early on, the controller was known as the PC and publications such as the PC Insider popularized the designation. This was fine until personal computers arrived around 1980 and this office/consumer item stole the PC acronym. To avoid confusion, controls people reverted to a PLC designation—a term that had been registered by Allen-Bradley for their newer model controllers. Allen-Bradley/Rockwell Automation has been quite gracious and has not attempted to restrict the use of the “PLC” term.
DEC was somewhat of a mystery with respect to the PLC industry. After the most noted start among the three GM suppliers, they quietly faded from the business about two years later. A major reason for this was the complexity of the PDP-14’s programming scheme. But why this wasn’t upgraded along with other improvements, as all PLC suppliers made, remains a question.
Shortly after its debut, PLC sales went from zero to more than a $1 billion in a little more than a decade. Equally significant is the fact that PLCs became the workhorse of factory automation. Other technologies such as CNC, motor drives, motion control, robotics, automatic ID systems, and vision systems became factory functional by becoming hitched to a PLC system.
Did you play a part in the history of the PLC? Do you have recollections from early use or development of PLCs? If so, post your comments below using the tools provided below.
Ken Ball is indebted to many friends in the industry, the references listed below, and the PLC-based Club of Detroit for contributed knowledge. Personal conversations with the Bedford Associates staff, Ed O’Connell, Ernie Vahala, Pete Bartlett, John Dute, Odo Struger, et al. were invaluable for insights into events and sequences. Although efforts were made to be as factual as possible, memories differ and errors occur. We would appreciate corrections or comments. Please direct them to the writer at 717-624-5381; or firstname.lastname@example.org .
Origination disputes and naming challenges
Who should be credited with inventing the PLC? Time-wise, Dick Morley’s January 1, 1968 memo is the earliest PLC documentation uncovered. In addition, Morley Patent 3,761,893 predates any other PLC patents.
However, Morley is quick to point out that the PLC really was developed by a team. Knowledge and skills of the entire Bedford Associates staff were utilized.
In later years, a competitor started action to challenge Modicon’s claim that Dick Morley was the “father” of the PLC. Attorneys could not find an early PLC patent listing Dick Morley. The reason: Morley’s invention record was entitled “Small Scale Digital Computer.” It described how to modify computer circuitry and components to perform PLC functions. The issued patent had become entitled “Digital Computer.” Continued legal action was not pursued.
Even if Modicon hadn’t been the first to build a PLC, the company would still have a claim to being originators, as their model was the only one of the first three installed in a factory to survive. In addition, their 084 initiated features that became industry standards.
GM’s Hydra-matic specifications
Requirements for a “standard machine controller” developed by GM Hydra-matic engineers stated that the unit would operate in a factory environment subject to nearby high voltages and vibrations. It would be of modular construction and provide status indications of input and output signals. Stored information or programs would not be altered or lost due to a system power failure of up to 12 hours.
Using solid state components, it would have 32 inputs expandable to 256 and 16 outputs and expandable to 128. Built-in isolation would accept 120vac digital inputs and provide at least sixteen 120 vac; 4 amp outputs. It was to be easily programmed and re-programmed. It was to have 1k of memory expandable to 4k.
According to GM, all three of the PLCs procured for evaluation met the specifications and were up and running going into 1970. However, several 084 features were strongly embraced by factory persons and led to the 084 being the product of choice.
The most important of these features was Modicon’s use of ladder logic programming. The PDQ and PDP-14 programs had to be written in Boolean. Boolean was fine for the computer-experienced engineers, but average plant engineers and electricians understood ladder logic and could work with the 084. It didn’t take other PLC makers long to emulate the Modicon ladder logic and this mode of programming has become a fixture with PLCs.