Control system hardware is a very diverse business. Siemens, ABB, Eaton, Rockwell and a universe of suppliers make up a marketplace of approximately $200 billion. At the supplier level the definition of control is pretty much anything from medium voltage switchgear, transformers, insulators and everything needed from the electric utility supply down to a 5 volt transistor logic switch.
PLCs, HMIs, AC induction motors, servo motors, and all the emerging communications products that are used to network these systems together make up the bulk of products sold in the giant controls marketplace. The forces acting in this marketplace are complex and difficult to comprehend. Except for one absolute certainty; competition and technology are constantly driving the cost of control lower.
For most modern control systems, the core processor is generally a PC derivative chip. But it hasn’t always been that way, and this provides some clues as to where things might be going in the future.
Early control systems were mechanical. Cams, gears, couplings, were fashioned to create machinery that produced complex textiles during the first industrial revolution. These systems have evolved slowly from water wheel powered systems to steam engine powered systems. When the electric motor, pneumatic and hydraulic actuators began to replace less sophisticated systems, control systems needed to keep pace. Electromechanical relays used to produce telegraph messages in the middle 1800’s later found use as controls for electrical systems.
Once it became possible to model processes in electrical terms, the trend to this type of solution became widely applied. When the relay reached its limit, and early computing systems became transistorized, the new frontier of the PLC was launched. The PLC presented an early digital solution based on a descriptive language that already existed which made the transition for low tech to new tech a little easier.
This story is repeated in all the major control disciplines and it was significantly impacted by the hardware solutions available at the time. PLC’s ran too fast for process controls and too slowly for CNC’s or motion controllers. PID algorithms were used to describe chemical processes and GCodes were used to describe machining operations. So to make the transition smoother, the separate languages evolved and processor platforms were often strained to meet speed requirements.
Fast forward 50 years to processors that run millions of times faster than the hardware of the 1960’s and not only can any processor run any equipment, but the choice of programming techniques is unlimited as well. Programming languages have migrated based on two elements; meeting the needs in describing the class of problems, but also in coping with the needs of the hardware and operating system.
How might hardware and software migrate in the future?
More next week.
Tell Us What You Think!