Industry before control computers
Long before electronic computers, factories relied on mechanical and electrical control. Cams, relays, governors and simple feedback devices kept machines running at a steady speed or switched equipment on in sequence. Large installations, such as chemical plants or power stations, depended on human operators who watched gauges and adjusted valves by hand.
Automation in the broad sense is therefore much older than computing. What computers added, starting around the 1950s and 1960s, was flexibility. A relay panel is wired for one task and must be physically rebuilt to change it. A computer can be reprogrammed, can handle many inputs at once, and can record what happened.
Early process control
Continuous industries were among the first to try computers for control, including oil refining, chemicals and electric power. These plants produce a steady flow of material, and small adjustments to temperature, pressure or flow matter a great deal. Engineers began connecting sensors to computers that could read measurements and recommend or make adjustments.
Early systems were often cautious. Many acted as monitors or advisors, logging data and alerting operators, rather than controlling the plant directly. Machines of that period were large, costly and not always reliable, and the consequences of a failure in a refinery or power plant are serious. Direct digital control, where a computer itself operated the controls, spread more gradually as hardware improved.
Machines that follow instructions
In discrete manufacturing, where separate parts are cut, shaped and assembled, a key development was the numerically controlled machine tool. Starting in the 1950s, engineers worked on machines that could follow coded instructions rather than a human hand on a wheel. The idea grew from numerical control programming, in which a part’s geometry is described as a series of positions and movements. Later, computers took over the control itself, and machine tool control became a software problem as much as a mechanical one.
This changed how factories handled variety. A machine controlled by a stored program could make a different part by loading a different program, which suited small production runs better than fixed tooling. It also tied design and production more closely together, since the same digital description could move from drawing to machine. That link is part of the story of computer-aided design.
Programmable controllers
Factory floors needed something more rugged and approachable than a general-purpose computer. By the late 1960s, the automotive industry in the United States was among those looking for a replacement for large banks of hard-wired relays, which were difficult to change when a production line was retooled. The result was the programmable logic controller, a hardened industrial computer designed to read inputs, run a program and switch outputs.
A major reason for their success was that they could be programmed in ways electricians already understood, often through diagrams resembling relay wiring. This made it easier for plant staff to maintain and modify them. Programmable controllers remain a standard part of industrial equipment, although today’s versions are far more capable than the first ones.
Robots and flexible production
Industrial robots, which began appearing in factories around the 1960s and spread widely later, are another form of computer-controlled automation. Early robots mostly handled repetitive, hazardous or heavy tasks such as welding, painting and moving parts. They worked best in structured settings where the same job was repeated with little variation.
It is easy to overstate what these machines could do. Robots typically needed carefully arranged surroundings, and tasks requiring fine judgment or adaptation to unpredictable parts remained hard to automate. Many factories also found that automation was most effective when processes were redesigned around it, rather than when machines were simply dropped into existing routines.
The business side
Industrial computing was never only about machines. Companies also used computers to schedule production, manage inventory and plan materials, linking the factory to the accounting office. Over time, these planning tools expanded into the integrated systems that coordinate purchasing, production and delivery. The wider story of business computing covers that administrative side, while the movement of goods between companies is a subject of its own in supply chain management.
Sensors and networks have extended the same idea. Equipment can now report its own condition, which helps maintenance staff schedule repairs before something fails. How well this works in practice varies widely between sites and industries.
Jobs, skills and safety
The effects on workers were mixed. Automation often reduced the number of people doing repetitive manual tasks, and communities that depended on a single plant could be hit hard when production moved or shrank. At the same time, demand grew for technicians, programmers and maintenance staff, and some dangerous or physically punishing tasks were taken over by machines.
Economists disagree about how to weigh these effects, and results differ by country, industry and decade. Globalization, changing consumer demand and trade policy also influenced manufacturing employment, so it is difficult to attribute job changes to computers alone.
Safety cuts both ways. Automated systems can keep people away from hazards and apply consistent control, but software faults, poor design or inadequate training can also contribute to accidents. Industrial standards and testing practices developed partly in response to such risks.
What the history suggests
Computers changed industry less through a single dramatic moment than through steady accumulation: more sensors, more programmable equipment, better links between design and production. The gains were often in consistency, flexibility and information rather than in removing people entirely. The costs, including retraining and local job losses, were real and unevenly spread, which is a pattern that recurs across the wider history of computing and its effects on work.