QUICK FACTS
Created Jan 0001
Status Verified Sarcastic
Type Existential Dread
Keywords
mechanical, hydraulic, pneumatic, electrical, electronic devices, computers, factories, airplanes, electricity, control systems

Automation

“Automation is a broad spectrum of technologies that aim to reduce human involvement in processes. It achieves this by predefining decision criteria,...”

Contents
  • 1. Overview
  • 2. Etymology
  • 3. Cultural Impact

Automation is a broad spectrum of technologies that aim to reduce human involvement in processes. It achieves this by predefining decision criteria, establishing relationships between subprocesses, and dictating specific actions. These predetermined elements are then embedded within machinery. The journey of automation has been paved with various tools, including mechanical , hydraulic , pneumatic , electrical , and electronic devices , with computers playing a pivotal role, often in combination. Complex systems, such as modern factories , airplanes , and ships, are prime examples of this intricate integration of technologies. The advantages of automation are manifold: it leads to significant labor savings, minimizes waste, reduces electricity and material costs, and crucially, enhances quality, accuracy, and precision.

The scope of automation extends to the utilization of diverse equipment and control systems , encompassing machinery in factories , boilers , heat-treating ovens , telephone networks , steering systems, ship stabilizers , aircraft , and a wide array of other vehicles , all operating with diminished human intervention. The spectrum of these applications is vast, ranging from a simple household thermostat managing a boiler to sophisticated industrial control systems overseeing tens of thousands of input measurements and output signals. The banking sector has also embraced automation, with applications varying from basic on-off controls to complex, multi-variable algorithms.

At its core, an automatic control loop functions by a controller comparing a measured process value against a desired set value. The resulting error signal is then processed to adjust an input to the process, thereby maintaining the system at its set point despite external disturbances. This is a classic application of negative feedback applied to a system. The mathematical underpinnings of control theory trace back to the 18th century, with significant advancements occurring in the 20th century. While the term “automatic” emerged earlier, derived from automaton , the term “automation” itself didn’t gain widespread traction until after 1947, when Ford established a dedicated automation department. This period coincided with the rapid adoption of feedback controllers , a trend significantly accelerated by technological innovations introduced in the 1930s that revolutionized numerous industries.

The World Bank ’s World Development Report of 2019 highlighted a compelling trend: the emergence of new industries and jobs in the technology sector has, in many cases, outpaced the economic disruption caused by workers being displaced by automation. However, the narrative isn’t entirely smooth. Job losses and subsequent downward mobility , often attributed to automation, have been cited as contributing factors to the resurgence of nationalist , protectionist , and populist political movements observed in countries like the US, UK, and France since the 2010s.

History

Early history

The meticulous tracking of time has been a long-standing human preoccupation, evident in the ingenuity of ancient civilizations. In Ptolemaic Egypt , around 270 BC, Ctesibius described a float regulator for a water clock . This mechanism, not unlike the ballcock in a modern flush toilet, represented the earliest known feedback-controlled system. However, the advent of the mechanical clock in the 14th century rendered water clocks and their feedback control systems largely obsolete.

The Persian BanĆ« MĆ«sā brothers, in their seminal work the Book of Ingenious Devices (circa 850 AD), detailed a number of automatic control mechanisms. They developed two-step level controls for fluids, a form of discontinuous variable structure controls , and also described a feedback controller . Prior to the Industrial Revolution, the design of feedback control systems was largely an empirical process, relying on trial-and-error and considerable engineering intuition. It wasn’t until the mid-19th century that the stability of these systems began to be rigorously analyzed using mathematics, laying the formal foundation for automatic control theory.

The centrifugal governor , an invention attributed to Christiaan Huygens in the 17th century, was initially employed to regulate the gap between millstones .

Industrial Revolution in Western Europe

The proliferation of steam engines spurred advancements in automation, driven by the necessity to precisely control engine speed and power output. The introduction of prime movers , or self-propelled machines, was a significant development in grain mills, furnaces, and boilers. The steam engine itself necessitated the development of automatic control systems, including temperature regulators (invented in 1624, attributed to Cornelius Drebbel ), pressure regulators (1681), float regulators (1700), and speed control devices. Another control mechanism, designed to automatically adjust the sails of windmills, was patented by Edmund Lee in 1745. Around the same period, in 1745, Jacques de Vaucanson unveiled the first automated loom. By the turn of the 19th century, Joseph Marie Jacquard had developed a punch-card system for programming looms, a precursor to modern computer programming.

In 1771, Richard Arkwright patented the first fully automated spinning mill, powered by water and known as the water frame . A truly pioneering achievement in industrial automation was the development of an automatic flour mill by Oliver Evans in 1785, marking the first completely automated industrial process.

The flyball governor serves as an early illustration of a feedback control system. An increase in speed would cause the counterweights to swing outward, activating a linkage that progressively closed the steam valve, thereby slowing the engine.

A centrifugal governor was first employed by Mr. Bunce of England in 1784 within a model steam crane . James Watt later adopted the centrifugal governor for use in a steam engine in 1788, after his partner Boulton observed one at a flour mill they were constructing. While the governor could not rigidly maintain a set speed—the engine would settle at a new constant speed in response to load fluctuations—it was effective in managing smaller variations, such as those caused by fluctuating heat loads in the boiler. Furthermore, it exhibited a tendency to oscillate when subjected to speed changes, rendering engines equipped with this governor unsuitable for applications demanding precise speed regulation, such as cotton spinning.

Despite these limitations, several improvements to the governor, coupled with advancements in steam engine valve cut-off timing, rendered the engine suitable for most industrial applications by the close of the 19th century. Interestingly, the development of the steam engine consistently outpaced scientific understanding, both in thermodynamics and control theory. The governor itself received limited scientific attention until James Clerk Maxwell published a seminal paper that established the initial theoretical framework for understanding control theory.

20th century

The introduction of relay logic, concurrent with the widespread electrification of factories, saw rapid adoption between 1900 and the 1920s. The expansion of central electric power stations and the operation of new high-pressure boilers, steam turbines, and electrical substations created a significant demand for sophisticated instruments and controls. Central control rooms became a standard feature in the 1920s, yet as late as the early 1930s, most process controls remained rudimentary on-off systems. Operators primarily relied on charts from recorders that plotted instrument data, manually adjusting valves or switches to effect corrections. Control rooms also employed color-coded lights to signal plant workers, who would then manually implement the necessary changes.

The development of the electronic amplifier in the 1920s, crucial for long-distance telephony, necessitated advancements in signal-to-noise ratio, which were achieved through negative feedback noise cancellation techniques. These telephony applications, along with others, significantly contributed to the burgeoning field of control theory. In the 1940s and 1950s, German mathematician Irmgard FlĂŒgge-Lotz pioneered the theory of discontinuous automatic controls, which found critical military applications during the Second World War in fire control systems and aircraft navigation systems .

The 1930s saw the introduction of controllers capable of making calculated adjustments in response to deviations from a set point, moving beyond simple on-off control. These controllers enabled manufacturing to continue its productivity gains, offsetting the diminishing impact of factory electrification alone.

Factory productivity experienced a substantial boost from electrification in the 1920s. However, productivity growth in U.S. manufacturing, which had been 5.2% annually from 1919–29, slowed to 2.76% annually from 1929–41. As noted by Alexander Field, investment in non-medical instruments saw a significant increase from 1929 to 1933 and remained robust thereafter.

The First and Second World Wars were catalysts for major advancements in mass communication and signal processing . Other pivotal developments in automatic controls during this era included breakthroughs in differential equations , stability theory , and system theory (1938), frequency domain analysis (1940), ship control (1950), and stochastic analysis (1941).

Beginning in 1958, the industry saw the emergence of various systems based on solid-state digital logic modules. These were the predecessors of programmable logic controllers (PLCs) and were introduced to replace electro-mechanical relay logic in industrial control systems for process control and automation. Early examples included systems from Telefunken/AEG (Logistat), Siemens (Simatic), Philips/Mullard/Valvo (Norbit), BBC (Sigmatronic), ACEC (Logacec), Akkord (Estacord), Krone (Mibakron), Bistat, Datapac, Norlog, SSR, and Procontic.

In 1959, Texaco’s Port Arthur Refinery became the first chemical plant to implement digital control . The widespread adoption of digital control in factories began to accelerate in the 1970s, driven by the declining cost of computer hardware .

Significant applications

The automatic telephone switchboard , introduced in 1892 alongside dial telephones, marked a significant step in telecommunications. By 1929, 31.9% of the Bell system operations were automated. Early automatic telephone switching relied on vacuum tube amplifiers and electro-mechanical switches, which were highly energy-intensive. The escalating call volume eventually raised concerns that the telephone system might consume all available electricity, prompting Bell Labs to initiate research into the transistor . The logical operations performed by telephone switching relays were, in fact, a foundational inspiration for the development of the digital computer.

The first commercially successful automated glass bottle-blowing machine was introduced in 1905. This machine, operated by a two-person crew working 12-hour shifts, could produce 17,280 bottles in 24 hours, a stark contrast to the 2,880 bottles produced daily by a crew of six men and boys in a manual shop. The cost of producing bottles by machine was a mere 10 to 12 cents per gross, compared to $1.80 per gross for manual glassblowers and their assistants.

Sectional electric drives, developed using control theory, found application in various machines where precise differential speeds between sections are critical. In steel rolling, for instance, metal elongates as it passes through roller pairs, necessitating progressively faster roller speeds. Conversely, in paper manufacturing, the paper sheet shrinks as it moves through steam-heated drying sections arranged in groups, requiring progressively slower roller speeds. The first application of a sectional electric drive was on a paper machine in 1919. A major advancement in the steel industry during the 20th century was continuous wide strip rolling, pioneered by Armco in 1928.

Automated pharmacology production

Prior to widespread automation, the production of many chemicals was a batch process. By 1930, with the increasing use of instruments and the nascent adoption of controllers, the founder of Dow Chemical Co. was advocating for continuous production methods.

James Nasmyth developed self-acting machine tools in the 1840s, designed to replace manual dexterity and allow operation by less skilled labor, even by boys. By the 1950s, machine tools were further automated with Numerical control (NC), utilizing punched paper tape, which soon evolved into computerized numerical control (CNC).

Today, automation is extensively integrated into virtually every manufacturing and assembly process. Prominent examples include electrical power generation, oil refining, chemical production, steel mills, plastics manufacturing, cement plants, fertilizer production, pulp and paper mills, automobile and truck assembly, aircraft production, glass manufacturing, natural gas separation plants, food and beverage processing, canning and bottling, and the manufacture of a vast array of component parts. Robots excel in hazardous environments, such as automotive spray painting, and are crucial for assembling electronic circuit boards. Automotive welding is frequently performed by robots, and automatic welders are employed in applications like pipeline construction.

Space/computer age

With the dawn of the space age in 1957, control system design, particularly in the United States, shifted away from the frequency-domain techniques of classical control theory. This shift led to a resurgence of interest in the differential equation techniques of the late 19th century, which were primarily couched in the time domain. During the 1940s and 1950s, German mathematician Irmgard Flugge-Lotz developed the theory of discontinuous automatic control. This theory found widespread application in hysteresis control systems , including navigation systems , fire-control systems , and electronics . Through the work of Flugge-Lotz and others, the modern era witnessed significant advancements in time-domain design for nonlinear systems (1961), navigation (1960), optimal control and estimation theory (1962), nonlinear control theory (1969), digital control and filtering theory (1974), and the advent of the personal computer (1983).

Advantages, disadvantages, and limitations

Perhaps the most frequently cited advantage of industrial automation is its association with faster production rates and reduced labor costs. Another significant benefit is the ability to replace arduous, physically demanding, or monotonous work with machines. Furthermore, tasks conducted in hazardous environments or those exceeding human capabilities can be safely performed by machines, which can operate reliably under extreme temperatures or in radioactive or toxic atmospheres. These automated systems can also be maintained with relatively simple quality checks. However, it’s important to acknowledge that not all tasks are currently amenable to automation, and the cost of automating certain operations can be prohibitive. The initial investment in installing automated machinery in factories is substantial, and neglecting system maintenance can lead to the loss of the very products being manufactured.

Moreover, some research suggests that industrial automation might have negative societal consequences beyond operational concerns, including worker displacement due to systemic job losses and potentially exacerbated environmental damage. However, these findings are often complex, subject to debate, and may be mitigated through careful planning and implementation.

The primary advantages of automation can be summarized as follows:

  • Increased throughput or productivity: Automated systems can operate continuously and at higher speeds than human workers.
  • Improved quality: Automation can lead to greater consistency and fewer errors in production.
  • Increased predictability: Automated processes are more predictable in their output and timing.
  • Improved robustness (consistency) of processes or product: Automation ensures that processes are carried out in a standardized manner, leading to consistent product quality.
  • Increased consistency of output: The uniformity of automated production minimizes variations.
  • Reduced direct human labor costs and expenses: Automation can significantly lower labor expenses.
  • Reduced cycle time: The time required to complete a task or process is often shortened.
  • Increased accuracy: Machines can perform tasks with a higher degree of precision than humans.
  • Relieving humans of monotonously repetitive work: Automation frees human workers from tedious and repetitive tasks.
  • Creation of new roles: While some jobs are displaced, automation also necessitates roles in the development, deployment, maintenance, and operation of automated processes.
  • Increased human freedom: By automating certain tasks, humans are afforded more time for other pursuits.

Automation primarily refers to machines replacing human actions, but it is also closely linked to mechanization, where machines replace human labor. In conjunction with mechanization, extending human capabilities in terms of size, strength, speed, endurance, visual range and acuity, hearing frequency and precision, and electromagnetic sensing and effecting, further advantages emerge:

  • Relieving humans of dangerous work stresses and occupational injuries : This includes reducing the incidence of back strains from lifting heavy objects.
  • Removing humans from dangerous environments: This encompasses hazardous locations such as fires, space, volcanoes, nuclear facilities, and underwater environments.

The primary disadvantages of automation include:

  • High initial cost: The investment required for automation systems can be substantial.
  • Faster production of defects: Unchecked automated processes can rapidly produce defective products if they malfunction.
  • Scaled-up problems upon failure: When automated systems fail, the consequences can be amplified, potentially releasing dangerous toxins, forces, or energies at an accelerated rate.
  • Underestimation of human adaptiveness: The nuances of human adaptability are often not fully understood by those implementing automation. Anticipating every contingency and developing fully pre-programmed automated responses for all situations can be challenging. Discoveries made during the automation process may necessitate unanticipated iterations to resolve issues, leading to unforeseen costs and delays.
  • Disruption of employment income: Individuals anticipating employment income may face significant disruption when automation is deployed and comparable income opportunities are not readily available.

Paradox of automation

The paradox of automation posits that as automated systems become more efficient, the critical role of human operators intensifies. While humans are less directly involved, their input becomes more crucial when issues arise. Cognitive psychologist Lisanne Bainbridge notably explored these issues in her influential paper “Ironies of Automation .” If an automated system malfunctions, it can propagate errors rapidly until the issue is rectified or the system is shut down. This is precisely where human operators become indispensable. A tragic illustration of this phenomenon occurred in Air France Flight 447 , where an automation failure placed the pilots in a manual control situation for which they were unprepared.

Limitations

This section requires additional citations to ensure verifiability. Please assist in improving this section by adding references to reliable sources. Unsubstantiated material may be subject to challenge and removal. (May 2019)

  • Current technology is not yet capable of automating all desired tasks.
  • Many automated operations involve significant capital investment and produce high volumes of goods, making malfunctions extremely costly and potentially hazardous. Consequently, some human personnel are necessary to ensure the proper functioning of the entire system, as well as to maintain safety and product quality.
  • As a process becomes increasingly automated, the potential for further labor savings or quality improvements diminishes. This exemplifies the principles of both diminishing returns and the logistic function .
  • As more processes are automated, the pool of non-automated processes shrinks. This represents an exhaustion of opportunities. However, new technological paradigms may emerge, establishing new benchmarks that surpass previous limitations.

Current limitations

A significant number of human roles in industrial processes currently fall outside the purview of automation. Human-level capabilities in pattern recognition , language comprehension , and language generation remain largely beyond the reach of modern mechanical and computer systems, although advancements like the Watson computer show promise. Tasks requiring subjective judgment or the synthesis of complex sensory data, such as the interpretation of scents and sounds, as well as high-level functions like strategic planning, still necessitate human expertise. In many instances, utilizing human labor remains more cost-effective than employing mechanical solutions, even when automation of industrial tasks is technically feasible. Consequently, algorithmic management , which involves the digital rationalization of human labor rather than its outright substitution, has emerged as an alternative technological strategy. Overcoming these current obstacles is theorized to be a pathway toward achieving post-scarcity economics.

Societal impact and unemployment

The increasing prevalence of automation often generates anxiety among workers regarding job security, as technology renders their skills or experience obsolete. Early in the Industrial Revolution , when inventions like the steam engine began to make certain job categories redundant, workers often resisted these changes vehemently. The Luddites , for example, were English textile workers who protested the introduction of weaving machines by destroying them. More recently, residents of Chandler, Arizona , have resorted to slashing tires and pelting rocks at self-driving cars in protest against their perceived threat to human safety and job prospects.

The level of anxiety surrounding automation, as reflected in public opinion polls, appears to correlate closely with the strength of organized labor in a given region or nation. For instance, while a study by the Pew Research Center indicated that 72% of Americans expressed concern about increasing automation in the workplace, 80% of Swedes viewed automation and artificial intelligence (AI) positively, attributed to the country’s enduringly strong unions and a more robust national safety net .

According to one estimate, 47% of all current jobs in the US could be fully automated by 2033. Furthermore, wages and educational attainment seem to be strongly negatively correlated with an occupation’s risk of automation. Erik Brynjolfsson and Andrew McAfee argue that “there’s never been a better time to be a worker with special skills or the right education, because these people can use technology to create and capture value. However, there’s never been a worse time to be a worker with only ‘ordinary’ skills and abilities to offer, because computers, robots, and other digital technologies are acquiring these skills and abilities at an extraordinary rate.” Conversely, others contend that highly skilled professional roles, such as those of a lawyer , doctor , engineer , or journalist , are also susceptible to automation.

A 2020 study published in the Journal of Political Economy concluded that automation has significant negative effects on employment and wages: “One more robot per thousand workers reduces the employment-to-population ratio by 0.2 percentage points and wages by 0.42%.” A 2025 study in the American Economic Journal found that the introduction of industrial robots between 1993 and 2014 led to a reduction in employment for both men and women by 3.7 and 1.6 percentage points, respectively.

Research conducted by Carl Benedikt Frey and Michael Osborne from the Oxford Martin School suggested that employees engaged in “tasks following well-defined procedures that can easily be performed by sophisticated algorithms” are at risk of displacement, estimating that 47% of jobs in the US were at risk. This study, initially released as a working paper in 2013 and published in 2017, predicted that low-paid physical occupations would be most vulnerable, based on a survey of expert opinions. However, a study published in McKinsey Quarterly in 2015 proposed that in most cases, computerization leads not to complete job replacement but to the automation of specific tasks within those jobs. The methodology employed in the McKinsey study has faced considerable criticism for its lack of transparency and reliance on subjective assessments. Similarly, the methodology of Frey and Osborne has been questioned for its perceived lack of evidence, historical context, and credible analytical framework. Furthermore, the Organisation for Economic Co-operation and Development (OECD ) estimated that across its 21 member countries, approximately 9% of jobs are automatable.

Economist Gilles Saint-Paul of Toulouse 1 University proposed a model suggesting that the demand for unskilled human capital declines at a slower rate than the demand for skilled human capital increases. In the long term, this trend has resulted in more affordable products, reduced average work hours , and the creation of new industries, such as those in robotics, computing, and design, which offer numerous high-paying, skill-based jobs. By 2030, it is projected that between 3% and 14% of the global workforce will be compelled to change job categories due to automation eliminating jobs within entire sectors. While the number of jobs lost to automation is often offset by jobs created through technological advancements, the nature of the jobs lost and gained rarely aligns perfectly, leading to increased unemployment, particularly among the lower-middle class. This phenomenon is largely observed in the US and other developed economies, where technological progress drives demand for highly skilled labor while simultaneously decreasing demand for middle-wage labor. Economists refer to this trend as “income polarization,” characterized by declining wages for unskilled labor and rising wages for skilled labor, a trend predicted to persist in developed economies.

Lights-out manufacturing

Lights-out manufacturing, a production system designed to operate without human workers to eliminate labor costs, gained traction in the US. General Motors, in 1982, implemented a “hands-off” manufacturing approach to “replace risk-averse bureaucracy with automation and robots.” However, this factory never achieved full “lights out” status.

The expansion of lights-out manufacturing necessitates:

  • Reliability of equipment: Systems must be highly dependable.
  • Long-term mechanic capabilities: Skilled maintenance personnel are crucial.
  • Planned preventive maintenance: Proactive maintenance schedules are essential.
  • Commitment from the staff: A dedicated workforce is required to manage and oversee the automated processes.

Health and environment

This section may contain original research. Please improve it by verifying the claims made and adding inline citations. Statements consisting only of original research should be removed. (March 2018)

The environmental impact of automation varies significantly depending on the specific technology, product, or engine involved. Some automated engines consume more energy resources than their predecessors, while others are more efficient. Hazardous operations, such as oil refining , the production of industrial chemicals , and various forms of metal working , have historically been early candidates for automation. The automation of vehicles holds the potential for a substantial environmental impact, though the nature of this impact could be either beneficial or detrimental, contingent on several factors. Because automated vehicles are projected to be significantly less prone to accidents than human-driven vehicles, certain safety features incorporated into current models (like anti-lock brakes or laminated glass ) might become unnecessary in self-driving versions. The removal of these safety features would reduce vehicle weight. Coupled with more precise acceleration and braking, and optimized fuel-efficient route planning, this could enhance fuel economy and decrease emissions. Despite these potential benefits, some researchers theorize that an increase in the production and ownership of self-driving cars could lead to a surge in vehicle usage, potentially negating any environmental advantages if these vehicles are driven more frequently.

The automation of homes and household appliances is also expected to influence the environment. A study conducted in Finland on the energy consumption of automated homes indicated that smart homes could reduce energy usage by monitoring consumption levels across different areas of the house and adjusting them to minimize energy waste (e.g., automatically reducing consumption during nighttime hours when activity is low). This study, along with others, suggested that the smart home’s ability to monitor and regulate consumption levels would lead to a reduction in unnecessary energy use. However, some research indicates that smart homes may not always be as efficient as non-automated homes. A more recent study suggests that while monitoring and adjusting consumption levels do decrease unnecessary energy use, the systems required for this monitoring also consume energy. The energy expended by these systems can sometimes offset their benefits, resulting in minimal or no ecological advantage.

Convertibility and turnaround time

Another significant shift in automation is the growing demand for flexibility and convertibility within manufacturing processes . Manufacturers are increasingly seeking the ability to seamlessly switch production from Product A to Product B without requiring a complete overhaul of the production lines . This emphasis on flexibility and distributed processes has led to the introduction of Automated Guided Vehicles equipped with Natural Features Navigation.

Digital electronics have also played a crucial role. Formerly analog-based instrumentation has been superseded by digital equivalents, offering improved accuracy, greater flexibility, and expanded capabilities for sophisticated configuration , parametrization , and operation. This transition was complemented by the fieldbus revolution, which enabled networked communication (utilizing a single cable) between control systems and field-level instrumentation, thereby eliminating extensive hard-wiring.

Discrete manufacturing plants have been rapid adopters of these technologies. In contrast, the more conservative process industries, with their longer plant lifecycles, have been slower to embrace these changes, and analog-based measurement and control systems remain prevalent. The increasing adoption of Industrial Ethernet on the factory floor is further accelerating these trends, enabling tighter integration of manufacturing plants within the broader enterprise, potentially extending to internet connectivity. Global competition has also heightened the demand for Reconfigurable Manufacturing Systems .

Automation tools

Engineers now possess numerical control capabilities over automated devices, leading to a rapid expansion in the range of applications and human activities. Computer-aided technologies (CAx) now form the foundation for mathematical and organizational tools used in the creation of complex systems. Notable examples of CAx include computer-aided design (CAD software) and computer-aided manufacturing (CAM software). The enhancements in product design, analysis, and manufacturing enabled by CAx have yielded significant benefits for industry.

Information technology , in conjunction with industrial machinery and processes , facilitates the design, implementation, and monitoring of control systems. A prime example of an industrial control system is the programmable logic controller (PLC). PLCs are specialized, hardened computers frequently employed to synchronize the flow of inputs from physical sensors and events with the flow of outputs to actuators and events.

An automated online assistant on a website, often featuring an avatar for enhanced human–computer interaction .

Human-machine interfaces (HMI) or computer human interfaces (CHI), formerly known as man-machine interfaces, are typically used for communication with PLCs and other computers. Service personnel who monitor and control these systems through HMIs may be referred to by various titles. In industrial process and manufacturing environments, they are known as operators or similar roles. In boiler houses and central utility departments, they are called stationary engineers.

Various types of automation tools are available:

Host simulation software (HSS) is a commonly used testing tool for evaluating equipment software, assessing performance against factory automation standards, including timeouts, response times, and processing times.

Cognitive automation

Cognitive automation, a subset of artificial intelligence (AI), represents an emerging category of automation powered by cognitive computing . Its primary focus is the automation of clerical tasks and workflows involving the structuring of unstructured data . Cognitive automation draws upon multiple disciplines, including natural language processing , real-time computing , machine learning algorithms , big data analytics , and evidence-based learning .

According to Deloitte , cognitive automation enables the replication of human tasks and judgment “at rapid speeds and considerable scale.” Such tasks include:

Recent and emerging applications

CAD AI

Artificially intelligent computer-aided design (CAD) systems can now leverage technologies like text-to-3D, image-to-3D, and video-to-3D to automate aspects of 3D modeling . AI-powered CAD libraries could also be developed using linked open data from schematics and diagrams . AI CAD assistants are increasingly employed as tools to streamline design workflows.

Automated power production

Technologies such as solar panels , wind turbines , and other renewable energy sources, when integrated with smart grids , micro-grids , and battery storage systems, can automate power production.

Agricultural production

Numerous agricultural operations are being automated through the use of machinery and equipment to enhance diagnosis, decision-making, and operational execution. Agricultural automation can alleviate the drudgery of farm work, improve the timeliness and precision of agricultural tasks, increase productivity and resource-use efficiency, build resilience, and enhance food quality and safety. Increased productivity can also free up labor, allowing agricultural households to allocate time to other activities.

The technological evolution in agriculture has led to progressive shifts towards digital equipment and robotics. Motorized mechanization, powered by engines, automates agricultural operations such as ploughing and milking. With the advent of digital automation technologies, it has also become possible to automate the diagnostic and decision-making processes in agriculture. For example, autonomous crop robots can harvest and seed crops, while drones can collect data to support automated application of inputs. Precision agriculture frequently incorporates such automation technologies.

Motorized mechanization has generally seen an increase in recent years. Sub-Saharan Africa remains the only region where the adoption of motorized mechanization has stagnated over the past decades.

Automation technologies are increasingly being utilized for livestock management, although adoption data is still emerging. Global sales of automatic milking systems have risen in recent years, but adoption is likely concentrated in Northern Europe, and remains scarce in low- and middle-income countries. Automated feeding machines for both cows and poultry also exist, but data and evidence regarding their adoption trends and driving factors are similarly limited.

Retail

Many supermarkets and even smaller retail outlets are rapidly implementing self-checkout systems, thereby reducing the need for checkout personnel. In the US, the retail industry employed 15.9 million people as of 2017, representing approximately one in nine Americans in the workforce. Globally, an estimated 192 million workers could be affected by automation, according to research by Eurasia Group .

A soft drink vending machine in Japan, an example of automated retail.

Online shopping can be considered a form of automated retail, as payment and checkout are handled through an automated online transaction processing system. The share of online retail has seen a significant jump, from 5.1% in 2011 to 8.3% in 2016. Notably, two-thirds of books, music, and films are now purchased online. Furthermore, the rise of automation and online shopping could decrease demand for physical shopping malls and retail property, which in the United States currently accounts for 31% of all commercial property, or approximately 7 billion square feet (650 million square meters). Amazon has captured a substantial portion of the recent growth in online shopping, accounting for half of the growth in online retail in 2016. Other forms of automation are also integral to online shopping, such as the deployment of automated warehouse robotics, exemplified by Amazon’s use of Kiva Systems .

Food and drink

KUKA industrial robots being used in a bakery for food production.

The food retail industry has begun to adopt automation in the ordering process. McDonald’s , for instance, has introduced touch screen ordering and payment systems in many of its restaurants, reducing the need for as many cashier employees. The University of Texas at Austin has established fully automated cafe retail locations. Some cafes and restaurants have implemented mobile and tablet “apps ” to streamline the ordering process, allowing customers to order and pay via their devices. Certain restaurants have automated food delivery to customer tables using conveyor belt systems . The use of robots is sometimes employed to replace waiting staff .

Construction

Automation in construction refers to the integration of methods, processes, and systems that enable greater machine autonomy in construction activities. The goals of construction automation can be multifaceted, including reducing jobsite injuries, decreasing activity completion times, and assisting with quality control and quality assurance .

Mining

Automated mining involves the elimination of human labor from the mining process. The mining industry is currently undergoing a transition towards automation. However, in some regions, particularly in the third world where labor costs are low, there may be less incentive to invest in automation for efficiency gains, and significant human capital may still be required.

Video surveillance

The Defense Advanced Research Projects Agency (DARPA ) initiated research and development into automated visual surveillance and monitoring (VSAM) programs between 1997 and 1999, followed by airborne video surveillance (AVS) programs from 1998 to 2002. Currently, a significant effort is underway within the computer vision community to develop fully automated tracking surveillance systems. Automated video surveillance monitors people and vehicles in real-time within busy environments. Existing automated surveillance systems are typically designed based on the environment they are intended to observe (indoor, outdoor, or airborne), the number of sensors they can manage, and the mobility of those sensors (stationary cameras versus mobile cameras). The primary objective of a surveillance system is to record the properties and trajectories of objects within a designated area, generate warnings, or notify authorities in the event of specific occurrences.

Highway systems

As demands for safety and mobility have escalated and technological possibilities have multiplied, interest in automation has surged. In an effort to accelerate the development and deployment of fully automated vehicles and highways, the U.S. Congress allocated over $650 million across six years for intelligent transport systems (ITS) and related demonstration projects through the 1991 Intermodal Surface Transportation Efficiency Act (ISTEA). Congress mandated in ISTEA that:

[T]he Secretary of Transportation shall develop an automated highway and vehicle prototype from which future fully automated intelligent vehicle-highway systems can be developed. Such development shall include research in human factors to ensure the success of the man-machine relationship. The goal of this program is to have the first fully automated highway roadway or an automated test track in operation by 1997. This system shall accommodate the installation of equipment in new and existing motor vehicles.

Full automation, commonly defined as requiring no driver control or very limited driver intervention, would be achieved through a combination of sensors, computers, and communication systems within vehicles and along roadways. Theoretically, fully automated driving could allow for closer vehicle spacing and higher speeds, thereby increasing traffic capacity in areas where road expansion is physically unfeasible, politically unpopular, or prohibitively expensive. Automated controls could also enhance road safety by reducing the incidence of driver error, which is a significant factor in many motor vehicle crashes. Additional potential benefits include improved air quality (resulting from more efficient traffic flow), increased fuel economy, and the development of spin-off technologies during the research and development phases of automated highway systems.

Waste management

Automated side loader operation.

Automated waste collection trucks reduce the need for a larger workforce and ease the labor demands associated with providing this service.

Business process

Business process automation (BPA) refers to the technology-enabled automation of complex business processes . It can assist in streamlining business operations for simplicity, achieving digital transformation , enhancing service quality , improving service delivery, or controlling costs. BPA involves the integration of applications, restructuring of labor resources, and the utilization of software applications across the organization. Robotic process automation (RPA; or RPAAI for self-guided RPA 2.0) is an emerging field within BPA that leverages AI. BPAs can be implemented in various business areas, including marketing, sales, and workflow management.

Home

Home automation, also known as domotics, signifies the increasing automation of household appliances and features within residential dwellings, primarily through electronic means that enable functionalities previously impractical, prohibitively expensive, or simply impossible. The rising adoption of home automation solutions reflects a growing reliance on these systems. Nevertheless, the enhanced comfort provided by these automation solutions is remarkable.

Laboratory

Automated laboratory instrument.

Automation is indispensable for numerous scientific and clinical applications. Consequently, it has been extensively integrated into laboratory settings. As early as 1980, fully automated laboratories were operational. However, the widespread adoption of automation in laboratories has been hindered by its high cost. This situation may evolve with the integration of low-cost devices with standard laboratory equipment. Autosamplers are common devices utilized in laboratory automation.

Logistics automation

Logistics automation involves the application of computer software or automated machinery to improve the efficiency of logistics operations. Typically, this pertains to operations within a warehouse or distribution center , with broader tasks managed by supply chain engineering systems and enterprise resource planning systems.

Industrial automation

Industrial automation primarily focuses on the automation of manufacturing , quality control , and material handling processes. General-purpose controllers for industrial processes include programmable logic controllers , stand-alone I/O modules , and computers. The fundamental aim of industrial automation is to replace human actions and manual command-response activities with mechanized equipment and logical programming commands. A significant trend is the increased use of machine vision for automatic inspection and robot guidance functions, alongside a continuous rise in the deployment of robots. Industrial automation is essential across a wide range of industries.

Industrial Automation and Industry 4.0

The rise of industrial automation is intrinsically linked to the “Fourth Industrial Revolution ”, more commonly known today as Industry 4.0. Originating in Germany, Industry 4.0 encompasses a multitude of devices, concepts, and machines, including the advancement of the industrial internet of things (IIoT). An “Internet of Things is a seamless integration of diverse physical objects in the Internet through a virtual representation.” These revolutionary advancements have brought a new perspective to the world of automation, revealing novel avenues for growth to enhance productivity and efficiency in machinery and manufacturing facilities. Industry 4.0 operates in synergy with the IIoT and software/hardware systems, fostering connectivity that, through communication technologies , introduces enhancements and improves manufacturing processes. The creation of smarter, safer, and more advanced manufacturing capabilities is now attainable with these emerging technologies. It establishes a manufacturing platform that is more reliable, consistent, and efficient than ever before. The implementation of systems such as SCADA , a supervisory data collection software, exemplifies the tools utilized in Industrial Automation today, and it is just one of many such systems. Industry 4.0 encompasses and influences numerous aspects of manufacturing, and its reach is expected to continue expanding.

Industrial robotics

Automated milling machines.

Industrial robotics is a specialized branch within industrial automation that contributes to various manufacturing processes, including machining, welding, painting, assembly, and material handling. Industrial robots employ a combination of mechanical, electrical, and software systems to achieve high levels of precision, accuracy, and speed that far surpass human capabilities. The genesis of industrial robots occurred shortly after World War II, driven by the United States’ need for more efficient production methods for industrial and consumer goods. The development of servos, digital logic, and solid-state electronics enabled engineers to construct superior and faster systems. Over time, these systems were refined to the point where a single robot can operate continuously, 24 hours a day, with minimal maintenance. In 1997, there were an estimated 700,000 industrial robots in use worldwide; this number had risen to 1.8 million by 2017. In recent years, AI has been integrated with robotics to create automatic labeling solutions, utilizing robotic arms for label application and AI for product recognition and learning.

Programmable Logic Controllers

Industrial automation incorporates programmable logic controllers (PLCs) into the manufacturing process. PLCs utilize a processing system that allows for variations in the control of inputs and outputs through simple programming. PLCs employ programmable memory to store instructions and functions such as logic, sequencing, timing, and counting. Using a logic-based language, a PLC can receive a variety of inputs and generate a range of logical outputs, with input devices typically being sensors and output devices being motors, valves, and so forth. PLCs share similarities with computers; however, while computers are optimized for calculations, PLCs are designed for control tasks and operation in industrial environments. They are engineered to require only basic logic-based programming knowledge and to withstand vibrations, high temperatures, humidity, and noise. The primary advantage offered by PLCs is their flexibility. With the same basic controllers, a PLC can manage a variety of different control systems, eliminating the need for rewiring to change control system configurations. This flexibility results in cost-effective systems for complex and varied control requirements.

PLCs can range from compact “building brick” devices with a limited number of I/O points housed integrally with the processor, to large rack-mounted modular units with thousands of I/O points, often networked to other PLCs and SCADA systems. They can be designed for diverse arrangements of digital and analog inputs and outputs (I/O), extended temperature ranges, immunity to electrical noise , and resistance to vibration and impact. Programs to control machine operation are typically stored in battery-backed-up or non-volatile memory .

The PLC originated in the automotive industry in the United States. Prior to the PLC, the control, sequencing, and safety interlock logic for automobile manufacturing primarily consisted of relays , cam timers , drum sequencers , and dedicated closed-loop controllers. As these components could number in the hundreds or even thousands, updating such facilities for annual model change-over was a time-consuming and expensive process, as electricians had to individually rewire relays to alter their operational characteristics.

When digital computers became available as general-purpose programmable devices, they were soon applied to control sequential and combinatorial logic in industrial processes. However, these early computers required specialist programmers and stringent environmental controls for temperature, cleanliness, and power quality. To address these challenges, the PLC was developed with several key attributes: it would tolerate the shop-floor environment, support discrete (bit-form) input and output in an easily extensible manner, not require extensive training to use, and allow for operational monitoring. Given that many industrial processes operate on timescales easily managed by millisecond response times, modern (fast, small, reliable) electronics greatly facilitate the construction of reliable controllers, allowing performance to be traded off for reliability.

Agent-assisted automation

Agent-assisted automation refers to the use of automation tools by call center agents to handle customer inquiries. The primary benefit of agent-assisted automation lies in ensuring compliance and preventing errors. Agents may sometimes lack complete training, or they might forget or overlook crucial steps in a process. Automation guarantees that the intended actions are performed during a call, consistently and reliably. There are two main categories: desktop automation and automated voice solutions.

Control

This section may require cleanup. It has been merged from Automatic control .

Open-loop and closed-loop

Fundamentally, there are two types of control loops: open-loop control (feedforward), and closed-loop control (feedback).

  • In open-loop control, the controller’s action is independent of the “process output” (or “controlled process variable”). A common example is a central heating boiler controlled solely by a timer, providing heat for a fixed duration regardless of the building’s actual temperature. The control action is the boiler’s on/off state, but it does not directly regulate the building temperature because this is an open-loop control of the boiler, not a closed-loop control of the temperature.

  • In closed-loop control, the controller’s action is dependent on the process output. Using the boiler analogy, this would involve a thermostat to monitor the building temperature and feed back a signal to the controller, ensuring it maintains the building at the temperature set on the thermostat. A closed-loop controller thus incorporates a feedback loop that enables the controller to exert a control action to bring the process output into alignment with the “reference input” or “set point.” For this reason, closed-loop controllers are also known as feedback controllers.

The definition of a closed-loop control system, according to the British Standards Institution , is “a control system possessing monitoring feedback, the deviation signal formed as a result of this feedback being used to control the action of a final control element in such a way as to tend to reduce the deviation to zero.”

Similarly, “A Feedback Control System is a system which tends to maintain a prescribed relationship of one system variable to another by comparing functions of these variables and using the difference as a means of control.”

Discrete control (on/off)

On-off control is one of the simplest forms of control. A thermostat used in household appliances, which either opens or closes an electrical contact, serves as an example. (Thermostats were originally developed as true feedback-control mechanisms, unlike the common on-off thermostats found in household appliances.)

Sequence control involves performing a programmed sequence of discrete operations, often based on system logic that considers various system states. An elevator control system is a typical example of sequence control.

PID controller

A block diagram of a PID controller within a feedback loop, where r(t) represents the desired process value or “set point,” and y(t) is the measured process value.

A proportional–integral–derivative controller (PID controller) is a control loop feedback mechanism (a type of controller ) widely employed in industrial control systems .

Within a PID loop, the controller continuously calculates an error value, e(t), as the difference between a desired setpoint and a measured process variable . It then applies a correction based on three terms: proportional (P), integral (I), and derivative (D), which give the controller its name.

The theoretical understanding and practical application of PID controllers date back to the 1920s. They are implemented in virtually all analog control systems, initially in mechanical controllers, then using discrete electronics, and more recently in industrial process computers.

Sequential control and logical sequence or system state control

Sequential control can be either fixed, following a predetermined sequence, or logical, adapting actions based on various system states. A timer on a lawn sprinkler system is an example of an adjustable but otherwise fixed sequence.

System states refer to the different conditions that can occur within a usage scenario or sequence of operations for a system. An elevator, for instance, utilizes logic based on its system state to execute specific actions in response to inputs and its current condition. For example, if an operator presses the button for floor ’n’, the system’s response will depend on whether the elevator is stationary or moving, ascending or descending, or if the doors are open or closed, among other factors.

Early development in sequential control utilized relay logic , where electrical relays engaged electrical contacts to either initiate or interrupt power to a device. Relays were initially used in telegraph networks before being adapted for controlling other devices, such as starting and stopping industrial electric motors or opening and closing solenoid valves . Using relays for control allowed for event-driven actions, where operations could be triggered out of sequence in response to external events. This offered greater flexibility compared to the rigid, single-sequence cam timers . More complex applications involved maintaining safe operating sequences for devices like swing bridges, where a lock bolt had to be disengaged before the bridge could move, and the lock bolt could only be released after safety gates were closed.

The total number of relays and cam timers in some factories could reach into the hundreds or even thousands. Early programming techniques and languages, such as ladder logic (where diagrams of interconnected relays resembled ladder rungs), were developed to manage these complex systems. Subsequently, specialized computers known as programmable logic controllers (PLCs) were designed to replace these hardware assemblies with a single, more easily reprogrammable unit.

In a typical hard-wired motor start and stop circuit (referred to as a control circuit), a motor is initiated by pressing a “Start” or “Run” button, which activates a pair of electrical relays. The “lock-in” relay maintains the energization of the control circuit once the push-button is released. (The start button is typically a normally open contact, and the stop button is a normally closed contact.) Another relay energizes a switch that powers the device responsible for engaging the motor starter switch (which comprises three sets of contacts for three-phase industrial power) in the main power circuit. Large motors operate at high voltages and experience significant in-rush current, making the speed of making and breaking contact critical. This can pose safety risks to personnel and property with manual switches. The “lock-in” contacts in the start circuit and the main power contacts for the motor are held engaged by their respective electromagnets until a “stop” or “off” button is pressed, de-energizing the lock-in relay.

This state diagram illustrates how UML can be employed in the design of a door system with controlled opening and closing functionalities.

Interlocks are commonly incorporated into control circuits. For instance, if a motor powers machinery requiring critical lubrication, an interlock can be added to ensure the oil pump is operational before the motor starts. Timers, limit switches, and electric eyes are other frequently used components in control circuits.

Solenoid valves are widely used with compressed air or hydraulic fluid to power actuators on mechanical components. While motors provide continuous rotary motion , actuators are generally better suited for intermittently creating limited movements for mechanical components, such as actuating mechanical arms, opening or closing valves , raising heavy press-rolls, or applying pressure in presses.

Computer control

Computers are capable of performing both sequential and feedback control. In industrial applications, a single computer typically handles both functions. Programmable logic controllers (PLCs), a type of specialized microprocessor , have largely replaced numerous hardware components such as timers and drum sequencers previously used in relay logic systems. General-purpose process control computers have increasingly supplanted stand-alone controllers, with a single computer capable of managing the operations of hundreds of individual controllers. Process control computers can analyze data from a network of PLCs, instruments, and controllers to implement typical control strategies (such as PID) for numerous individual variables, or in some cases, to execute complex control algorithms utilizing multiple inputs and mathematical manipulations. They can also analyze data, generate real-time graphical displays for operators, and produce reports for operators, engineers, and management.

The control of an automated teller machine (ATM) serves as an example of an interactive process where a computer executes a logic-derived response to user selections based on information retrieved from a networked database. This ATM process shares similarities with other online transaction processes. The various logical responses are referred to as scenarios. Such processes are typically designed using use cases and flowcharts , which guide the development of the software code. The earliest known feedback control mechanism was the water clock, invented by the Greek engineer Ctesibius (285–222 BC).

See also