Robotics

Summary

Robotics is an interdisciplinary scientific field concerned with the design, development, operation, and assessment of electromechanical devices used to perform tasks that would otherwise require human action. Robotics applications can be found in almost every arena of modern life. Robots, for example, are widely used in industrial assembly lines to perform repetitive tasks. They have also been developed to help physicians perform difficult surgeries and are essential to the operation of many advanced military vehicles. Among the most promising robot technologies are those that draw on biological models to solve problems, such as robots whose limbs and joints are designed to mimic those of insects and other animals.

Definition and Basic Principles

Robotics is the science of robots—machines that can be programmed to carry out a variety of tasks independently, without direct human intervention. Although robots in science fiction tend to be androids or humanoids (robots with recognizable human forms), most real-life robots, especially those designed for industrial use, do not resemble humans physically. Robots typically consist of at least three parts: a mechanical structure (most commonly a robotic arm) that enables the robot to physically affect either itself or its task environment; sensors that gather information about physical properties such as sound, temperature, motion, and pressure; and some kind of processing system that transforms data from the robot's sensors into instructions about what actions to perform. Some devices, such as the search-engine bots that mine the internet daily for data about links and online content, lack mechanical components. However, they are nevertheless often considered robotic because they can perform repeated tasks without supervision.

Many robotics applications also involve the use of artificial intelligence. This is a complex concept with a shifting definition, but in its most basic sense, a robot with artificial intelligence possesses features or capabilities that mimic human thought or behavior. For example, one aspect of artificial intelligence involves creating parallels to the human senses of vision, hearing, or touch. The friendly voices at the other ends of customer-service lines, for example, are increasingly likely to be robotic speech-recognition devices capable not merely of hearing callers' words but also of interpreting their meanings and directing the customers' calls intelligently.

More advanced artificial intelligence applications give robots the ability to assess their environmental conditions, make decisions, and independently develop efficient plans of action for their situations—and then modify these plans as circumstances change. Chess-playing robots do this each time they assess the state of the chessboard and make a new move. The ultimate goal of artificial intelligence research is to create machines whose responses to questions or problems are so humanlike as to be indistinguishable from those of human operators. This standard is the so-called Turing test, named after the British mathematician and computing pioneer Alan Turing.

Background and History

The word "robot" comes from a Czech word for “forced labor” that the Czech writer Karel Čapek used in his 1921 play R.U.R. about a man who invents a humanlike automatic machine to do his work. During the 1940s, as computing power grew, the influential science-fiction writer Isaac Asimov began applying "robotics" to the technology behind robots. The 1950s saw the development of the first machines that could be called robots. These prototypes took advantage of such new technologies as transistors (compact, solid-state devices that control electrical flow in electronic equipment) and integrated circuits (complex systems of electronic connections stamped onto single chips) to enable more complicated mechanical actions. In 1959, an industrial robot was designed that could churn out ashtrays automatically. Over the ensuing decades, public fascination with robots expanded far beyond their actual capabilities. It was becoming clear that creating robots that could accomplish seemingly simple tasks—such as avoiding obstacles while walking—was a surprisingly complex problem.

During the late twentieth century, advances in computing, electronics, and mechanical engineering led to rapid progress in the science of robotics. These included the invention of microprocessors, single integrated circuits that perform all the functions of computers' central processing units, production of better sensors and actuators, and developments in artificial intelligence and machine learning, such as a more widespread use of neural networks. (Machine learning is the study of computer programs that improve their performance through experience.)

Cutting-edge robotics applications are continuously being developed by an interdisciplinary research cohort of computer scientists, electrical engineers, neuroscientists, psychologists, and others. These applications combine greater mechanical complexity with more subtle information processing systems than were once possible. Homes may not be populated with humanoid robots who can hold conversations, but mechanical robots have become ubiquitous in industry. Also, unpiloted robotic vehicles and planes are essential in warfare, search-engine robots crawl the World Wide Web every day collecting and analyzing data about internet links and content, and robotic surgical tools are indispensable in health care. All this is evidence of the extraordinarily broad range of problems robotics addresses.

How It Works

Sensing. Robots must gather as much information as possible about the physical features of their environments in order to move and react to conditions. They do so through a large array of sensors designed to monitor different physical properties. Simple touch sensors consist of electric circuits that are completed when levers receive enough pressure to press down on switches. Robotic dogs designed as toys, for example, may have touch sensors in their backs or heads to detect when they are being petted and signal them to respond accordingly. More complex tactile sensors can detect properties such as torque (rotation) or texture. Such sensors may be used, for example, to help an assembly-line robot's end effector control the grip and force it uses to turn an object it is screwing into place.

Light sensors consist of one or more photocells that react to visible light with decreases in electrical resistance. They may serve as primitive eyes, allowing unpiloted robotic vehicles, for example, to detect the bright white lines that demarcate parking spaces and maneuver between them. Reflectance sensors emit beams of infrared light, measuring the amounts of that light that reflects from nearby surfaces. They can detect the presence of objects in front of robots and calculate the distances between the robots and the objects, allowing the robots to either follow or avoid the objects. Temperature sensors rely on internal thermistors (resistors that react to high temperatures with decreases in electrical resistance). Robots used to rescue human beings trapped in fires may use temperature sensors to navigate away from areas of extreme heat. Similarly, altimeter sensors can detect changes in elevation, allowing robots to determine whether they are moving up or down slopes.

Other sensor types include magnetic sensors, sound sensors, accelerometers, and proprioceptive sensors that monitor the robots' internal systems and tell them where their own parts are located in space. After robots have collected information through their sensors, algorithms (mathematical processes based on predefined sets of rules) help them process that information intelligently and act on it. For example, a robot may use algorithms to help it determine its location, map its surroundings, and plan its next movements.

Motion and Manipulation. Robots can be made to move around spaces and manipulate objects in many different ways. At the most basic level, a moving robot needs to have one or more mechanisms consisting of connected moving parts, known as links. Links can be connected by prismatic or sliding joints, in which one part slides along the other, or by rotary or articulated joints, in which both parts rotate around the same fixed axis. Combinations of prismatic and rotary joints enable robotic manipulators to perform a host of complex actions, including lifting, turning, sliding, squeezing, pushing, and grasping. Actuators are required to move jointed segments or robot wheels. Actuators may be electric or electromagnetic motors, hydraulic gears or pumps (powered by compressed liquid), or pneumatic gears or pumps powered by pressurized gas. Electric circuits control the actuators to coordinate the robots' movements.

Motion-description languages are a type of computer programming language designed to formalize robot motions. They consist of sets of symbols that can be combined and manipulated in different ways to identify whole series of predefined motions in which robots of specified types can engage. Motion-description languages were developed to simplify the process of manipulating robot movements by allowing different engineers to reuse common sets of symbols to describe actions or groups of actions rather than having to formulate new algorithms to describe every individual task they want robots to perform.

Control and Operation. A continuum of robotic control systems ranges from fully manual operation to fully autonomous operation. On the one hand, a human operator may be required to direct every movement a robot makes. For example, some bomb disposal robots are controlled by human operators working only a few feet away, using levers and buttons to guide the robots as they pick up and remove the bombs. On the other side of the spectrum are robots that operate with no human intervention at all, such as the KANTARO—a fully autonomous robot that slinks through sewer pipes, inspecting them for damage and obstructions. Many robots have control mechanisms lying somewhere between these two extremes.

Robots can also be controlled from a distance. Teleoperated systems can be controlled by human operators situated either a few centimeters away, as in robotic surgeries, or millions of miles away, as in outer space applications. "Supervisory control" is a term given to teleoperation in which the robots themselves can perform the vast majority of their tasks independently; human operators are present merely to monitor the robots' behavior and occasionally offer high-level instructions.

Artificial Intelligence. Three commonly accepted paradigms, or patterns, are used in artificial intelligence robotics: hierarchical, reactive, and hybrid. The hierarchical paradigm, also known as a top-down approach, organizes robotic tasks in sequence. For example, a robot takes stock of its task environment, creates a detailed model of the world, uses that model to plan a list of tasks it must carry out to achieve a goal and proceeds to act on each task in turn. The performance of hierarchical robots tends to be slow and disjointed since every time a change occurs in the environment, the robot pauses to reformulate its plan. For example, if such a robot is moving forward to reach a destination and an obstacle is placed in its way, it must pause, rebuild its model of the world, and begin lurching around the object.

In the reactive (or behavioral) paradigm, also known as a bottom-up approach, no planning occurs. Instead, robotic tasks are carried out spontaneously in reaction to a changing environment. If an obstacle is placed in front of such a robot, sensors can quickly incorporate information about the obstacle into the robot's actions and alter its path, causing it to swerve momentarily.

The hybrid paradigm is the one most commonly used in artificial intelligence applications being developed during the twenty-first century. It combines elements of both the reactive and the hierarchical models.

Applications and Products

Industrial Robots. In the twenty-first century, almost no factory operates without at least one robot—more likely several—playing some part in its manufacturing processes. Welding robots, for example, consist of mechanical arms with several degrees of movement and end effectors in the shape of welding guns or grippers. They are used to join metal surfaces together by heating and then hammering them, and produce faster, more reliable, and more uniform results than human welders. They are also less vulnerable to injury than human workers. Another common industrial application of robotics is silicon-wafer manufacturing, which must be performed within meticulously clean rooms so as not to contaminate the semiconductors with dirt or oil. Humans are far more prone than robots to carry contaminants on them.

Six major types of industrial robots are defined by their different mechanical designs. Articulated robots are those whose manipulators (arms) have at least three rotary joints. They are often used for vehicle assembly, die casting (pouring molten metal into molds), welding, and spray painting. Cartesian robots, also known as gantry robots, have manipulators with three prismatic joints. They are often used for picking objects up and placing them in different locations, or for manipulating machine tools. Cylindrical robots have manipulators that rotate in a cylindrical shape around a central vertical axis. Parallel robots have both prismatic and rotary joints on their manipulators. Spherical robots have manipulators that can move in three-dimensional spaces shaped like spheres. SCARA (Selective Compliant Assembly Robot Arm) robots have two arms connected to vertical axes with rotary joints. One of their arms has another joint that serves as a wrist. SCARA robots are frequently used for palletizing (stacking goods on platforms for transportation or loading).

Service Robots. Unlike industrial robots, service robots are designed to cater to the needs of individual people. Robopets, such as animatronic dogs, provide companionship and entertainment for their human owners. The Sony Corporation's AIBO (Artificial Intelligence roBOt) robopets use complex systems of sensors to detect human touch on their heads, backs, chins, and paws, and can recognize the faces and voices of their owners. They can also maintain their balance while walking and running in response to human commands. AIBOs also function as home-security devices, as they can be set to sound alarms when their motion or sound detectors are triggered. Consumer appliances, such as iRobot Corporation's robotic vacuum cleaner, the Roomba, and the robotic lawn mover, the RoboMower, developed by Friendly Robotics, use artificial intelligence approaches to safely and effectively maneuver around their task environments while performing repetitive tasks to save their human users time.

Even appliances that do not much resemble public notions of what robots should look like often contain robotic components. For example, digital video recorders (DVRs) such as TiVos contain sensors, microprocessors, and a basic form of artificial intelligence that enable them to seek out and record programs that conform to their owners' personal tastes. Some cars can assist their owners with driving tasks such as parallel parking and lane-keeping assistance. The auto industry has continued to explore the reality of autonomous cars (self-driving vehicles), with the goal of making mass-produced models available to the public that would function flawlessly on existing roads. In 2020, the technology company Google was testing a ride-sharing system called Waymo that used self-driving vehicles. In 2023, Mercedes-Benz introduced the first Level 3 system, Drive Pilot, for sale in the United States. Level 3, which is a driver-assistance device, requires state approval. Tesla also released a similar model, the Autopilot driver-assistance system, and in 2024, after much beta-testing, released The Full Self-Driving system, which built off Tesla’s standard Autopilot driver-assistance system. Full Self-Driving is a monthly paid subscription available to Tesla owners.

Many companies or organizations rely on humanoid robots to provide services to the public. The Smithsonian National Museum of American History, for example, has used an interactive robot named Minerva to guide visitors around the museum's exhibits, answering questions and providing information about individual exhibits. Other professional roles filled by robots include those of receptionists, floor cleaners, librarians, bartenders, and secretaries. At least one primary school in Japan even experimented with a robotic teacher developed by a scientist at the Tokyo University of Science. However, an important pitfall of humanoid robots is their susceptibility to the uncanny valley phenomenon. This is the theory that as a robot's appearance and behavior become more humanlike, people will respond to it more positively—but only up to a point. On a line graph plotting positive response against degree of human likeness, the response dips (the "uncanny valley") as the likeness approaches total realism but does not perfectly mimic it. In other words, while people will prefer a somewhat anthropomorphic robot to an industrial-looking one, a highly humanlike robot that is still identifiably a machine will cause people to feel revulsion and fear rather than empathy.

Medical Uses.Robotic surgery has become an increasingly important area of medical technology. In most robotic surgeries, a system known as a master-slave manipulator is used to control robot movements. Surgeons look down into electronic displays showing their patients' bodies and the robots' tool tips. The surgeons use controls attached to consoles to precisely guide the robots' manipulators within the patients' bodies. A major benefit of robotic surgeries is that they are less invasive—smaller incisions need to be made because robotic manipulators can be extremely narrow. These surgeries are also safer because robotic end effectors can compensate for small tremors or shakes in the surgeons' movements that could seriously damage their patients' tissues if the surgeons were making the incisions themselves. Teleoperated surgical robots can even allow surgeons to perform operations remotely without the need to transport patients over long distances. Surgical robots such as the da Vinci system are used to conduct operations such as prostatectomy, cardiac surgery, bariatric surgery, and various forms of neurosurgery.

Humanoid robots are also widely used as artificial patients to help train medical students in diagnosis and procedures. These robots have changing vital signs such as heart rates, blood pressure, and pupil dilation. Many are designed to breathe realistically, express pain, urinate, and even speak about their conditions. With their help, physicians-in-training can practice drawing blood, performing cardiopulmonary resuscitation (CPR), and delivering babies without the risk of harm to real patients.

Other medical robots include robotic nurses that can monitor patients' vital signs and alert physicians to crises and smart wheelchairs that can automatically maneuver around obstacles. Scientists have developed nanorobots the size of bacteria that can be swallowed and sent to perform various tasks within human bodies, such as removing plaque from the insides of clogged arteries.

Robot Exploration and Rescue. One of the most intuitive applications of robotic technology is the concept of sending robots to places too remote or too dangerous for human beings to work in—such as outer space, great ocean depths, and disaster zones. The six successful crewed moon landings of the Apollo program carried out during the late 1960s and early 1970s are dwarfed in number by the uncrewed robot missions that have set foot not only on the moon but also on other celestial bodies, such as planets in the solar system. The wheeled robots Spirit and Opportunity, for example, began analyzing material samples on Mars and sending photographs back to Earth in 2004. Roboticists have also designed biomimetic robots inspired by frogs that take advantage of lower gravitational fields, such as those found on smaller planets, to hop nimbly over rocks and other obstacles. In 2020 the wheeled robot Perseverance landed on Mars to find signs of ancient life and return samples of rock and soil to Earth.

Robots are also used to explore the ocean floor. The Benthic Rover, for example, drags itself along the seabed at depths up to 2.5 miles below the surface. It measures oxygen and food levels, takes soil samples, and sends live streaming video up to the scientists above. The rover is operated by supervisory control and requires very little intervention on the part of its human operators.

Rescue robots seek out, pick up, and safely carry injured humans trapped in fires, under rubble, or in dangerous battle zones. For example, the US Army's Bear (Battlefield Extraction-Assist Robot) is a bipedal robot that can climb stairs, wedge itself through narrow spaces, and clamber over bumpy terrain while carrying weights of up to three hundred pounds.

Military Robots. Militaries have often been among the leading organizations pioneering robotics, as robots have the potential to complete many military tasks that might otherwise prove dangerous to humans. While many projects, such as bomb-removal robots, have proven highly useful and have been widely accepted, other military applications have proven more controversial. For example, many observers and activists have expressed concern over the proliferation of drones, particularly unpiloted aerial vehicles (UAVs) capable of enacting military strikes such as rocket or missile launches while going virtually undetected by radar. The US government has used such technology (including the Predator drone) to successfully destroy terrorist positions, including in remote territory that would otherwise be difficult and dangerous to access, but there have also been notable examples of misidentified targets and civilian collateral damage caused by drone strikes. Opponents of drones argue that the potential for mistakes or abuse of their capabilities is dangerous.

Careers and Course Work

Courses in advanced mathematics, physics, computer science, electrical engineering, and mechanical engineering make up the foundational requirements for students interested in pursuing careers as robotics engineers. Earning a bachelor of science degree in any of these fields is an appropriate preparation for graduate work in a similar area. In most circumstances, either a master's degree or a doctorate is a necessary qualification for the most advanced future career opportunities in both academia and industry. However, an advanced degree is not generally required for a career as a robotics technician—someone who maintains and repairs robots rather than designs them.

Students should take substantial course work in more than one of the primary fields of study related to robotics (physics, mathematics, computer science, and engineering) because designing and testing robots requires skills drawn from multiple disciplines. In addition, anyone desiring to work in robotics should possess skills that go beyond the academic, including good physical coordination, excellent manual dexterity, and an aptitude for mechanical details. A collaborative mindset is also an asset, as robotics work tends to be done by teams.

Careers in the field of robotics can take several different shapes. The manufacturing industry is the biggest employer of robotics engineers and technicians. Within this industry, robotics professionals might focus on developing, maintaining, or repairing production-line robots used in factory assembly. Other industries in which robotics engineers and technicians often find work include aviation, agriculture, nuclear energy, telecommunications, electronics, mining, health care and medicine, and education.

Many roboticists prefer employment within academic settings. Such professionals divide their time between teaching university classes on robotics and conducting their own research projects. Others find work in government agencies such as the National Aeronautics and Space Administration (NASA) and DARPA, focusing on large-scale robotics applications in such areas as space exploration, warfare, and disaster management.

Social Context and Future Prospects

In the twenty-first century, the presence of robots in factories all over the world is taken for granted. Meanwhile, robots are also increasingly entering daily life in the form of automated self-service kiosks at supermarkets, electronic lifeguards that detect when swimmers in pools are struggling, and cars whose robotic speech-recognition software enables them to respond to verbal commands. A science-fiction future in which ubiquitous robotic assistants perform domestic tasks such as cooking and cleaning may not be far away, but technological limitations must be overcome for that to become a reality.

However, advancements in the natural language processing (NLP) field have demonstrated the capabilities of robots in daily life. For example, the voice-activated personal assistant software in smartphone and computer operating systems, such as Apple's Siri, can assist in making calls, looking up information, and programming one's personal calendar. Another example is OpenAI's chat GPT, an artificial intelligence chatbot that uses NLP to generate human-like dialogue and respond to user questions. Though the program has proven problematic in academia, raising questions about plagiarism and students' original ideas and work, it has helped countless industries speed productivity.

Robots that provide nursing care or companionship to the infirm are not merely becoming important parts of the health care industry but may also provide a solution to the problem increasingly faced by countries in the developed world—a growing aging population who need more caretakers than can be found among younger adults. Another area where robots can be particularly useful is in performing dangerous tasks that would otherwise put a human's life at risk. The use of robotics in the military remains a growing field—not only in the controversial use of combat drones, but also for tasks such as minesweeping. Robots are also more and more heavily used in space exploration, and robots have been used in other dangerous fields as well; during the coronavirus disease 2019 (COVID-19) pandemic declared in 2020, robots were used to disinfect contaminated areas, take the temperature of patients, and monitor mask-wearing and social distancing among citizens. The robotics company Boston Dynamics has developed prototypes of extremely human- and animal-like robots with arms and legs that can walk, jump, climb stairs, and even dance. Some of the company's robots are used by utility companies, construction sites, and police departments.

There are also concerns about the growing use of robots to perform tasks previously performed by humans, however. As robotics technology improves and becomes less expensive, companies may well turn to cheap, efficient robots to do jobs that are typically performed by immigrant human labor, particularly in such areas as agriculture and manufacturing. Meanwhile, some observers are concerned that the rise of industrial and professional service robots is already eliminating too many jobs held by American workers. Many of the jobs lost in the 2008–09 recession within the struggling automotive industry never came back because costly human workers were replaced by cheaper robotic arms. However, the issue is more complicated than that. In certain situations, the addition of robots to a factory's workforce can actually create more jobs for humans. Some companies, for example, have been able to increase production and hire additional workers with the help of robot palletizers that make stacking and loading their products much faster.

Safety concerns can sometimes hinder the acceptance of new robotic technologies, even when they have proven to be less likely than humans to make dangerous mistakes. Robotic sheep shearers in Australia, for example, have met with great resistance from farmers because of the small risk that the machines may nick a major artery as they work, causing the accidental death of a sheep. And while fully autonomous vehicles, including self-driving cars, are seen by many automotive and technology companies alike as a major area of innovation, crashes and other accidents by several prototypes in the 2010s and 2020s drew concerns from regulators and the general public. It is critical in the field of robotics to not only develop the technology necessary to make a design a reality, but to understand the cultural and economic landscape that will determine whether a robot is a success or failure.

Bibliography

Clarke, Warren, and Cherise Threewitt. “What Does Tesla's Full-Self Driving Mode Do?” US News, 4 Mar. 2024, cars.usnews.com/cars-trucks/advice/tesla-full-self-driving. Accessed 7 June 2024.

Faust, Russell A., ed. Robotics in Surgery: History, Current and Future Applications. New York: Nova Science, 2007. Print.

Floreano, Dario, and Claudio Mattuissi. Bio-Inspired Artificial Intelligence: Theories, Methods, and Technologies. Cambridge: MIT P, 2008. Print.

Gutkind, Lee. Almost Human: Making Robots Think. 2006. Reprint. New York: Norton, 2009. Print.

“Introducing ChatGPT.” OpenAI, 30 November 2022, openai.com/index/chatgpt/. Accessed 7 June 2024.

Jones, Joseph L. Robot Programming: A Practical Guide to Behavior-Based Robotics. New York: McGraw-Hill, 2004. Print.

"Our Friends Electric." Economist. Economist Newspaper, 7 Sept. 2013. Web. 14 Nov. 2014.

Popovic, Marko B. Biomechanics and Robotics. Boca Raton: CRC, 2013. Print.

Niku, Saeed. Introduction to Robotics: Analysis, Control, Applications. 3rd ed., Wiley, 2020.

Samani, Hooman. Cognitive Robotics. CRC Press, 2016.

Siciliano, Bruno, et al. Robotics: Modeling, Planning, and Control. London: Springer, 2009. Print.

Siciliano, Bruno, and Oussama Khatib, eds. The Springer Handbook of Robotics. Springer, 2016.

Springer, Paul J. Military Robots and Drones: A Reference Handbook. Santa Barbara: ABC-CLIO, 2013. Print.

Tucker, Ian. "The Five: Robots Helping to Tackle Coronavirus." The Guardian, 31 May 2020, www.theguardian.com/technology/2020/may/31/the-five-robots-helping-to-tackle-coronavirus. Accessed 28 Feb. 2022.

Tucker, Sean. "Self-Driving Cars: Everything You Need to Know." Kelley Blue Book, 3 Aug. 2021, www.kbb.com/car-advice/self-driving-cars/. Accessed 28 Feb. 2022.