Robot

A robot is a programmable machine capable of completing one or more tasks. Common robot functionality includes the ability to sense its environment, perform computations, make decisions, and carry out actions in the real world. While robots typically reduce or remove the need for human intervention, the level of autonomy associated with each machine varies considerably. It ranges from human-controlled bots to fully autonomous machines able to perform tasks without any human interactions. The word "robot" is derived from the Czech word robota, meaning “drudgery” or “servitude.” It was coined in the 1920 play called Rossum’s Universal Robots, by Czech writer Karel Čapek.

Robotics, an interdisciplinary field dedicated to the design, construction, and use of robots, has advanced considerably in the last fifty years. With the rise of new technologies, such as artificial intelligence (AI) and enhanced manufacturing techniques, robots have become more efficient and capable of more complex tasks. Robots can perform a number of tasks better than humans, including the automation of manual or repetitive activities that can lead to injuries when performed manually or tasks that human workforces are incapable of performing due to size limitations or hazardous environments.

Robotics is a diverse field, including mechanical robots and software-based robots in fields such as robotic process automation (RPA) that automate specific business processes with the aim of streamlining operations. Common characteristics of mechanical robots include electrical components controlled by programable computer systems. The main components of a mechanical robot can be separated into the following:

• Control system
• Power supply
• Sensors
• Actuators
• End effectors

Robots have a large number of use cases across many industries, including healthcare, space exploration, agriculture, and the smart home. Robots are perhaps most well-known for use in manufacturing. From 2012 to 2022, the global sales volume of industrial robots tripled, peaking in 2018 at roughly 422,000 units. This growth is largely driven by automotive and electronics manufacturing, which accounted for 23% and 31% of new installations, respectively, in 2020. These drivers mean the prevalence of industrial robotics is higher in countries with active automotive and electronics manufacturing, such as Japan, China, South Korea, Germany, and the United States. South Korea has 930 robots for every 10,000 manufacturing employees—seven times higher than the global average.

The word "robot" dates back to the early twentieth century and brothers Karel and Josef Čapek. Against the backdrop of Prague's difficult and overcrowded working and living conditions, playwright Karel began to write a story inspired by the idea of people no longer herded like sheep but machines, only able to work and unable to reason. The play's story concerned the manufacture of artificial people from human-made organic material, freeing humans of their work and drudgery and finally leading to humankind's destruction after overproduction. When developing a name for these artificial workers, he turned to his brother Josef, who suggested he call them "robots" from the Czech word robota, meaning hard work or forced labor.

In the summer of 1920, Karel finished the play titled Rossum's Universal Robots (RUR). It was published in November 1920 and first performed in January 1921. The word "robot" quickly gained popularity in many languages as an expression of artificial intelligence machines invented by humans.

The study of artificial machines and robots dates back to ancient civilizations such as the Greeks, Romans, and Egyptians. In 320 BC, the Greek philosopher Aristotle wrote:

If every tool, when ordered, or even of its own accord, could do the work that befits it… then there would be no need either of apprentices for the master workers or of slaves for the lords.

The first biomechanical automation on record occurred in 1737 when French inventor Jacques de Vaucanson developed the "Flute Player," a mechanical device capable of playing twelve songs. Vaucanson also completed the canard digérateur," or "digesting duck," in 1738. A life-sized duck made of gilded copper sitting atop a large plinth, the machine reportedly contained hundreds of moving parts, with much of its workings located inside the plinth. The duck performed several animatronic actions, including waddling, flapping its wings, drinking water, making duck noises, and even "eating," "digesting," and "defecating." In reality, Vaucanson's duck produced droppings that were soggy breadcrumbs stored internally in a separate container. The original duck remained in private collections until it was destroyed in a museum fire in 1879.

Modern reconstruction of Vaucanson's digesting duck.

In 1936, Alan Turing published the paper "On Computable Numbers," introducing the concept of a theoretical computer named the Turing Machine. "Cybernetics or Control and Communication in the Animal and the Machine" was published in 1948 by MIT professor Norbert Wiener. The book focuses on the concept of communication and control in electronic, mechanical, and biological systems. In 1949, the neurophysiologist and inventor William Grey Walter developed a pair of battery-operated robots that looked like tortoises. The robots could move objects, locate a source of light, and find their way back to a charging station. 1950 saw two key publications: Isaac Asimov's "Three Laws of Robotics" and Alan Turing's paper "Computing Machinery and Intelligence," in which he proposed a method to determine if a machine is intelligent, now known as the Turing Test.

The first industrial robots appeared in 1954. Barrett Electronics produced the first electric vehicle, which no longer required a human driver, operating as an Autonomous Guided Vehicle (AGV). General Motors purchased Unimate, a hydraulic arm capable of lifting heavy loads, considered to be the first industrial robot developed in the USA. In the following years, several versions of the hydraulic arm were developed by George Devol and his partner, Joseph Engelberger, becoming introduced into General Motors facilities in 1961. The arm was used to lift and stack metal parts.

The first mobile robot, known as "Shakey," was developed at the Stanford Research Institute (SRI) Artificial Intelligence Center from 1966 to 1972. Shakey had the ability to perceive its surroundings and perform tasks that required planning, route-finding, and the rearranging of simple objects.

Shakey, the first mobile robot developed at SRI from 1966 to 1972.

In 1969, Victor Scheinman, a mechanical engineering student working in the Stanford Artificial Intelligence Lab (SAIL), designed the Stanford Arm. The robotic arm, one of the first to be controlled by a computer, had six joints and was capable of mimicking the movements of a human arm. In 1978, Professor Hiroshi Makino of Yamanashi University, Kofu, Japan, created the first Selective Compliance Assembly Robot Arm (SCARA) prototype. A pick-and-place robotic arm, the four-axis programmable SCARA robot, could pick up objects before turning and placing them in another location. First introduced to commercial assembly lines in 1981, SCARA robots remain in use. 1985 saw the first documented robot-assisted surgical procedure occur (PUMA 560 robotic surgical arm), and two remotely-operated robots were used to survey the site following the Three Mile Island nuclear power plant accident in 1979.

The early 1990s saw the founding of a number of important robotics companies, including iRobot in 1990 (the company behind the Roomba vacuum cleaner) and Boston Dynamics in 1992. In 1997, the Sojourner rover (part of NASA's Pathfinder mission) landed on Mars. The free-ranging rover returned considerable data, including 17,000 images and chemical analyses of martian soil and rocks.

In 2000, Cynthia Brazeal created a robotic head, called Kismet, programmed to provoke and react to emotions. The Roomba vacuum robot was created by iRobot in 2002, going on to become one of the first robots to find success in the commercial sector. In 2003, Kiva systems (now Amazon Robotics) invented the Kiva robot, capable of maneuvering around warehouses transporting goods. 2004 saw Boston Dynamics unveil BigDog, a quadruped robot controlled by humans. More nimble than previous iterations, the robot contained fifty sensors and an onboard computer to manage the gait and keep it stable.

In 2011, NASA and General Motors collaborated to send Robonaut 2, a human-like assistant to space on board the space shuttle Discover. Robonaut 2 became a resident onboard the International Space Station. The first license for a self-driving car was issued in 2012 in Nevada. The license was for a Toyota Prius modified with Google technology. Sophia, a humanoid robot from Hanson Robotics, was unveiled in 2016. The robot is capable of facial recognition, expressions, and verbal communication.

Sophia, a humanoid robot from Hanson Robotics, first unveiled in 2016.

Given robotics is a large and varied field, it is possible to divide robots into categories based on a number of factors (use case, autonomy, price, fixed/mobile, form factor, etc.). Common types of robots include those below:

AMRs are defined by their ability to understand and move through their surroundings without direct human oversight or the use of fixed/predetermined paths. To do this, AMRs employ sophisticated sensors and onboard processing units that enable them to map their environments (both fixed obstructions and variable obstacles) and plan their tasks in an efficient manner. The output of these sensors is generally fed into a warehouse's control system such that multiple AMRs can effectively coordinate and enhance the flexibility of the facility's operations. AMRs perform critical tasks that do not require human creativity. These non-value-adding tasks include transporting, picking up, and dropping off products or picking, checking, and packing an order. Typically AMRs can be divided into three main use cases:

• Moving inventory within a facility
• Assisting in the picking process
• Flexible sortation solutions

AMRs are used across numerous industries, including warehouses, logistical companies, agriculture businesses, and healthcare institutions. When implemented successfully they can improve productivity safely.

Example of an AMR.

AGVs are material handling solutions used to autonomously transport goods through manufacturing facilities, warehouses, and distribution centers. While AMRs have the ability to freely move through environments, AGVs rely on predefined paths or tracks and can require additional operator oversight. Therefore, AMRs offer greater flexibility and efficiency, as they can change their path in real time and recalculate a more efficient route based on new circumstances.

AGVs are driverless vehicles with onboard technology guiding their movements along predefined paths. There are many ways AGVs navigate through a site; two of the most popular are the following:

• Reflector navigation—Reflectors are installed within a facility and scanned by each automated vehicle, allowing it to determine its position based on the distance to the reflectors. This method of navigation offers high accuracy and robustness.
• Natural navigation—Reference points are used, such as walls, racking, and fixed objects to calculate location. It is a common navigation method for warehouses or distribution centers with constant internal setups.

AGVs also include built-in safety scanners, obstacle detection sensors, load sensors, and vision cameras to improve safety.

Example of an articulated robot.

Also known as robotic arms, articulated robots can have anywhere from two to ten rotary joints, with each additional joint or axis adding a greater degree of motion. Most articulated industrial robots have four to six axes, six being the most common. Articulated robots start with a base, vertical to the ground containing the first joint. The main robot body is connected to the base through the first revolute joint. A second joint runs perpendicular to the body, connecting the shoulder to the body. At the end of the shoulder, a parallel revolute joint attaches the shoulder to the robot arm. Additional joints are then used at the end of the robot arm to attach to the end effector.

This design closely mimics the human arm. Servo motors are used to power the fast and precise movements of each axis. Articulated robots are commonly used for arc welding, material handling, machine tending, and packaging.

Humanoid robots describe machines that perform human-centric functionality, often with human-like appearances. Many humanoid robots could also be described as AMRs, as they use technology to sense, plan, and carry out their tasks. Humanoids are used to perform human tasks with greater speed and efficiency, sometimes even interacting in a similar manner as humans. Examples of humanoid use cases include personal assistance/caregiving, education/entertainment, search and rescue, and manufacturing.

Cobots, or collaborative robots, are designed to operate alongside or directly with human workers. While many industrial robots work separately from human workers, cobots are designed to share the same space helping them accomplish more. Often cobots are used to eliminate repetitive manual tasks that may be strenuous or difficult for humans to perform on a daily basis. Cobots are designed to automate these types of tasks. There is significant crossover between industrial robots and cobots; however, cobots have unique features to enable collaboration with humans.

Hybrid robots are the combination of various types to create a new solution more capable for a specific workplace. An example could be an AMR with a robotic arm to handle and transport packages within a warehouse. To combine functionality, companies also have to consider combining the software used to operate the hybrid solution.

Also called bots, software robotics describes computer programs that perform tasks with a level of autonomy. They are not physical, mechanical robots with a body or chassis, instead existing online. Common examples include the following:

• Chatbots
• Spam bots
• Search engine crawler bots
• Monitoring bots

It is possible to divide robots based on their level of autonomy with the two extremes being fully dependent or independent robots. Dependent robots are non-autonomous, requiring human interaction to enhance and supplement their actions. Examples of robots that require human intervention are pre-programmed robots that operate in a controlled environment performing simple, monotonous tasks. This includes articulated robots on an assembly line that are programmed to perform a specific task longer, faster, and more efficiently than a human. Other examples of semi-autonomous machines include teleoperated robots that are controlled by humans from a safe distance. Teleoperated robots often work in extreme conditions and circumstances.

Independent or autonomous robots function without the need for human operator control. They require sophisticated control systems and technology to understand the task being performed and the environment in which it is taking place. They often perform tasks in open environments that do not require human supervision. Autonomous robots use sensors to perceive the world around them, then apply decision-making algorithms to take the optimal next step based on their data and mission. Independent robots have disrupted many sectors, eliminating jobs previously performed by human workers. Examples of autonomous robots include those below:

• Cleaning bots
• Lawn trimming bots
• Hospitality bots
• Autonomous drones
• Medical assistant bots

Mechanical robots can contain a wide range of components:

The control system, the robot's central processing unit, includes all the components that determine the machine's actions. Control systems are programmed to decide how the robot utilizes its specific components to perform the required task.

The power supply provides the energy the robot requires to function. While stationary robots may connect to a wall outlet and run on AC power, most robots have some form of internal power supply or battery. Factors to consider when designing robotic power supplies include safety, weight, lifecycle, maintenance, and replaceability.

Sensors provide the data robots need to process to interact with their environment. This could be a camera feed, photoresists that measure light, or microphones. A robot's sensors produce electrical signals that are fed into the control system to process and find the most logical action to perform.

Mechanical robots have movable components requiring actuators. These components translate signals from the control system into movements to complete the assigned task.

End effectors are physical, generally external components that allow robots to perform their role. End effectors could consist of interchangeable tools based on their current tasks.

Probably the oldest and most well-known robot use case is manufacturing. The accelerated use of robotics and other technology to automate and optimize manufacturing techniques is referred to as industry 4.0. This includes the better use of manufacturing data, AI algorithms to improve decision-making and enhance processes, and robotics to automate manufacturing and logistical tasks. Robots for manufacturing and logistics perform a number of roles:

• Material handling
• Pick-and-place
• Material dispensing
• Material removal
• Palletization and de-palletizing
• Welding
• Assembly

Generally speaking, robotics are used for two main applications: automating repetitive and menial tasks to increase output, accuracy, and safety and more flexible automation that can successfully adapt to new variables. Often this flexible automation is paired with remote monitoring or programming options so operators can view and make changes while off-site. Robots used for manufacturing and logistics include AMRs, AGVs, articulated robots, and Cobots.

Estimates show there are roughly 2.7 million industrial robots in use around the world, with around 400,000 new robots coming online every year. The global industrial robotics industry is valued at $43.8 billion, by revenue. The North American robotics industry revenue is expected to have a compound annual growth rate of 11.67 percent through 2026.

A number of robots for the smart home have been released or are under development. Perhaps the most well-known example is the Roomba autonomous vacuum cleaner. iRobot, the company behind the Roomba, uses AI and machine learning for their newer models, upgrading the navigation and control software that maps users' homes to identify "cleaning zones."

Robots have a large number of use cases in the healthcare industry:

• Robots to assist with surgeries
• Bots to help recovery and physical therapy
• Autonomous blood drawing devices
• Service robots to clean and transport supplies

Robotics can help make healthcare processes more efficient and improve the level of patient care.

Robots have been used for a number of space missions as they can perform experiments without space agencies having to worry about creating habitable, safe conditions for astronauts. Examples of active robot space missions include Mars rovers such as Perseverance and Curiosity from NASA and China's Tutu 2 lunar rover.