Robotics

Robotics
Interdisciplinary engineering and scientific field concerned with robots
Discipline	Engineering and applied science
Principal fields	Mechanical engineering, electrical engineering, control engineering and Computer Science
Core topics	Kinematics, dynamics, motion planning, feedback control and robot perception
Types of systems	Industrial manipulators, mobile robots, service robots and medical robots
Applications	Manufacturing, logistics, healthcare, agriculture and scientific research
Vocabulary standard	ISO 8373:2021
Related fields	Artificial Intelligence, Physical AI, Industrial Automation
Security topics	Robot Cybersecurity, Machine Identity

Robotics is an interdisciplinary field of engineering and science concerned with the design, construction, control, operation and application of robots. It combines mechanical systems, electronics, sensors, computing and control methods to develop machines capable of performing physical tasks. Robotics includes both the machines themselves and the methods used to describe their motion, interpret sensor information, plan actions and control interactions with the environment.

The International Organization for Standardization's ISO 8373 vocabulary defines robots in terms of programmed, actuated mechanisms that possess some degree of autonomy and perform locomotion, manipulation or positioning. Under this definition, autonomy does not require human-level intelligence or continuous independent decision-making. Robots may execute predefined programs, respond to sensor feedback or use more advanced planning methods.

Robotics draws on mechanical and electrical engineering, computer science, mathematics, control theory and, in some applications, artificial intelligence. Its applications include manufacturing, logistics, agriculture, healthcare, scientific research and work in hazardous or otherwise difficult-to-access environments.

History

The term robot entered public discourse through Czech writer Karel Čapek's play R.U.R. (Rossum's Universal Robots), published in 1920 and first performed in 1921. Its artificial workers were fictional beings rather than examples of modern electromechanical robotics. The related term robotics became established through Isaac Asimov's science-fiction writing in the 1940s. These literary uses influenced the vocabulary of the field but should be distinguished from its engineering history.

Modern industrial robotics developed through advances in programmable machinery, feedback control, electrical drives and computer technology. An important milestone was the installation of the Unimate robotic arm at a General Motors plant in 1961. Developed from the work of George Devol and Joseph Engelberger, Unimate performed industrial material-handling operations and became an early example of commercially deployed programmable robotic machinery.

Subsequent developments expanded robotics beyond fixed industrial manipulators. Research in mobile navigation, machine vision, manipulation, sensing and computing supported robots capable of operating in less structured environments. Modern robotics encompasses industrial machines, mobile platforms, medical and assistive systems, as well as experimental robots developed for research and exploration.

Definitions and scope

Robots and robotic devices

A robot is not necessarily humanoid. Robots can take the form of articulated arms, wheeled platforms, legged machines or other actuated mechanisms suited to particular tasks.

ISO 8373:2021 provides standardized terminology for robotics and distinguishes robots from other robotic devices. The distinction is significant because not every automated mechanism satisfies the standard's definition of a robot. For example, the degree of autonomy and the mechanism's intended function matter when classifying a system.

Robotics is also not synonymous with artificial intelligence. A robot can perform useful functions through conventional programming and feedback control without using machine learning. Conversely, an artificial intelligence model can operate entirely in software without controlling a physical machine.

Industrial, service and medical robotics

The International Federation of Robotics (IFR) uses ISO terminology when describing industrial and service robots.

An industrial robot is a reprogrammable, multipurpose manipulator that can be controlled automatically and programmed in at least three axes for industrial automation applications. Industrial robots may be fixed in place or mounted on mobile platforms.

A service robot carries out useful tasks for people or equipment in personal or professional applications. These include activities such as cleaning, logistics and assistance. ISO 8373:2021 also recognizes medical robots as a separate classification.

These categories describe application and function; they are not determined solely by the external appearance of a robot.

Scientific and engineering foundations

Kinematics

Robot kinematics studies the geometry of motion without directly considering the forces that produce it.

In a robotic manipulator, forward kinematics determines the position and orientation of the end effector from the robot's joint configuration. Inverse kinematics addresses the reverse problem: determining joint configurations that achieve a specified end-effector pose.

Kinematic models use coordinate frames, rotations and rigid-body transformations. They are fundamental to robot manipulation, motion planning and control.

Dynamics

Robot dynamics describes the relationship between motion and the forces or torques acting on a robot.

Dynamic models account for quantities such as mass, inertia, gravity, joint forces and actuator inputs. They are important when calculating motor torques, analyzing mechanical loads, designing controllers and evaluating dynamic performance.

Motion planning

Motion planning concerns the selection of a feasible route or sequence of configurations between a robot's initial state and a desired goal.

Depending on the task, a planner may need to consider obstacles, joint limits, kinematic constraints, collision avoidance and the geometry of the robot's environment.

Trajectory generation additionally addresses how a robot moves along a path over time, subject to limits on quantities such as velocity and acceleration.

Feedback control

Feedback control uses measurements of the current system state to adjust actuator commands and reduce the difference between actual and desired behavior.

Robot controllers may regulate joint position, velocity, force or torque. Control methods range from established feedback controllers to model-based and optimization-based techniques. Their selection depends on the robot, task and operating conditions.

Robot hardware and components

Mechanical structure

A robot's mechanical structure determines its possible movements, operating workspace and ability to transmit forces.

Common mechanical arrangements include serial-link manipulators, Cartesian mechanisms, parallel mechanisms, wheeled mobile bases and legged platforms.

The number and arrangement of joints influence the robot's degrees of freedom. Payload, reach, stiffness, speed and positioning accuracy are relevant engineering characteristics when selecting or designing a robotic system.

Actuators

Actuators convert supplied energy into mechanical motion or force. Electric motors, hydraulic systems and pneumatic systems are used in different robotic applications.

The choice of actuator affects performance characteristics such as controllable force, speed, efficiency, mechanical complexity and maintenance requirements.

End effectors

An end effector is the device attached to a robotic manipulator that performs an operation or interacts with an object.

End effectors include mechanical grippers, vacuum grippers and process-specific tools. Their design depends on the object or operation involved, as well as the forces and precision required.

Sensors

Robots use sensors to measure internal conditions and obtain information about their surroundings.

Internal measurements commonly include joint positions, motor motion and inertial information. Environmental sensing may use cameras, ranging systems, force and torque measurements, or tactile sensors.

Measurements are subject to noise, calibration errors and limitations of the sensing method. Reliable robotic operation therefore depends on appropriate sensor selection, calibration and interpretation.

Perception, localization and navigation

Robot perception

Robot perception involves obtaining and interpreting information from sensors to estimate features of a robot's surroundings or its own state.

Computer vision methods are used to analyze images, identify objects and interpret visual observations. Other sensing methods may provide depth, range, force or movement information.

Sensor fusion combines measurements from multiple sources. Because each sensor has its own limitations, combining information can improve state estimation when the measurement models and uncertainty are handled appropriately.

Localization and mapping

Localization is the process of estimating a robot's position and orientation with respect to a reference frame or map.

Mapping is the process of constructing a representation of the surrounding environment from available observations.

In simultaneous localization and mapping (SLAM), a robot estimates its own pose while constructing or updating an environmental map. The two estimation problems are coupled because the robot's interpretation of sensor data depends on its estimated position, while its position estimate can depend on features of the map.

Autonomous navigation

Mobile robots use localization, environmental sensing, route planning and motion control to travel toward a destination.

Navigation systems must account for constraints such as obstacles, available free space, mechanical limitations and changes in the environment.

Different systems use different combinations of maps, localization techniques and path-planning algorithms. Not all mobile robots navigate autonomously; some follow externally prescribed routes or operate under remote control.

Major types of robots

Robots can be classified by mechanical configuration, operating environment or intended application. These classifications are not mutually exclusive; a mobile platform may also carry a robotic manipulator.

Robot type	Main characteristics	Representative applications
Articulated robot	Manipulator with multiple rotary joints	Welding, assembly, material handling and machine tending
Cartesian robot	Motion organized along linear coordinate axes	Positioning, transfer operations and automated handling
SCARA robot	Manipulator configuration suited to selected high-speed planar operations	Assembly and pick-and-place tasks
Parallel robot	Moving platform supported by multiple kinematic chains	High-speed handling and specialized positioning
Mobile robot	Platform capable of moving within an operating environment	Transport, inspection and navigation
Legged robot	Locomotion through articulated legs	Research, inspection and mobility over selected terrain
Humanoid robot	Mechanical arrangement incorporating human-like body features	Research, manipulation, mobility and service tasks
Medical robot	Robotic equipment developed for medical applications	Surgical assistance and other regulated medical uses

Robot configuration and application categories are discussed in the Springer Handbook of Robotics and in the industrial and service robot classifications maintained by the IFR.

Control, programming and autonomy

Robot behavior is determined by the interaction of mechanical design, control software, sensing, task requirements and operating conditions.

A robot operating in a structured production environment may follow a programmed sequence, using sensors and feedback to maintain specified performance. Other robots may need to select routes, react to new obstacles or adjust actions in environments that cannot be fully predicted in advance.

Autonomy describes an ability to perform intended tasks without continuous human intervention. The degree of autonomy varies across robotic systems and should not be confused with an unrestricted ability to make decisions.

Teleoperation, supervised operation, shared control and autonomous execution are different approaches to human involvement in robotic tasks. A system's operating mode must be considered when assessing its capabilities and safety requirements.

Artificial intelligence and robot learning

Artificial intelligence supports some robotic applications, particularly those involving complex perception, decision-making or adaptation.

Machine learning has been applied to tasks including visual recognition, manipulation and learned control. Approaches within robot learning include reinforcement learning, learning from demonstration and methods that use simulation to develop or evaluate policies.

However, machine learning is not a necessary component of every robot. Conventional motion planning and feedback control remain important even in systems that incorporate learned perception or decision-making.

Research on robot foundation models investigates models intended to support multiple robotic tasks. Their usefulness depends on factors including training data, embodiment, task conditions and evaluation in the relevant operating environment.

The behavior of a learned model alone does not establish that an entire robotic system is safe. Physical deployment still requires appropriate system-level testing and risk assessment.

Applications

Industrial manufacturing

Robots are used in industrial environments for operations such as welding, assembly, painting, material handling, inspection and machine tending. Their suitability depends on task requirements, production integration, reliability and safety controls.

Warehousing and logistics

Robotic systems are used for selected storage, transport, picking and handling operations. Mobile platforms and robotic manipulators may operate separately or as components of an integrated logistics system.

Healthcare and assistance

Robotics applications in healthcare include medical robotic equipment, rehabilitation technologies and other assistance-related systems. Requirements vary by application, especially where the robot interacts directly with people.

Agriculture and inspection

Robotic methods are applied to selected agricultural, environmental measurement and inspection tasks. Their design depends on the sensing, manipulation and mobility requirements of the operating environment.

Research and exploration

Robots are used as experimental platforms for studying locomotion, manipulation, perception, autonomy and human-robot interaction. Robotic systems also support scientific exploration in environments where direct human operation can be difficult or hazardous.

Industry and adoption

The International Federation of Robotics publishes industry statistics covering industrial and service robots.

In its World Robotics 2026 reporting, released in September 2026, the IFR reported that the worldwide operational stock of industrial robots reached approximately five million units in 2025, an increase of about 9% from the preceding year. These figures concern the industrial-robot category and should not be interpreted as a count of every type of robot in use worldwide.

Robot safety

Robots can create hazards through unintended motion, collisions, loss of control, electrical or mechanical faults, and interactions with tools or other equipment. Safety therefore concerns the complete robotic application, including its working environment and foreseeable interactions with people.

Risk reduction can involve appropriate mechanical design, protective measures, safety-related control functions, restricted access, monitored operating conditions and procedures for installation and maintenance.

For industrial robotics, the principal international safety standards include:

• ISO 10218-1:2025 — Requirements addressing industrial robots.
• ISO 10218-2:2025 — Requirements addressing industrial robot applications and robot cells.

Part 1 concerns the robot itself, while Part 2 addresses integration into an industrial application or cell. Their scope should not be assumed to cover every category of service, medical or consumer robot.

ISO 13482:2014 addresses safety requirements for specified categories of personal care robots. Its scope differs from that of industrial robot safety standards.

A robot operating near people must be assessed according to its intended use, hazards and applicable requirements. The presence of a collaborative operating mode or a particular robot design does not, by itself, establish that an application is safe.

Robot identity and cybersecurity

Networked robots may exchange instructions, operational data, software updates and status information with controllers, operators and external services.

These capabilities introduce cybersecurity concerns alongside traditional mechanical and electrical safety considerations. Threats may involve unauthorized commands, compromised software, insecure communication or disruption of services used by the robot.

Relevant protective measures can include:

• Machine identity — Establishing identities for robots, controllers and related components.
• Authentication — Verifying entities before accepting commands or establishing trusted communications.
• Authorization — Restricting permitted operations and access to resources.
• Software integrity — Protecting updates and executable software against unauthorized modification.
• Monitoring — Recording and analyzing relevant security events.
• Network protection — Applying appropriate network architecture and access restrictions.

NIST has studied cybersecurity controls and their operational effects in robotic manufacturing workcells. Its guidance on operational technology security also emphasizes the need to consider reliability, performance and safety when designing security measures.

Cybersecurity measures must be evaluated as part of the overall robotic system. Strong authentication, for example, does not by itself guarantee collision avoidance, correct motion or functional safety.

Research challenges

Robotics research continues to address several difficult engineering problems.

These include reliable operation under uncertain sensing conditions, manipulation of unfamiliar objects, motion through complex environments, dexterous control, energy use, safe interaction with people and the transfer of methods from simulation into physical operation.

For learning-enabled systems, generalization beyond training conditions and the evaluation of failures remain important technical considerations. For connected fleets, software maintenance, equipment identity and coordination across multiple systems add further engineering requirements.

Progress in an individual algorithm or component does not necessarily establish equivalent performance in a fully integrated robotic system.

Selected academic books

The following publications provide technical coverage of robotics and related research areas:

• Siciliano, Bruno, and Oussama Khatib, editors (2016). Springer Handbook of Robotics. 2nd edition. Springer. ISBN 978-3-319-32552-1.
• Lynch, Kevin M., and Frank C. Park (2017). Modern Robotics: Mechanics, Planning, and Control. Cambridge University Press. ISBN 978-1-107-15630-2.
• Thrun, Sebastian, Wolfram Burgard, and Dieter Fox (2005). Probabilistic Robotics. MIT Press. ISBN 978-0-262-20162-9.
• LaValle, Steven M. (2006). Planning Algorithms. Cambridge University Press. ISBN 978-0-521-86205-9.

See also

• Physical AI
• Embodied AI
• Autonomous Systems
• Industrial Automation
• Computer Vision
• Sensor Fusion
• Machine learning
• Robot Foundation Models
• Robot Cybersecurity
• Machine Identity
• Human-Robot Interaction
• Robot Safety Risk Assessment