An actuator is a device that converts supplied energy into mechanical force, torque, or motion in response to a command or physical stimulus. Actuators open valves, position machine tools, move robot joints, adjust aircraft control surfaces, and perform countless other operations in which a system must act on its surroundings. Their energy may come from electricity, pressurized liquid or gas, heat, or stored mechanical energy. Some actuators produce continuous rotation; others provide limited angular movement, linear displacement, vibration, or deformation.
In an automated system, an actuator commonly forms the physical output stage: a controller determines the required action, an energy-conversion device produces force, and a mechanical connection delivers that force to the load. A sensor performs the complementary task of measuring a physical condition. The boundary of an actuator depends on context: the term may identify a bare cylinder or electromagnetic device, or an integrated assembly containing a motor, transmission, sensors, electronics, and protective equipment.
Actuators are central to automation because computation alone cannot change a valve opening, grip an object, or stabilize a moving mechanism. Their behavior determines not only how much work a machine can perform, but also its precision, responsiveness, energy consumption, and response to faults. Understanding an actuator therefore requires consideration of its energy source, mechanical structure, control system, and interaction with its load.
Historical development and scope
Actuation predates electronic control by centuries. Water-powered mechanisms, windmill controls, clockwork, and spring-driven devices converted available energy into useful movement, although their makers did not necessarily classify the components as actuators. An important distinction developed between the mechanism that decided when movement should occur and the mechanism that supplied the force. A float might detect a liquid level while a lever operated a valve; a rotating cam could prescribe a motion while a powered shaft supplied the required work.
The development of the steam engine made controlled mechanical power increasingly important. In 1788, James Watt incorporated a centrifugal governor into a rotative steam-engine installation. Changes in shaft speed altered the position of rotating weights, which acted through a linkage on a steam throttle. The governor combined sensing, mechanical computation, and valve actuation without electronic components. Later steam and hydraulic servomechanisms separated the small effort needed to issue a command from the much larger force needed to move machinery.
Electromagnetic technology introduced a different route to controlled motion during the nineteenth century. Electromagnets operated telegraph mechanisms, relays, bells, and switching equipment, while the development of the electric motor made continuous electrically powered motion practical. Hydraulic and pneumatic machinery developed alongside electric drives rather than simply being displaced by them. Different energy carriers offered different advantages: electricity was readily distributed and controlled, while fluid pressure could produce large forces in compact cylinders.
During the twentieth century, aircraft, machine tools, industrial process plants, and guided systems increased demand for accurate powered positioning. Servomechanisms used feedback to make output position or speed follow a command despite disturbances. Numerical control and subsequently computer control extended this principle to coordinated machine axes. The first Unimate industrial robot entered factory service at a General Motors plant in 1961; its hydraulic actuation reflected the importance of fluid power in early programmable manipulation.
Modern usage consequently covers devices with very different levels of complexity. A thermal element that opens a radiator valve, a piezoelectric positioning stack, and an integrated robotic joint can all be actuators. Conversely, not every source of mechanical power is normally called one. A propulsion engine is usually described as an engine, while a device that adjusts its throttle or guide vanes is described as an actuator. The distinction is functional rather than a strict division between physical technologies.
Energy conversion and performance characteristics
An actuator does not create energy. It converts an input into mechanical output while dissipating some energy through electrical resistance, friction, fluid leakage, throttling, or other losses. For linear movement, instantaneous mechanical power is
where is force and is velocity in the force direction. For rotation,
where is torque and is angular velocity. These relations explain why maximum force or torque alone is an inadequate description: a device may generate a large static force while having little capacity for rapid motion.
Output capability is usually described by several interacting limits. Stroke is the available linear travel, while angular range specifies permitted rotation. Speed, acceleration, peak force, continuous force, and permissible load direction further constrain operation. Continuous ratings are commonly set by thermal equilibrium, whereas short-duration peak ratings may depend on current limits, material stress, pressure limits, or allowable temperature rise. A motor that tolerates a large starting torque for a fraction of a second may overheat if required to sustain it.
Precision also has several distinct meanings. Resolution is the smallest commanded or measured increment. Accuracy describes agreement between actual and intended output, while repeatability describes the consistency of repeated movements under specified conditions. A mechanism can repeat a position closely while remaining systematically offset from the target. These properties depend on measurement location and loading: excellent motor-shaft positioning does not guarantee equally accurate motion at the end of a flexible linkage.
Stiffness expresses resistance to displacement under load. For a linear elastic approximation, . High stiffness can improve resistance to external disturbances, but may also transmit impacts directly into a machine. Compliance can be intentional, particularly where contact forces must be moderated. Friction introduces other effects: static friction can prevent small commanded movements, while changes between sticking and sliding can produce irregular motion. Backlash creates lost movement when a transmission reverses direction, and hysteresis makes output depend on the history of the input.
Dynamic performance concerns how quickly and faithfully the output follows changing commands. Rise time, settling time, overshoot, and frequency response describe different aspects of this behavior. The useful bandwidth of an assembled mechanism is limited not only by its actuator but also by load inertia, structural resonances, sensing, and control delays. Ratings therefore require operating conditions. Force-to-mass ratios, efficiency figures, and positioning specifications are meaningful only when their measurement boundaries, duty cycles, cooling arrangements, and load assumptions are stated.
Electrical and electromagnetic actuators
Electrical actuators convert electrical energy into motion through electromagnetic forces or, in some cases, electrically induced material deformation. Motor-based systems are particularly widespread because electrical power and command signals are readily distributed. A DC motor produces torque through interaction between current and a magnetic field; electronic drives regulate current, speed, and position. A brushless DC motor performs commutation electronically rather than with sliding brush contacts. Induction and synchronous motors are also used in actuator assemblies, depending on required speed range, precision, cost, and operating environment.
For a simplified permanent-magnet motor, torque is approximately proportional to current:
Rotation also generates a back electromotive force approximately proportional to angular speed:
These relationships connect mechanical performance to the electrical supply and drive electronics. Increasing speed raises the back electromotive force, leaving less voltage available to drive current through the winding. Increasing torque requires more current and usually increases resistive heating. The resulting torque–speed envelope is therefore constrained by voltage, current, temperature, and mechanical limits rather than by a single maximum rating.
A stepper motor advances through a sequence of electromagnetic equilibrium positions as its phases are energized. It can provide convenient incremental positioning without continuous position measurement, but open-loop operation does not prove that the load followed every commanded step. Excessive acceleration, resonance, or an unexpected load can cause loss of synchronism. Microstepping subdivides commanded electrical increments and can improve smoothness, but nominal subdivision is not equivalent to guaranteed absolute positioning accuracy.
Solenoids generally provide a limited stroke by pulling a ferromagnetic armature toward a lower-reluctance magnetic configuration. Their force varies strongly with air gap and magnetic saturation. They are common in latches, contactors, and valve mechanisms. A voice-coil actuator instead places a current-carrying coil in a magnetic field and can produce approximately current-proportional force over a designed travel range. Loudspeaker drives, optical positioning devices, and disk-drive head mechanisms illustrate this operating principle. Linear motors extend electromagnetic motor action along a straight path, avoiding a rotary-to-linear transmission.
The complete electric actuator frequently includes power electronics, position sensing, a brake, and a transmission. Drives may return energy to a suitable supply or storage system during deceleration, a principle related to regenerative braking. Holding a stationary load can still consume substantial power if continuous current is required. A mechanical brake or self-locking transmission can reduce that demand, but changes the system’s response to faults and external forces. Consequently, electric actuation ranges from inexpensive on–off devices to sophisticated servo axes with tightly managed electrical, thermal, and mechanical behavior.
Hydraulic and pneumatic actuators
Fluid-power actuators convert the energy of pressurized liquid or gas into mechanical output. In a hydraulic cylinder, pressure acts on a piston connected to a rod. For a simplified double-acting cylinder, the available force is
where and are chamber pressures, and are their effective areas, and represents friction and related resistance. Because the rod occupies part of one face, the effective areas commonly differ. Extension and retraction can therefore have different force and speed capabilities at the same supply conditions.
Flow determines speed as well as pressure determining force. Neglecting leakage and compression, piston velocity is , where is chamber inflow. As an illustrative calculation, an effective area of exposed to a pressure difference of produces an ideal force of . Real output is lower when friction, opposing pressure, and other losses are included. The example shows why hydraulic cylinders can deliver substantial force from a relatively small cross-section.
Hydraulic systems use liquids whose low compressibility supports high stiffness and powerful motion. They are common in excavators, presses, lifting equipment, and aircraft systems. Their complete architecture may include a pump, reservoir, filters, valves, accumulators, piping, and cooling equipment. Directional or servo valves regulate flow, while pressure-control devices limit loads. Throttling across valves dissipates energy as heat; pump-controlled arrangements can reduce some of these losses. Entrained air, flexible hoses, fluid compressibility, seal friction, and structural compliance all affect practical dynamics.
Pneumatic actuators use compressed gas, usually air. Cylinders, rotary vane units, and rack-and-pinion mechanisms are widely used for clamping, sorting, packaging, and valve operation. Their construction can be simple and robust, and exhausted air often returns directly to the surroundings. However, air compressibility makes position and force depend strongly on chamber volume, pressure, temperature, and valve flow. Precision pneumatic positioning is possible with appropriate feedback and modeling, but its dynamics differ substantially from those of an incompressible-fluid approximation.
Both technologies can incorporate springs, cushioning, end-position sensing, and load-holding valves. A single-acting actuator may use pressure for motion in one direction and a spring or external load for return. A double-acting actuator uses pressure to drive both directions. Stored pressure can remain hazardous after the main supply is removed, and a suspended load may move if a hose fails or a valve leaks. Thus, the actuator’s apparent simplicity must be distinguished from the complexity of its power supply, control components, and load-restraint arrangements.
Piezoelectric, thermal, and deformable-material actuators
Some actuators produce movement through changes in material dimensions rather than through a conventional motor or piston. A piezoelectric actuator uses the converse piezoelectric effect: an applied electric field produces strain in an appropriate material. In a simplified one-dimensional description, free strain is , where is a relevant piezoelectric coefficient and is electric field. The complete behavior is tensorial and depends on material orientation, mechanical stress, and electrical boundary conditions.
Multilayer piezoelectric stacks combine many thin active layers to obtain useful displacement at practical drive voltages. They can provide high stiffness, rapid response, and very fine positioning, but direct strain is small. Mechanical amplification increases travel at the cost of output force and stiffness. Ceramic elements are vulnerable to tensile and bending loads, so assemblies often apply compressive preload. Hysteresis, creep, temperature changes, and mounting stresses affect accuracy. Closed-loop sensing or charge-control techniques can reduce some errors, but do not remove structural and environmental limitations.
Thermal actuators exploit expansion or temperature-driven phase changes. A bimetallic strip bends because bonded metals expand by different amounts. Wax elements produce useful displacement as their working material expands through a temperature range, as in many engine thermostats. Electrically heated beams can actuate microscale mechanisms. Thermal devices may be compact and mechanically simple, but their response is constrained by heat transfer. Heating and cooling rates are generally unequal, and sustained operation may require continuous energy input even when little external mechanical work is being done.
A shape-memory alloy can recover a programmed shape through a reversible solid-state transformation. Nickel–titanium alloys are prominent examples. A loaded wire may contract on heating and be extended again by a bias spring or opposing load during cooling. Such devices can provide substantial force for their mass, but transformation hysteresis, cooling time, fatigue, and dependence on stress complicate precise control. Magnetostrictive actuators provide another material-based route: selected ferromagnetic materials change dimensions in response to magnetic fields.
Deformable actuators also include dielectric elastomer membranes, ionic polymer devices, and fluid-driven soft structures. In a dielectric elastomer actuator, electrostatic attraction between compliant electrodes compresses the thickness of an insulating elastomer and expands its area. Dielectric breakdown, electrode durability, and mechanical instability constrain operation. Pneumatic soft chambers instead bend or elongate through designed differences in wall geometry and reinforcement. Polymers and hydrogels support additional stimulus-responsive mechanisms, while microelectromechanical systems frequently employ electrostatic or thermal actuation. These technologies extend actuation to small scales and compliant forms, but their force, displacement, lifetime, and environmental sensitivity differ markedly.
Transmissions and mechanical integration
The energy-conversion element is often only one part of an actuator. Its natural output may not match the load’s required speed, force, travel, or geometry. A transmission transforms that output and connects it to the machine. A gear reduction, for example, decreases output speed and increases available torque, subject to losses. If the reduction ratio is , then ideally ; with efficiency , the corresponding approximation is .
Reduction also changes how the load is experienced at the motor. Under an ideal rigid-transmission model, a load inertia is reflected to the motor as . This can improve the match between a small motor and a heavy load. However, gears introduce their own inertia, friction, backlash, compliance, noise, and wear. High-ratio transmissions may resist backdriving, but this behavior depends on geometry and friction rather than on ratio alone. A transmission that holds a load under one condition may not constitute a reliable brake under vibration, wear, or altered lubrication.
Rotary motion is commonly converted into linear motion by screws. A ball screw uses recirculating balls to transmit load between a screw and nut with relatively low friction. Sliding lead screws are simpler and may be self-locking for some combinations of lead angle and friction, but usually dissipate more energy. For a screw with lead , an approximate driving relationship is
A smaller lead increases mechanical advantage but requires more revolutions for a given travel. Long screws also face critical-speed, buckling, alignment, and lubrication constraints.
Belts, cables, chains, racks, cams, and linkages offer other transformations. Cable drives can place a heavy motor away from a moving joint, reducing distal mass, but cable stretch and tension management influence accuracy. Flexure mechanisms guide motion through elastic deformation rather than sliding or rolling contacts, avoiding backlash within a limited range. Direct-drive arrangements eliminate a transmission stage and its associated play, but require the motor to supply load torque or force without mechanical multiplication.
Integration also determines which forces pass through which components. A cylinder rod or screw is not automatically an adequate guide for large side loads; separate bearings or rails may be needed to support them. Mounting stiffness, alignment, seals, lubrication, corrosion protection, and cable routing can dominate practical performance. Position sensors mounted before a transmission cannot directly observe every output error caused by its compliance or lost motion. Mechanical design and sensing location are therefore inseparable from the actuator’s nominal specifications.
Feedback control and dynamic behavior
Actuators connect the abstract commands of control theory to physical systems. In open-loop operation, the command is issued without using measured output to correct the result. This is adequate for many end-to-end movements, timed operations, or loads whose behavior is sufficiently predictable. Closed-loop operation uses negative feedback to compare a measured quantity with a reference and adjust the input. The controlled quantity may be position, velocity, force, torque, pressure, or a combination of these.
A simplified motor-driven axis can be represented by
where is total inertia, represents viscous damping, is load torque, and is motor current. This model explains why the same motor behaves differently when attached to different loads. Real mechanisms additionally exhibit friction, elasticity, backlash, electrical dynamics, and nonlinear limits. A transfer function can describe a linearized operating condition, but it does not automatically capture impacts, contact changes, saturation, or large variations across a working range.
Servo drives often use nested control loops. An inner current loop regulates electromagnetic torque, a velocity loop regulates speed, and an outer position loop regulates displacement. Their response rates are coordinated so that each outer loop acts on an adequately responsive inner system. Feedforward terms can supply predicted effort for acceleration, friction, or gravity, reducing the correction left to feedback. Encoder resolution, measurement noise, sampling intervals, and computation delays affect the achievable result.
Physical limits create important control problems. If a controller requests more voltage, flow, current, or force than the actuator can deliver, its output saturates. Integral control can then accumulate an excessive internal command, producing overshoot or delayed recovery after the limitation ends; anti-windup methods address this behavior. Structural resonances and time delays can reduce stability margins. Raising controller gain may improve disturbance rejection over part of the frequency range while amplifying noise or exciting flexible modes elsewhere.
Contact with the environment changes the objective. A robot approaching a rigid surface may need controlled interaction rather than strict position tracking. Force control regulates contact effort, while impedance control specifies a relationship among force, position, and motion. A series elastic actuator places a deliberate elastic element in the force path; measured deflection can estimate force through or its rotational equivalent. Compliance can improve impact tolerance and force sensitivity, but adds dynamics and limits stiffness. Communication networks and programmable logic controllers coordinate many actuators, while fast inner loops often remain within local drives.
Applications across industry and everyday technology
Industrial production uses actuators both to move products and to regulate the processes that transform them. An industrial robot coordinates several powered joints to position a tool or workpiece. Machining centers require controlled axis motion, while presses prioritize high force and structural rigidity. Packaging lines combine rapid indexing, gripping, cutting, and sealing, often using electric servos for coordinated motion and pneumatic cylinders for simple repetitive actions. The actuator choice follows the operation’s load, cycle time, cleanliness, and precision requirements rather than a universal preference for one technology.
In process plants, actuators adjust valves, dampers, and other final control elements. A pneumatic diaphragm actuator can move a valve stem against a spring, while an electric motor actuator may operate a large isolation valve through a gearbox. The force needed depends on pressure differences, seal friction, valve geometry, and flow-induced loads. Valve motion and process response are not identical: a small stem displacement can produce a nonlinear change in flow, and wear can alter the relationship. Local positioners help make valve travel follow the command.
Transportation contains numerous specialized forms of actuation. An automobile uses actuators for throttle positioning, braking functions, steering assistance, transmission shifting, door locking, and climate control. Aircraft employ hydraulic, electrohydraulic, and electromechanical arrangements for flight controls and utility systems, with architectures selected for particular loads and reliability requirements. In railway systems, actuators operate doors, brakes, switches, and suspension-related equipment. These applications expose devices to vibration, temperature variation, contamination, and long service intervals.
Buildings and infrastructure use actuators on a different timescale. Heating and air-conditioning systems adjust dampers and water valves, sometimes over tens of seconds rather than milliseconds. Waterworks and irrigation systems operate gates and valves over large distances. Wind turbines use pitch actuators to rotate blades and yaw drives to orient the nacelle; both influence energy capture and loading. Solar-tracking mechanisms trade additional energy collection against drive consumption, mechanical complexity, and maintenance.
Small-scale applications emphasize low mass, compactness, or precision. Camera lenses use miniature motors or voice-coil devices for focus and stabilization. Precision optical instruments and electron microscopes use specialized positioning mechanisms where vibration, drift, and vacuum compatibility matter. Surgical instruments and powered prosthetic devices impose additional constraints on size, interaction forces, cleanability, and reliability. Across robotics, especially mobile and legged systems, actuator design is closely coupled to available battery energy: peak output, holding losses, transmission efficiency, and regenerative capability all affect operating duration.
Reliability, safety, and engineering trade-offs
Actuator failures can involve loss of motion, unintended motion, reduced force, loss of position accuracy, or an inability to stop. Electrical windings may overheat, bearings may wear, seals may leak, and transmissions may develop excessive play. A sensor fault can be as consequential as a mechanical failure if it causes a controller to apply inappropriate effort. Fault analysis therefore considers the complete chain from command generation and power supply to sensing, energy conversion, transmission, and load support.
Removing electrical power does not necessarily remove mechanical danger. Springs, hydraulic accumulators, compressed air, gravity loads, and rotating masses can retain energy. A brake may engage when power is lost, but its capacity, engagement time, and wear state remain relevant. In electric drives, a safe-torque-off function prevents torque-generating power from being supplied through the drive under specified conditions; it does not by itself restrain a descending load or guarantee immediate cessation of motion. Emergency stopping and secure load holding are distinct engineering functions.
A fail-safe state is defined by the application. A valve may be required to close, open, or remain in place after loss of its control signal or power source. Spring-return actuators provide one method of reaching a designated position, while stored fluid energy or a separate electrical supply can provide others. No single default position is safe for every process. In some installations, closing a valve limits a release; in others, continued flow is needed for cooling. The actuator’s failure behavior must therefore be considered within the larger process and its hazards.
Applicable standards address different parts of this problem. ISO 4413 covers general rules and safety requirements for hydraulic fluid-power systems, and ISO 4414 addresses pneumatic systems. ISO 13849-1 concerns safety-related parts of machinery control systems, while IEC 61508 provides a framework for functional safety of electrical, electronic, and programmable electronic safety-related systems. These frameworks do not make an actuator safe in isolation: installation, diagnostic coverage, component reliability, environmental conditions, and the required safety function remain essential.
Selection is consequently a multi-objective engineering task. Higher reduction can reduce motor size but increase friction and lost motion; higher stiffness can improve positioning but worsen impact loads; smaller packaging can impede cooling. Efficiency must be evaluated over the operating cycle, including standby, holding, acceleration, and energy recovery, rather than only at a favorable steady-state point. Condition monitoring can track temperature, current, vibration, pressure, leakage, or position error to detect deterioration. Such measurements require interpretation against load and operating history, since a change in current or travel time may reflect a changed task as well as a developing fault.