{
  "id": "variable-frequency-drive",
  "title": "Variable-Frequency Drive",
  "lang": "en",
  "source_lang": "en",
  "summary": "A variable-frequency drive controls an AC motor’s speed and torque by electronically adjusting the frequency and voltage of its electrical supply.",
  "category": "technology",
  "subcategory": "engineering",
  "type": "concept",
  "aliases": [
    "VFD",
    "Variable-frequency inverter",
    "Adjustable-frequency drive",
    "Frequency converter",
    "AC drive"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/variable-frequency-drive/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/variable-frequency-drive/",
      "json": "https://aiwiki.page/zh/variable-frequency-drive.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A variable-frequency drive (VFD) is an electronic power-conversion system that controls the speed and torque of an alternating-current electric motor by varying the frequency and voltage supplied to it. Instead of connecting a motor directly to a fixed-frequency supply, the drive synthesizes an electrical output suited to the required operating condition. VFDs are used in equipment ranging from small ventilation fans to large compressors, rolling mills, and marine propulsion systems. Most industrial installations employ three-phase induction motors, although drives also operate permanent-magnet and other synchronous machines.\n\nVariable-frequency control makes motor speed an adjustable process variable rather than a quantity largely fixed by the electricity supply. This permits controlled acceleration, accurate regulation of flow or motion, reduced mechanical shock, and—in suitable applications—substantial reductions in electricity consumption. The drive is therefore both a power converter and a control system, combining semiconductor switches, measurement circuits, embedded computation, and protection functions.\n\nThe term is narrower than variable-speed drive, which also includes direct-current drives, hydraulic systems, and mechanical transmissions. It overlaps with AC drive and frequency converter, although a frequency converter need not be intended for motor control. A complete driven installation includes more than the electronic converter: the motor, cables, mechanical load, supply network, and supervisory controller all influence its performance."
    },
    {
      "heading": "Motor speed, frequency, and torque",
      "text": "The operating principle of a VFD follows from the relationship between electrical frequency and the rotating magnetic field of an AC machine. For a motor with \\(p\\) magnetic poles supplied at frequency \\(f\\), the synchronous speed is\n\n\\[n_s=\\frac{120f}{p},\\]\n\nwhere \\(n_s\\) is expressed in revolutions per minute and \\(f\\) in hertz. A four-pole machine therefore has a synchronous speed of 1,500 revolutions per minute at 50 hertz and 1,800 revolutions per minute at 60 hertz. Changing the supply frequency changes this reference speed without requiring a different pole arrangement or a mechanical transmission.\n\nAn induction motor normally runs below synchronous speed when delivering torque. The relative difference is called slip:\n\n\\[s=\\frac{n_s-n_r}{n_s},\\]\n\nwhere \\(n_r\\) is rotor speed. Slip allows the rotating stator field to induce rotor currents, whose interaction with the field produces torque. A drive operating a loaded induction motor at 25 hertz does not, by frequency adjustment alone, guarantee precisely half its full-frequency shaft speed: slip and the load-dependent control response must also be considered. Synchronous motors instead maintain rotor motion in synchronism with the fundamental rotating field during stable operation.\n\nFrequency cannot generally be changed independently of voltage. Neglecting stator resistance and other voltage drops, the magnetic flux established by a sinusoidal supply is approximately proportional to voltage divided by frequency. Applying rated voltage at a much lower frequency would therefore tend to overexcite the motor, producing excessive magnetizing current and heating. Basic drives reduce voltage approximately in proportion to frequency to maintain a suitable flux level.\n\nBelow the motor’s base speed, a properly matched motor and drive can often provide an approximately constant-torque operating region, subject to current and thermal limits. Above base speed, the drive usually cannot increase its output voltage further. Increasing frequency then reduces available flux, creating a field-weakening region in which available torque generally falls. An approximately constant-power region may be possible over part of this range because mechanical power is\n\n\\[P=T\\omega,\\]\n\nwith torque \\(T\\) and angular speed \\(\\omega\\). Neither constant torque below base speed nor constant power above it is unlimited: cooling, rotor strength, bearings, available current, and the machine’s electromagnetic characteristics constrain the usable envelope."
    },
    {
      "heading": "Power-conversion architecture",
      "text": "The most common industrial VFD is a voltage-source converter with three main stages: an input rectifier, an intermediate direct-current link, and an output inverter. The rectifier converts the incoming AC supply to DC. The DC link stores and smooths electrical energy, while the inverter switches that DC voltage to produce a controlled multiphase output. This arrangement largely decouples motor frequency from supply frequency, although the supply voltage and converter design still constrain the available output.\n\nA standard three-phase input stage uses six diodes arranged as a bridge. In a capacitor-input drive, conduction occurs mainly when the instantaneous rectified supply voltage exceeds the DC-link voltage, producing non-sinusoidal input current. The DC link commonly contains electrolytic or film capacitors, sometimes combined with an inductor. A precharge circuit limits the initial current needed to charge these capacitors; without it, connecting an uncharged link directly to the supply could produce a severe current surge.\n\nThe output stage normally comprises three semiconductor half-bridges, one for each motor phase. Industrial low-voltage drives have widely used the insulated-gate bipolar transistor (IGBT), accompanied by antiparallel diodes that provide current paths during switching and energy exchange with the motor. Metal–oxide–semiconductor field-effect transistors are important in other voltage and power ranges. Gate-driver circuits command the switches while enforcing timing constraints: simultaneous conduction of the upper and lower switches in one leg would short-circuit the DC link.\n\nNot all drives use this architecture. Current-source converters employ a substantial DC-link inductance and regulate current rather than maintaining a stiff DC voltage. Cycloconverters directly construct a low-frequency output from portions of the input waveform and have been used in very large, low-speed installations. Matrix converters use an array of bidirectional switches to connect input phases to output phases without a conventional large DC-link energy store. Each topology entails different compromises in voltage capability, commutation, regeneration, filtering, and protection.\n\nDrive ratings also require context. In IEC terminology, low-voltage equipment extends up to 1,000 volts AC, while industrial medium-voltage drives commonly serve motors rated in the kilovolt range. Large drives may use cascaded converter cells or other multilevel arrangements rather than a single two-level bridge. Their input transformers, cell bypass systems, insulation coordination, and maintenance provisions make them substantially different installations from compact cabinet-mounted drives."
    },
    {
      "heading": "Switching waveforms and modulation",
      "text": "A modern drive does not usually deliver a smooth sinusoidal voltage directly from its power switches. Instead, it applies a sequence of voltage pulses whose fundamental component has the desired frequency, phase, and amplitude. The motor’s inductance limits rapid current changes, so the resulting current can be much closer to sinusoidal than the terminal voltage. This distinction is essential: the output frequency that determines motor speed is not the same as the much higher switching frequency of the converter.\n\nIn pulse-width modulation (PWM), the duration and arrangement of switching pulses determine the synthesized output. A basic method compares sinusoidal reference signals with a high-frequency triangular carrier. Changing the reference frequency changes the electrical output frequency; changing its amplitude changes the fundamental voltage within the modulator’s operating limits. Space-vector modulation instead represents the inverter’s available switching states as vectors and allocates their application times to approximate a desired voltage vector.\n\nSwitching frequency involves competing effects. Raising it generally reduces current ripple and can move prominent acoustic components beyond the most sensitive range of human hearing. However, each switching transition dissipates energy in the power devices, so more transitions usually increase converter losses and cooling requirements. Drives therefore specify permissible switching frequencies in relation to output current, ambient temperature, and overload capability. A high carrier-frequency setting may require current derating rather than being a cost-free improvement.\n\nThe usable fundamental output voltage is limited by the DC-link voltage and modulation method. At higher voltage demands, a drive can enter overmodulation, eventually approaching six-step operation. This increases fundamental-voltage utilization but introduces additional low-order waveform distortion and changes the behavior of the current controller. Consequently, the highest-speed operating region may differ electrically from the region in which the inverter produces nearly sinusoidal motor currents.\n\nMedium-voltage converters often use multilevel inverter structures to build an output from several voltage steps. Cascaded H-bridge, neutral-point-clamped, and related arrangements can reduce the voltage increment imposed by an individual transition and improve waveform quality for a given device switching rate. These advantages come with additional devices, isolated supplies or transformer windings, capacitor-voltage management, and more complex fault handling. Waveform quality is therefore a system-level outcome, not a property determined solely by the number of switching pulses."
    },
    {
      "heading": "Scalar, vector, and direct torque control",
      "text": "The simplest widespread control method is scalar control, often called volts-per-hertz or \\(V/f\\) control. The drive commands frequency and schedules voltage to maintain approximately constant air-gap flux. At low frequency, stator resistance accounts for a significant share of the applied voltage, so a simple proportional relationship is insufficient; voltage boost or more detailed compensation may be added. Slip compensation can also increase commanded frequency as load rises, reducing the resulting speed drop.\n\nScalar control is suitable for many fans, pumps, and other loads that do not require rapid torque response or precise low-speed regulation. It can also operate several compatible motors from one drive, provided the installation includes appropriate individual motor protection. Its limitation is that flux and torque are not independently regulated with the precision available from more elaborate methods. Transient performance depends strongly on the machine and load.\n\nField-oriented control, also called vector control, transforms measured or estimated three-phase electrical quantities into rotating coordinates. Using mathematical techniques associated with control theory, it separates current components related to magnetic flux and torque. Inner current-control loops then regulate these components, while an outer loop can regulate speed or another process variable. For an induction motor, the controller must account for rotor flux and slip; for a permanent-magnet synchronous motor, it must establish the relevant relationship to rotor position.\n\nRotor position and speed can be obtained from an encoder or estimated from electrical measurements. Sensorless control eliminates a dedicated shaft sensor but does not eliminate measurement or dependence on a motor model. At very low speed, voltage drops, parameter errors, and weak speed-dependent electrical signals make estimation more difficult. Some methods use high-frequency signal injection and machine saliency to obtain information near standstill, but their applicability depends on the motor and required performance.\n\nDirect torque control regulates estimated torque and flux more directly, traditionally by selecting inverter switching states according to their errors. Later implementations include modulated and predictive variants. The labels used by manufacturers do not necessarily identify identical algorithms, and practical performance depends on sampling rates, motor identification, current sensing, switching strategy, and tuning. Accurate motor data—including rated current, voltage, frequency, speed, and electrical parameters—are important across advanced methods. A drive’s automatic identification routine may measure some parameters at standstill or during rotation, but it cannot remove the physical limits of the motor or load."
    },
    {
      "heading": "Load characteristics and energy use",
      "text": "Energy savings from variable-frequency operation depend principally on the mechanical process. In a centrifugal pump or fan operating under conditions where the affinity laws apply approximately, flow varies with rotational speed, developed head or pressure rise varies with speed squared, and shaft power varies with speed cubed:\n\n\\[Q\\propto n,\\qquad H\\propto n^2,\\qquad P_{\\mathrm{shaft}}\\propto n^3.\\]\n\nFor geometrically unchanged equipment under comparable operating conditions, reducing speed to 80 percent suggests a shaft-power requirement of approximately \\(0.8^3=0.512\\), or 51.2 percent of the original value. This is an idealized relationship, not a guaranteed reduction in electrical input.\n\nThe process operating point matters. A pump works at the intersection of its characteristic curve and the system curve. In a system dominated by frictional losses, reducing speed can substantially reduce both flow and required head. In a system with large static elevation head, the pump must still overcome that head, and the simple cubic power relationship does not describe the entire operating change. Pump efficiency, motor efficiency, and converter losses also vary with operating point.\n\nVariable-speed control can avoid losses associated with throttling valves, inlet guide vanes, or bypass arrangements. Applications include air-conditioning ventilation, cooling-water circulation, irrigation, and wastewater treatment. However, minimum flow requirements, solids transport, compressor stability, and process constraints can restrict permissible speeds. The lowest electrical input is not necessarily the operating point that maintains the required service or equipment condition.\n\nConstant-torque loads behave differently. A conveyor lifting a steady load, for example, requires roughly the same load torque at different speeds, while its instantaneous mechanical power changes approximately in proportion to speed. Slowing the conveyor may reduce power but also reduce throughput; energy per unit of product need not fall correspondingly. Positive-displacement pumps, extruders, and machine tools each have distinct torque–speed requirements that determine whether speed variation primarily improves efficiency, controllability, or production flexibility.\n\nThe drive itself introduces losses through semiconductor conduction, switching, magnetic components, control electronics, and cooling equipment. Modern converters can be highly efficient near their intended operating range, but a single efficiency figure does not describe every speed and load. Motor losses also change under inverter operation. Evaluation therefore distinguishes converter efficiency, motor efficiency, and total driven-system performance. Standards in the IEC 61800-9 series address energy-efficiency assessment for power drive systems, emphasizing operating points and system boundaries rather than assuming that installing a drive automatically saves energy."
    },
    {
      "heading": "Acceleration, braking, and coordinated motion",
      "text": "Controlled starting is a major benefit of a VFD. A directly connected induction motor can draw a starting current several times its rated current, whereas a drive can raise frequency and voltage progressively while limiting current. This reduces supply disturbance and permits acceleration to match the mechanical process. The required acceleration torque follows from the combined rotating inertia:\n\n\\[T_{\\mathrm{motor}}-T_{\\mathrm{load}} =J_{\\mathrm{eq}}\\frac{d\\omega}{dt},\\]\n\nwhere \\(J_{\\mathrm{eq}}\\) is the equivalent inertia referred to the motor shaft. A short acceleration time therefore requires greater torque, even when the steady running load is modest.\n\nRamps may be linear or shaped to reduce abrupt changes in acceleration. Smooth profiles limit mechanical shock in couplings, belts, and transported material. In pumping systems, controlled speed changes can reduce hydraulic transients, although pressure behavior also depends on pipe geometry, valves, and fluid dynamics. Current limits, torque limits, prohibited speed bands, and process interlocks often modify the commanded ramp.\n\nDuring deceleration, an overhauling load or the stored kinetic energy of the rotating system can drive the motor as an electric generator. Energy then flows into the DC link. A conventional diode input bridge cannot return that energy to the AC supply, so the link voltage rises unless the energy is consumed, stored, or dissipated. A braking chopper can connect a suitably rated resistor across the DC link to convert excess energy into heat. Both the resistor’s pulse-energy capability and its continuous thermal rating matter.\n\nA regenerative input converter can return energy to the supply, enabling regenerative braking where the operating cycle justifies it. A common DC bus can instead transfer energy from braking axes to motoring axes, as in coordinated production machinery. Such arrangements require deliberate protection and energy-balance design; merely connecting DC terminals does not establish a safe shared system. DC-injection braking, another available function, creates a stationary magnetic field to produce braking torque but mainly dissipates the associated energy as heat rather than recovering it.\n\nDrives also coordinate with production controllers. Speed references and status information may pass through analog signals, digital inputs, or industrial communication networks connected to a programmable logic controller. Applications in printing, winding, and papermaking may regulate tension, synchronize roller surface speeds, or compensate for changing roll diameter. Servo systems emphasize tightly controlled position and rapid dynamic response; although their converter technology overlaps with VFD technology, the terms describe different performance expectations rather than mutually exclusive electrical architectures."
    },
    {
      "heading": "Supply harmonics and electromagnetic compatibility",
      "text": "A drive affects both the supply network and the motor circuit, but the mechanisms differ. On the supply side, a conventional six-pulse diode rectifier draws non-sinusoidal current. Under balanced idealized conditions, its characteristic current harmonics have orders\n\n\\[h=6k\\pm1,\\]\n\nwhere \\(k\\) is a positive integer. The fifth and seventh harmonics are therefore prominent examples. Their actual magnitudes depend on line impedance, DC-link inductance, loading, supply distortion, and other installation details.\n\nHarmonic current produces additional heating and interacts with network impedance to distort supply voltage. A drive may have a displacement power factor close to unity while still having a lower true power factor because waveform distortion increases root-mean-square current without a corresponding increase in useful real power. Assessing the effect on an electrical grid or plant distribution system consequently requires more than measuring the phase angle of the fundamental current.\n\nMitigation methods include AC line reactors, DC-link chokes, passive harmonic filters, multipulse rectifiers, and active front ends. Twelve-pulse and higher-pulse arrangements combine phase-shifted rectifier supplies to cancel selected harmonics under suitable conditions. An active front end uses controlled switching to regulate input current and can support regeneration, but it introduces its own filtering, control, and compatibility requirements. IEEE 519 addresses harmonic control at the point of common coupling; its limits are not simply a universal specification for each individual drive.\n\nOn the motor side, rapid switching produces high-frequency differential-mode and common-mode voltages. Common-mode current can flow through cable capacitance, motor capacitance, protective conductors, and nearby structures. These paths can interfere with instrumentation and communications if the installation does not control high-frequency current return paths. Shield termination, cable routing, bonding, and filter placement therefore influence performance alongside the drive’s internal circuitry.\n\nElectromagnetic compatibility encompasses both emissions and immunity. The IEC 61800-3 standard addresses EMC requirements for adjustable-speed electrical power drive systems and distinguishes operating environments and equipment categories. Input EMC filters, output reactors, \\(dv/dt\\) filters, and sinusoidal filters serve different purposes and are not interchangeable. Their suitability depends on cable length, switching frequency, motor insulation, leakage current, and converter design. An installation can satisfy one objective, such as reducing conducted interference, while creating another constraint, such as increased protective-conductor current."
    },
    {
      "heading": "Motor insulation, bearings, and thermal limits",
      "text": "The motor connected to a drive experiences electrical stresses unlike those produced by a nearly sinusoidal fixed-frequency supply. Fast voltage edges travel along the motor cable as waves. Where cable and motor impedances differ, reflections can increase the instantaneous voltage at the motor terminals. Cable length, edge rise time, DC-link voltage, and termination characteristics determine the resulting stress; under unfavorable conditions, reflected steps can approach twice the incident step amplitude before further interactions are considered.\n\nRepeated steep-front pulses place particular demands on winding insulation, especially between nearby turns that may not initially share the voltage evenly. Motors described as inverter-duty incorporate insulation and thermal provisions appropriate to specified converter conditions, but the designation does not remove all cable-length or voltage constraints. NEMA MG 1 and the IEC 60034 series contain provisions relevant to converter-fed machines, with requirements varying by machine and insulation system. Output filtering may be needed where a motor’s insulation capability and the drive waveform are not compatible.\n\nCommon-mode voltage can also create shaft voltage and bearing current. Capacitive coupling through the motor can charge the rotor relative to the frame; when the lubricant film breaks down electrically, discharge pulses may pass through a bearing. Repeated discharge can damage raceways and produce characteristic surface patterns, including fluting. Other bearing-current mechanisms occur in larger machines. Countermeasures include appropriate bonding, shaft grounding, insulated bearings, common-mode filtering, and suitable cable arrangements, selected according to the actual current path.\n\nCooling presents a separate limitation. A self-ventilated motor’s shaft-mounted fan moves less air at low speed. Sustained rated torque at reduced speed may therefore cause excessive temperature even if current remains within the motor’s nominal rating. Independently powered ventilation or a reduced continuous-torque rating can address this operating condition. Temperature sensing and thermal models improve protection, but their effectiveness depends on correct parameterization and knowledge of the motor’s cooling behavior.\n\nHigh-speed operation introduces mechanical as well as electrical limits. Bearings, rotor balance, fan strength, and maximum safe rotational speed may prevent operation at frequencies the converter can readily generate. Flexible shafts and structures can have resonances within the adjustable speed range, while PWM and electromagnetic forces can excite audible vibration. Skipping a narrow frequency band may avoid sustained operation at a resonance, but it does not remove the transient crossing or replace a mechanical assessment. The motor–drive combination must therefore be treated as an electromechanical system rather than as independently rated components."
    },
    {
      "heading": "Protection, functional safety, and reliability",
      "text": "A VFD normally monitors output current, DC-link voltage, power-device temperature, and selected supply and motor conditions. Protective responses include current limiting, controlled stopping, pulse inhibition, and fault trips. Typical detected conditions include overcurrent, DC overvoltage, undervoltage, overheating, and some forms of ground fault. The coverage and response time vary by product: an electronic overload model does not detect every possible insulation defect, and converter protection is not equivalent to comprehensive protection of the entire installation.\n\nUpstream short-circuit protection and isolation remain necessary. Fuses or a circuit breaker must be coordinated with the converter and the available fault current. A drive’s stated short-circuit current rating concerns a specified equipment arrangement and associated protective devices; it should not be interpreted as the inverter’s ability to interrupt any fault without external protection. Stored energy in the DC-link capacitors also remains relevant after the incoming supply has been disconnected.\n\nFunctional-safety features address hazardous motion rather than ordinary process regulation. Safe torque off disables the drive’s ability to generate torque through designated safety-related circuitry. It does not electrically isolate the motor terminals, guarantee that an already moving shaft stops immediately, or hold a gravity-loaded axis in position. Safe stop functions can combine controlled deceleration with subsequent torque removal, while other functions monitor speed or motion under defined safety conditions.\n\nIEC 61800-5-2 specifies functional-safety requirements for safety-related power drive systems. Their use within machinery requires consideration of the entire safety function, including input devices, logic, wiring, mechanical brakes, and validation. An ordinary stop command over a control network is not automatically a safety function. Likewise, automatically restarting after a power interruption may be acceptable for an unattended ventilation process but hazardous for accessible machinery; restart behavior is part of the machine-level design.\n\nReliability depends strongly on temperature, contamination, electrical stress, and duty cycle. Cooling fans wear, heat-exchanger surfaces foul, and some capacitor technologies age as a function of time, temperature, and ripple current. Power modules experience thermal cycling as loads change, while vibration and corrosive atmospheres affect connections and circuit boards. Modern drives can report operating hours, temperatures, fault histories, and estimated component condition. These data support maintenance planning but do not provide an exact remaining lifetime. Networked drives also form part of industrial cybersecurity systems because unauthorized changes to speed limits, control modes, or firmware can affect physical equipment."
    },
    {
      "heading": "Historical development and emerging technologies",
      "text": "Before practical solid-state converters, adjustable-speed industrial drives commonly used DC motors, mechanical transmissions, or electromechanical conversion equipment. The Ward Leonard system, introduced in the 1890s, used a motor–generator arrangement to vary the voltage supplied to a DC motor, providing effective speed control at the cost of additional rotating machinery. AC motor speed could also be changed through pole switching or specialized rotor arrangements, but these methods did not provide the broad, continuously adjustable characteristics of modern electronic drives.\n\nPower electronics progressively changed this situation. Commercial thyristors appeared in the late 1950s, enabling controlled rectifiers and high-power converters with no mechanically moving switching elements. Early AC drive systems used current-source, six-step, and other converter arrangements suited to the available devices. Their size, waveform quality, torque behavior, and control complexity constrained their use compared with later transistor-based systems.\n\nA decisive conceptual advance was the development of field-oriented control around the beginning of the 1970s, associated particularly with the work of K. Hasse and Felix Blaschke. It provided a framework for controlling AC-machine torque and flux with a degree of separation analogous to that available in separately excited DC machines. During subsequent decades, microprocessors, digital signal processors, improved current sensors, and faster power devices made such calculations practical in commercial equipment. Direct torque control emerged as another influential approach during the 1980s.\n\nThe spread of IGBTs from the 1980s onward helped establish compact PWM voltage-source drives across a broad industrial power range. Increasing computational capacity enabled automatic motor identification, sensorless operation, built-in process controllers, and extensive communications. VFDs became common components of automation rather than specialized devices reserved for unusually demanding installations. Their application also broadened beyond induction motors to permanent-magnet and synchronous-reluctance machines, whose efficiency and operating characteristics can be advantageous in appropriate duties.\n\nNewer designs increasingly use wide-bandgap semiconductors, particularly silicon carbide, where their switching and voltage capabilities offer benefits. Faster switching can reduce converter losses or filter size, but can also intensify insulation stress and common-mode interference unless the complete system is designed accordingly. Other development areas include improved sensorless low-speed control, predictive control, modular medium-voltage converters, and closer integration of motors and drives. Condition monitoring increasingly combines electrical measurements with vibration and temperature data, sometimes using digital twin models to compare expected and observed behavior. The engineering trade-offs remain concrete: efficiency, dynamic response, electromagnetic compatibility, serviceability, and lifetime do not necessarily improve together."
    }
  ],
  "links": [
    "electric-motor",
    "semiconductor",
    "magnetic-field",
    "induction-motor",
    "rectified-linear-unit",
    "power-inverter",
    "capacitor",
    "insulated-gate-bipolar-transistor",
    "control-theory",
    "air-conditioning",
    "irrigation",
    "wastewater-treatment",
    "electric-generator",
    "resistor",
    "regenerative-braking",
    "programmable-logic-controller",
    "printing",
    "papermaking",
    "power-factor",
    "electrical-grid",
    "ball-bearing",
    "circuit-breaker",
    "cybersecurity",
    "automation",
    "silicon-carbide",
    "digital-twin"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}