Electronics Guide

Magnetic levitation transportation, universally shortened to maglev, replaces the wheel-on-rail contact of a conventional railway with a controlled magnetic gap. The vehicle carries no wheels in normal running, touches nothing, and is pushed along by a linear motor rather than by torque applied to an axle. These are not proposals. Maglev trains have carried paying passengers in daily revenue service since 1984, when a 600-meter shuttle opened at Birmingham Airport in England. Today the operating lines include the Shanghai maglev, which has run between Longyang Road and Pudong International Airport since 2003; Linimo in Aichi Prefecture, Japan, open since 6 March 2005; the Incheon Airport line in South Korea, open since 3 February 2016; the Changsha Maglev Express, open since 6 May 2016; and Beijing subway Line S1, open since 30 December 2017. Several shorter Chinese lines have followed.

This distinction matters because maglev is frequently confused with hyperloop, and the two are not the same thing. A hyperloop vehicle runs inside a tube held at low pressure, which is what allows its proponents to claim speeds above 1,000 kilometers per hour; no hyperloop has carried a paying passenger, and its best-funded developer wound down at the end of 2023. A maglev train runs in the open air, at atmospheric pressure, on an elevated or at-grade guideway, and it stops at ordinary stations. The levitation physics overlap, but almost nothing else does. Air resistance, not vacuum engineering, sets the practical ceiling on maglev speed. Weather, not tube integrity, is the environmental hazard. The engineering problems that dominate a maglev project are the tolerance of the guideway, the segmentation of the wayside propulsion supply, and the safety case for a railway that has no track circuits and no adhesion.

This article covers the electronics and control engineering of maglev as it is actually built and operated: the two suspension families in service, the levitation control loop and its sensors, the linear motors that propel these vehicles, onboard power transfer, guideway construction and its cost consequences, braking without friction, operational control in the absence of a rail circuit, and the record of the systems that run today. For the deeper treatment of the levitation physics shared with evacuated-tube concepts, including null-flux coil geometry, see Hyperloop Technologies.

Operating Railways, Not Proposals

Maglev has an unusual history for a transportation technology: the engineering was proved decades ago, and the obstacle has always been commercial rather than technical. Germany operated a full-scale test facility at Emsland from 1984, reaching 450 km/h with the Transrapid 07 vehicle on 17 June 1993, yet never built a revenue line at home. Japan has run superconducting test vehicles at Miyazaki and then Yamanashi since the 1970s and holds the world speed record for a crewed rail vehicle, yet its first revenue maglev line, Linimo, runs at 100 km/h over less than nine kilometers. The systems that carry passengers today are therefore a mixture of one high-speed airport link built with exported German technology and a set of medium-low-speed urban lines.

Two speed families are worth separating from the start, because they use different electronics. High-speed maglev, represented by the Transrapid vehicles in Shanghai and the Japanese SCMaglev, places the motor in the guideway and operates in the 300 to 500 kilometers per hour band. Medium-low-speed maglev, represented by the Japanese HSST design used on Linimo, the Korean UTM design used at Incheon, and the Chinese lines at Changsha, Beijing, and elsewhere, places the motor on the vehicle and operates around 100 to 140 kilometers per hour. The first competes with short-haul aviation and high-speed rail. The second competes with light rail and elevated monorail, and its selling points are low noise, tight curve radii, steep gradients, and the absence of wheel and rail wear.

A third category exists only as prototypes and test articles, treated near the end of this article. Throughout, service speeds and record speeds are named as such, because the two are routinely conflated and differ by more than a hundred kilometers per hour even on the same technology.

What Air Changes

Running at atmospheric pressure changes the engineering in ways that are easy to overlook when maglev is discussed alongside evacuated-tube concepts. Aerodynamic drag rises roughly with the square of speed, and the power needed to overcome it rises roughly with the cube, so the propulsion rating of a high-speed maglev is set almost entirely by air resistance rather than by anything magnetic. Levitation is comparatively cheap; moving air out of the way is not. This is why the Shanghai line, reported to have reduced its peak service speed from 431 to 300 kilometers per hour in 2021, gave up only about fifty seconds of journey time on a thirty-kilometer run while saving a substantial fraction of its traction energy.

Air also supplies what a sealed pod must carry. Cabin ventilation draws from outside, and cooling systems reject heat into the surrounding atmosphere rather than into the vehicle structure or a radiating tube wall. In exchange, the vehicle must survive weather, and where the route runs in tunnels it faces the pressure transients that trouble conventional high-speed rail rather than anything peculiar to a vacuum tube. Both are taken up below.

Two Ways to Levitate a Train

Every maglev in service uses one of two suspension principles. They differ in the size of the gap, in whether the suspension is stable without a controller, in whether the vehicle needs wheels, and in what the guideway must be built to. The physics of both, including the figure-eight null-flux coil arrangement, is developed at length in Hyperloop Technologies and in Linear Motors and Magnetic Levitation. What follows concentrates on the consequences for a railway that must run in air, stop at platforms, and survive weather.

Electromagnetic Suspension and the Millimeter Gap

Electromagnetic suspension, abbreviated EMS, lifts the vehicle by attraction. Electromagnets carried on the vehicle reach up around a ferromagnetic rail fixed to the underside of the guideway and pull the vehicle toward it from below. The Transrapid design holds a nominal air gap of 10 millimeters at all speeds, and the Japanese HSST design used on Linimo holds about 8 millimeters. Because the force grows as the gap shrinks, the equilibrium is unstable: nothing passive holds it. A linearized model of an attractive suspension places an unstable pole near the square root of twice the gravitational acceleration divided by the nominal gap, which for a ten-millimeter gap corresponds to roughly seven hertz. The control loop must therefore be fast enough to dominate a disturbance that doubles in a fraction of a second, and it must run continuously from the moment the vehicle lifts to the moment it sets down.

The compensations are substantial. EMS works at zero speed, so the vehicle levitates in the station, at a stand, and while creeping. It needs no wheels for normal operation, only landing skids for emergencies and for parking. The magnets that provide levitation can serve a second purpose as the excitation field of the propulsion motor, which is exactly what the Transrapid design does. Against this, the small gap places a severe demand on guideway accuracy, because a one-millimeter construction error consumes a tenth of the available clearance, and it makes ice, debris, and settlement of the supporting structure into operational problems rather than nuisances.

Electrodynamic Suspension and the Superconducting Magnet

Electrodynamic suspension, abbreviated EDS, lifts the vehicle by repulsion. Strong magnets on the vehicle induce currents in coils or conducting sheets in the guideway as the vehicle moves past, and the induced currents repel. The Japanese SCMaglev is the only EDS system built at full railway scale. Its bogies carry superconducting magnets, and the guideway sidewalls carry two sets of coils, one arrangement wound in a figure-eight pattern and cross-connected beneath the track so that the vehicle is pushed back toward the null position whenever it deviates from it. The result is a levitation gap of about 100 millimeters, ten times the Transrapid figure.

The price of that gap is speed. Induced current requires relative motion, so an EDS vehicle produces no useful lift at rest and little at low speed. The SCMaglev runs on retractable rubber-tired wheels until roughly 150 kilometers per hour, at which point the induced currents are sufficient to lift it clear, and it lowers the wheels again on the way down. Every departure and arrival therefore involves a transition between two entirely different suspension regimes, with landing gear extension and retraction sequenced against measured levitation force. The wheels, their brakes, and their tires become safety-critical items with a duty cycle unlike anything on a conventional train.

The superconducting magnets bring their own systems. They operate as persistent-current coils, so they consume no continuous excitation power once charged, but they require cryogenic refrigeration on board, monitoring for the onset of a quench, and quench-protection circuitry to dump the stored energy safely if the winding goes normal. They also produce strong stray fields, which drives magnetic shielding of the passenger compartment and of any equipment sensitive to field. Japan Central Railway announced in 2025 a next-generation development vehicle intended to use high-temperature superconducting magnets, which the company states will require less cooling and simpler refrigeration equipment; that vehicle is a development article, not a production design.

The Levitation and Guidance Control Loop

On an EMS vehicle, the levitation controller is the system on which everything else depends. Loss of control does not degrade performance; it drops the vehicle onto its skids. The loop is duplicated, distributed along the length of the train, and designed so that no single sensor, processor, or power stage can bring the vehicle down.

Sensing the Gap and the Motion

Gap measurement uses inductive or eddy-current sensors that observe the distance between the magnet pole face and the guideway rail with sub-millimeter resolution and sample at kilohertz rates. Each magnet typically carries more than one sensor, and the channels are compared so that a drifting or failed sensor can be identified and excluded rather than believed.

Gap alone is not enough, because gap is measured relative to the guideway, and the guideway is not an inertial reference. It deflects under the weight of the passing vehicle, it steps slightly at the joints between beams, and it is built to a tolerance rather than to perfection. A controller that drove gap error to zero against every guideway feature would faithfully transmit the shape of the track into the passenger compartment. Practical designs therefore add accelerometers on the magnet frame, giving an inertial reference, and combine the two measurements so that the magnet follows the guideway while the car body does not. A secondary suspension between the levitation frame and the car body, typically air springs, filters what remains.

Control Law and Power Stage

The controller is usually a state-feedback design over gap, gap rate, and magnet current or flux, wrapped around an inner current loop, implemented on a digital signal processor or field-programmable gate array with a deterministic execution period. Gains are scheduled against operating point, because the plant is nonlinear: the force depends on the square of the current and inversely on the square of the gap.

The power stage is a four-quadrant chopper per magnet, fed from the vehicle direct-current bus. Its rating is set less by average power than by the rate at which it must change current. A levitation magnet is a large inductance, and the achievable rate of change of current is the applied voltage divided by that inductance, so the bus voltage effectively sets how fast the suspension can respond to a disturbance. The average power is modest by traction standards; the bandwidth is not.

Redundancy and Touchdown

Magnets are grouped into many independently controlled segments spaced along both sides of the vehicle, each with its own sensors, controller, and chopper. A failure removes one segment, redistributes its share of the load to its neighbors, and reduces margin, but it does not remove levitation. Batteries on board hold levitation and control for a defined period after loss of wayside power, so that the vehicle can coast to a halt while still floating. When the reserve is exhausted, or when the failure is severe enough, the vehicle sets down on landing skids, which are a designed operating mode rather than a failure of structure.

Lateral Guidance

Vertical support is only half the problem. High-speed EMS designs add separate guidance magnets on the sides of the vehicle undercarriage that act against steel guidance rails fixed to the guideway, controlled by loops of the same kind as the levitation loops. Medium-low-speed designs of the HSST family generally exploit geometry instead: the levitation magnet module straddles the rail so that a lateral displacement produces a restoring force without a dedicated guidance actuator, which suits an urban vehicle that never runs fast enough to need more. On the SCMaglev, lateral centering comes from the null-flux coil arrangement itself and requires no active control at speed.

The vertical and lateral loops are not independent. Curve entry, crosswind gusts, and passenger load shifts couple them, and combined controllers distribute force across the magnet groups to hold the vehicle centered in three dimensions while limiting current in any one magnet. Because the vehicle wraps around the guideway rather than resting on top of it, the geometry itself makes derailment impossible; the loops are protecting ride quality and clearance, not preventing the vehicle from leaving the track.

Linear Propulsion and Wayside Power

A maglev has no adhesion, so tractive effort must come from a magnetic field acting directly on the vehicle. The machine that does this is a linear motor: a rotary motor conceptually cut open and laid flat, so that the traveling field moves along a track instead of rotating in a bore. The design question is which half of the motor goes on the vehicle.

Why the Motor Lives in the Guideway

High-speed maglev puts the three-phase winding in the guideway and the field on the vehicle. This is the long-stator linear synchronous motor. In the Transrapid design the stator packs and windings are mounted along the underside of the guideway cantilevers, and the vehicle's levitation magnets serve simultaneously as the excitation of the motor, so no separate propulsion magnet is required. In the SCMaglev the ground coils in the sidewalls perform propulsion while the superconducting magnets provide the field.

Four consequences follow, and together they explain why every high-speed maglev has made this choice. First, the vehicle carries no traction converter and no traction transformer, so its mass is dominated by structure and payload rather than by propulsion equipment. Second, the power that must be transferred to the moving vehicle collapses from megawatts of traction to tens of kilowatts of auxiliary load, which makes contactless transfer feasible. Third, thrust is limited by installed wayside converter capacity rather than by what a vehicle can carry, so acceleration can be maintained to high speed and steep gradients become tractable; Transrapid literature cites route gradients up to ten percent, against roughly four percent for conventional high-speed rail. Fourth, and least welcome, the motor must be installed over the entire route whether or not a vehicle is present, which moves a large share of the system cost into the fixed infrastructure.

Being synchronous, the motor locks the vehicle to the traveling field. Speed is the supply frequency multiplied by twice the pole pitch, which means that at 500 kilometers per hour the wayside converters run at frequencies of a few hundred hertz, and it also means that vehicle speed is a wayside command rather than a driver's choice. Thrust is set by current magnitude and by the load angle between the field and the vehicle magnets. To keep that angle controlled, the wayside converter needs to know where the vehicle is, not merely how fast it is going, so vehicle position measured on board and reported over a redundant radio link becomes part of a control loop closed by equipment the vehicle never touches.

Block Switching Between Substations

Energizing an entire route at once would be absurd, so the long stator is divided into sections, each connected through switchgear to a converter in a wayside substation. Only the section occupied by a vehicle is energized. Published maglev practice puts section lengths in the range of a few hundred meters to a few kilometers, with each section fed by a converter rated in the megawatts. The route infrastructure therefore includes propulsion switch stations at intervals along the guideway, alongside the radio antenna masts that carry the control link.

The interesting engineering is at the boundaries. A vehicle at 430 kilometers per hour covers about 120 meters every second, so it crosses from one section to the next in a time that leaves no room for a gap in thrust. Designs handle this by arranging that adjacent sections can be fed from alternate converters, so the next section is energized and phase-aligned before the vehicle reaches it and the two overlap briefly while thrust transfers. The handover must preserve both the magnitude and the phase of the field relative to the vehicle, since a phase error appears immediately as a thrust transient the passengers feel.

Section switching also produces a safety property for free. A vehicle can move only where its section is energized, and a section can be energized for only one vehicle. Separation between trains is therefore enforced by the propulsion system itself rather than by a separate signaling overlay, a point developed further below.

Short-Stator Propulsion on Urban Lines

Medium-low-speed maglev makes the opposite choice. Linimo, the Incheon Airport line, the Changsha Maglev Express, and Beijing Line S1 all carry linear induction motors on the vehicle, working against a passive aluminum reaction rail with a steel backing fixed to the guideway. Power comes from a conductor rail, 1,500 volts direct current on both Linimo and Line S1, feeding an onboard variable-voltage variable-frequency inverter of the sort used in any modern light-rail vehicle.

The engineering trade is straightforward. A linear induction motor with a large mechanical air gap suffers significant magnetizing current and end effects, so its efficiency and power factor are poor compared with a rotary machine, and it must carry its own inverter and cooling. At 100 kilometers per hour, none of that matters much. What matters is that the guideway becomes a simple structure carrying a levitation rail, a reaction plate, and a conductor rail, rather than a distributed motor, which brings the cost per route-kilometer down to something an urban transit budget can absorb. These vehicles are also conventional in every operational respect but levitation: Linimo runs driverless under automatic train control and automatic train operation, exactly as a rubber-tired automated people mover would.

Onboard Power and the Auxiliary Load

Moving the traction motor into the guideway does not eliminate the vehicle's need for electricity. An EMS vehicle must energize its levitation and guidance magnets whenever it floats, including at a stand in a station, and every vehicle needs air conditioning, lighting, control electronics, door actuators, and communications. Supplying that load to a vehicle that touches nothing is a design problem in its own right.

The Transrapid answer is layered. At standstill and low speed, contact power rails on the guideway supply the vehicle directly, which is unproblematic because the vehicle is barely moving. Above roughly 80 kilometers per hour, linear generator windings integrated into the pole faces of the levitation magnets harvest energy from the harmonic content of the long-stator field, so the vehicle draws its auxiliary power from the same traveling field that propels it, without contact. Later Transrapid vehicles were designed to remove the need for physical contact at any speed. Onboard batteries buffer the transitions and, more importantly, hold levitation and control for a defined interval if wayside power is lost.

The SCMaglev has a different load profile and reached a different answer. Its superconducting magnets need no continuous excitation power, but the cryogenic refrigeration does, and so do the usual hotel loads. The original L0 Series vehicles carried gas turbine generators in the end cars, a solution that put a combustion engine on a train. The improved L0 Series introduced from 2020 replaced them with inductive power transfer from the guideway, eliminating the turbines. That change is worth noting as a general trend: contactless power transfer at the tens-of-kilowatts scale has matured enough to displace an onboard prime mover on a passenger vehicle.

Medium-low-speed lines sidestep the problem entirely. A conductor rail already supplies traction power, so auxiliaries, levitation magnets, and the inverter all draw from the same 1,500-volt bus, and the vehicle looks electrically like a light-rail car with an unusual set of loads.

The Guideway Sets the Cost

Ask what makes a high-speed maglev expensive and the answer is not the vehicle. It is the guideway, because the guideway is simultaneously a bridge, a motor, a power distribution system, and a precision machine surface, and it must be all of those over every kilometer of the route.

What the Beam Carries

A Transrapid guideway beam supports the weight of the vehicle and transfers it to the ground, and it also provides the mounting for the functional components: the guidance rails, the slide rails on which a set-down vehicle rests, the stator packs, the motor windings, the power rails, and the vehicle location reference flags. The standard cross-section is a trapezoidal box girder with the functional components mounted under each cantilever, so the vehicle wraps around the beam from below. Three standard beam types were developed, with double-span lengths of about 62, 25, and 6.2 meters and heights of 2, 1, and 0.4 meters respectively, covering elevated, at-grade, and tunnel applications.

Stiffness is a functional requirement rather than a structural nicety. The beam deflects under the vehicle, and that deflection appears directly in the ten-millimeter gap the levitation controller is trying to hold. Manufacturers developed beams in steel, in prestressed concrete, and in a hybrid form that combines a prestressed concrete main section with steel cantilever elements fabricated in 3.1-meter modules, on the reasoning that steel accepts the precision machining that the functional mounting surfaces demand while concrete provides mass, stiffness, and damping at lower cost. Beams are prefabricated by largely automated processes in climate-controlled facilities, often purpose-built alongside the route, precisely because the required accuracy is not achievable with ordinary field construction.

The arithmetic explains the fuss. At 430 kilometers per hour the vehicle crosses a 25-meter beam in roughly one-fifth of a second, and any departure of the mounting surfaces from the design contour is a direct fraction of a ten-millimeter clearance. Straight beams are built with a two percent cross-fall for drainage, and curved beams may be canted up to twelve degrees to offset lateral acceleration. The Shanghai project used single-span hybrid beams of about 25 meters standard length and required more than 2,500 of them for a thirty-kilometer double-track line.

Foundations, Settlement, and Service Life

The tolerance problem does not end when construction does. Foundations settle, and the guideway is designed for a service life of about eighty years with regular maintenance, so the alignment must be surveyed and adjusted over decades rather than set once. Soil conditions dominate the substructure cost: on the Shanghai alignment, poor ground required many of the piles supporting the columns to reach depths beyond forty-eight meters. A conventional railway absorbs a few millimeters of settlement in the ballast. A maglev guideway resolves it in shims, or it appears in the ride.

Switching

A maglev cannot use a conventional turnout, because there is no wheel flange to guide and the vehicle wraps around a continuous beam. Transrapid solves this with a bending switch: a multi-span welded steel beam that is elastically bent into position by electromechanical rack-and-pinion drives mounted on alternate supports, then locked. The low-speed version weighs about 300 tons over a 78-meter beam and permits 100 kilometers per hour in the diverging position; the high-speed version weighs about 600 tons over a 148-meter beam and permits 200 kilometers per hour diverging, with full speed straight. Both are designed for roughly one million cycles, or twenty to thirty years of typical service.

These are impressive machines and also a structural constraint on network design. A junction is a several-hundred-ton actuator with a movement time measured in seconds, not the light, fast, cheap device that lets a conventional railway build dense networks and complex station throats. Maglev topologies therefore tend toward point-to-point lines with few junctions, which reinforces the commercial difficulty discussed at the end of this article.

Braking Without Adhesion

A conventional train stops by pressing friction material against wheels or rails and by regenerating through traction motors whose torque reaches the rail through adhesion. A maglev has neither friction interface nor adhesion, so braking must be reconstructed from other principles, in layers, with each layer independent of the failure that disabled the one above it.

The primary service brake on a high-speed maglev is the linear motor itself. Commanding the traveling field behind the vehicle turns the motor into a generator and returns energy through the wayside converter. Deceleration is limited by passenger comfort rather than by any physical contact, and it does not vary with weather, contamination, or wheel condition, which removes the single largest source of uncertainty in conventional train braking. The catch is architectural: the motor belongs to the wayside. If the propulsion supply fails, the vehicle has no service brake at all, and this is precisely why an independent onboard brake is not optional.

The independent layer on the Transrapid design is an eddy-current brake. Brake magnets on the vehicle act against the conductive side guidance rails, inducing currents in them and generating a retarding force, powered from onboard batteries so that it works with the wayside dead. Two properties shape how it is used: the force falls off as speed falls, so it cannot bring a vehicle to a complete stop, and the dissipated energy heats the rail, which limits how often it can be applied. Below the speed at which it becomes ineffective, the vehicle sets down on its landing skids and friction against the slide rails completes the stop. Setting down is a last resort, because it is preferable to keep the vehicle levitating on battery power until it halts naturally where circumstances allow.

The SCMaglev has the wheels it needs for low-speed running already, so its lowest braking layer is conventional: disc brakes on the retractable landing wheels. Its primary brake is likewise regenerative through the ground coils, and aerodynamic brake panels have been used on Japanese test vehicles as an additional high-speed layer. Medium-low-speed lines, whose motors are on board, keep braking on board too, combining regenerative and dissipative braking through the vehicle inverter with a fallback that sets the vehicle down.

One property is common to all of them and deserves emphasis. Because braking does not depend on adhesion, a maglev is not subject to wheel slide, to the leaf-film and low-adhesion problems that plague autumn railway operation, or to the seasonal derating of braking distance that follows from them. Braking distance is a deterministic function of speed and available brake capacity, which simplifies the safety case considerably.

Operational Control Without Track Circuits

Conventional railway signaling rests on a piece of physics that a maglev does not have. A steel axle bridging two running rails short-circuits a track circuit, proving that a train occupies a section and doing so in a way that fails safe when a rail breaks or a circuit dies. A maglev vehicle touches nothing and there are no running rails to shunt. Train detection, and with it the whole apparatus of interlocking and separation, must be rebuilt on different foundations.

High-speed maglev rebuilds it on the propulsion system. Because the wayside drives the vehicle, the wayside already commands its speed and position; a vehicle cannot move where no section is energized, and a section is energized for one vehicle at a time, so exclusive occupancy is a property of the power distribution rather than an assertion by a signaling system. Position is measured on board against reference markers installed along the guideway, one of the functional components carried on every beam, and reported to the operation control system over a redundant radio link. Speed is not selected by a driver and cannot be exceeded, because the vehicle is synchronous with a field whose frequency the wayside sets. Overspeed protection, the central preoccupation of conventional automatic train protection, is intrinsic.

Several classical railway hazards disappear with the wheel. Derailment is impossible for a vehicle that wraps around its guideway. Broken rails, wheel failures, and hot axle boxes have no analogue. Buckling of track under heat affects a guideway differently, since it is a series of bridge spans on bearings with designed expansion gaps rather than a continuous welded rail.

What remains is obstruction and human procedure, and the fatal accident in the history of the technology came from exactly there. On 22 September 2006, a Transrapid vehicle on the Emsland test track struck a maintenance carriage standing on the guideway, at a speed of about 170 kilometers per hour, killing twenty-three people. The cause was procedural: the vehicle was released before the maintenance vehicle had cleared the track. No property of magnetic levitation contributed, and none prevented it. The lesson generalizes beyond maglev. A system whose technical safety argument is strong can still be defeated by the interface between the automated system and the people working on it, and the possession and work-authorization procedures that a mature railway spends a century refining are not supplied by the propulsion architecture.

Practical maglev safety therefore leans on the same disciplines as any other guided transport system. Elevated guideway keeps the running surface out of reach of road traffic and pedestrians. Intrusion detection, obstacle detection, and positive interlocking of maintenance access address what elevation does not. Evacuation planning must account for a vehicle stopped on an elevated beam with no adjacent walkway unless one is deliberately provided. And because commands travel to the vehicle over a radio link, the control channel needs authentication, integrity protection, and monitoring, treated as part of the safety case rather than as an information-technology concern. Approaches to these problems, developed for conventional railways, are covered in Signaling and Train Control.

Ride Quality, Aerodynamics, and Weather

Ride quality is the visible output of everything described above, and it is assessed with the same instruments used on conventional railways: whole-body vibration measured and weighted according to ISO 2631, and ride indices of the Sperling type that condense acceleration spectra into a single comfort number. Accelerometers in the car body, at floor level and at seat interfaces, provide the measurements, and the same data serve maintenance by revealing guideway sections whose geometry has drifted.

The maglev advantage is real but narrower than the marketing suggests. Removing the wheel removes the rolling-contact excitation that dominates the vibration spectrum of a conventional train, and it removes the impulsive input from rail joints and wheel flats. It does not remove guideway geometry error, aerodynamic buffeting, or the disturbance from crosswinds, and at high speed those are what remain. The levitation controller must reject them within the millimeters it has, and the secondary suspension must absorb what the controller deliberately does not.

Weather is the operating environment that separates a maglev railway most sharply from an evacuated-tube concept. Crosswinds load an elevated vehicle laterally and the guidance loops must hold against gusts, which in practice means anemometers along exposed sections and speed restrictions when wind exceeds thresholds. Snow and ice accumulate on the functional surfaces of the guideway, and on an EMS design a few millimeters of ice is a significant fraction of the clearance. Lightning strikes elevated structures, so the guideway, the wayside converters, and the vehicle electronics need coordinated surge protection and bonding. Thermal expansion moves the beams over their bearings and opens or closes the joints between spans, and the levitation controller crosses those joints hundreds of times a minute.

Tunnels introduce a problem shared with conventional high-speed rail rather than with hyperloop. A vehicle entering a tunnel at high speed compresses the air ahead of it, generating a pressure wave that travels to the far portal and can radiate as an audible micro-pressure wave, and the pressure transients inside the tunnel act on the car body and on passengers' ears. Sealed bodies, controlled cabin ventilation, portal hoods, and long tapered noses are the standard mitigations. They matter disproportionately for the Chuo Shinkansen, roughly ninety percent of which is planned to run in tunnel.

The Systems in Service

The operating record is short enough to state completely, which is unusual for a transportation technology more than forty years old and is itself the most important fact about maglev.

Shanghai

The Shanghai maglev is the only high-speed maglev in commercial passenger service anywhere. It runs 29.86 kilometers from Longyang Road station to Pudong International Airport, using Transrapid TR08 vehicles built by the Siemens and ThyssenKrupp joint venture, on a hybrid guideway built by a Chinese consortium. A preview run took place on 31 December 2002 and commercial service began in 2003, at a reported construction cost of about ten billion yuan.

Its speeds are frequently misquoted, so it is worth stating them precisely. The peak commercial service speed was 431 kilometers per hour, giving a journey time of about seven minutes and twenty seconds, and the line is reported to have operated at a reduced peak of 300 kilometers per hour since 2021, giving a journey time of about eight minutes and ten seconds. Separately, and not in passenger service, the line recorded 501 kilometers per hour during testing on 12 November 2003. The 501 figure is a record; the 431 and 300 figures are service speeds.

Linimo

Linimo, formally the Aichi Rapid Transit Tobu Kyuryo Line, opened on 6 March 2005 to serve the Expo 2005 world exposition near Nagoya and remained afterward as an ordinary urban railway. It runs 8.9 kilometers with nine stations at a maximum of 100 kilometers per hour, floating about 8 millimeters above the track. It uses the Japanese HSST electromagnetic suspension design with onboard linear induction motors, draws 1,500 volts direct current from a conductor rail, and operates without a driver under automatic train control and automatic train operation, the first uncrewed commercial urban maglev. Ridership fell sharply after the exposition, from roughly 31,000 passengers per day during Expo to about 12,000 in the following six months, and has since settled near 16,500 per day.

Incheon

The Incheon Airport maglev opened on 3 February 2016, running 6.1 kilometers with six stations from Incheon International Airport, developed by the Korea Institute of Machinery and Materials with Hyundai Rotem. It uses electromagnetic suspension with linear induction motor propulsion and a design maximum of 110 kilometers per hour, although it has operated well below that in practice. Its history illustrates the commercial fragility of demonstration lines: service was suspended from 14 July 2022 and resumed on 17 October 2025 after more than three years, reopening as a tourism and experience facility with limited daily hours and long headways rather than as everyday public transportation.

Changsha and Beijing

The Changsha Maglev Express opened on 6 May 2016, linking Changsha South railway station to Changsha Huanghua International Airport over 18.55 kilometers using Chinese medium-low-speed maglev technology. It began at 100 kilometers per hour with rolling stock rated for 120, and reporting places an increase of the maximum service speed to 140 kilometers per hour with new rolling stock in July 2021, giving an end-to-end journey of about nineteen and a half minutes. An extension with additional airport terminal stations has been under construction.

Beijing subway Line S1 opened on 30 December 2017 and was the third Chinese maglev line in commercial operation. It runs 10.2 kilometers with eight stations in the western suburbs at a service maximum of 100 kilometers per hour, with six-car trains about 89 meters long built by CRRC Tangshan, drawing 1,500 volts direct current from a third rail and propelled by onboard linear motors. Additional short Chinese medium-low-speed lines have followed, including tourist-oriented services at Fenghuang in Hunan and at Qingyuan in Guangdong.

Lines No Longer Running

Two European systems came first and are gone. The Birmingham maglev shuttle linked Birmingham Airport to Birmingham International railway station from 1984 to 1995, a 600-meter run at 42 kilometers per hour using electromagnetic suspension and linear induction motors, and was the first commercial maglev anywhere. The Berlin M-Bahn operated briefly around 1989 to 1991 before reunification restored the conventional line whose alignment it occupied.

Transrapid in Germany: What Happened

Germany developed the technology that runs in Shanghai and never built a revenue line of its own, and the reasons are instructive because none of them are magnetic.

The Transrapid Versuchsanlage Emsland, a 31.5-kilometer test facility, opened in 1984 and served as the development and demonstration site for eighteen distinct guideway beam designs and for successive vehicle generations through the TR08, which entered test service there in the autumn of 1999. The facility recorded 450 kilometers per hour with the TR07 on 17 June 1993, a test record rather than a service speed. By the early 2000s more than 330,000 passengers had ridden demonstration runs there.

Two events ended the program. The first was the accident of 22 September 2006 described above, which killed twenty-three people and, whatever its procedural cause, removed the political room for a domestic project. The second was arithmetic. The proposed Munich airport link was cancelled on 27 March 2008 after its estimated cost rose from about 1.85 billion euros to well over three billion, the increase driven principally by tunneling and civil engineering rather than by the maglev equipment. The Emsland facility closed in 2011 when its operating license expired, and although demolition was approved in 2012, dismantling was repeatedly deferred while alternative uses were explored, and the site was cleared much later.

The pattern generalizes. In every maglev project that has failed to proceed, the decisive number has been the cost of building a new, dedicated, high-tolerance guideway through developed land, not the cost or the readiness of the levitation and propulsion technology.

The Chuo Shinkansen and the 600-Kilometer Class

Japan has pursued superconducting maglev since the 1970s, and its test program holds the speed records. On 2 December 2003 the crewed MLX01 reached 581 kilometers per hour on the Yamanashi test line, and on 21 April 2015 the crewed L0 Series reached 603 kilometers per hour on the same line, which by then had been extended from its original 18.4-kilometer priority section to 42.8 kilometers. Both are test records set on a test track, not service speeds.

The revenue application is the Chuo Shinkansen, a 285.6-kilometer line from Shinagawa in Tokyo to Nagoya, roughly ninety percent of it in tunnel, planned for a maximum service speed of 505 kilometers per hour and a journey time of about forty minutes. Major construction began on 17 December 2014. The project has been repeatedly delayed, principally by a dispute with Shizuoka Prefecture over the effect of tunneling on the flow of the Oi River. The original target of 2027 has been abandoned, and by 2025 the operator no longer projected an opening before the middle of the following decade; reporting in 2026 indicated that the prefecture had accepted a revised environmental mitigation plan, potentially allowing the disputed section to begin. No opening date is fixed, and none should be assumed. A later extension to Osaka remains a plan rather than a schedule.

China has pursued the high-speed EMS route instead. CRRC unveiled a prototype vehicle designed for 600 kilometers per hour at Qingdao in July 2021, using electromagnetic suspension with a long-stator linear synchronous motor, and testing has continued since. It is a prototype: no line exists for it to run on in commercial service, and its design speed is neither a record nor a service speed.

Where Maglev Fits

Comparison with conventional high-speed rail is fairer on service speeds than on records. The wheel-rail speed record stands at 574.8 kilometers per hour, set by a specially prepared French TGV trainset on 3 April 2007, against the crewed maglev record of 603 kilometers per hour; the gap is under thirty kilometers per hour. In service the picture is different in the other direction: modern high-speed rail runs at 320 to 350 kilometers per hour on many routes, while the single high-speed maglev in service has run at 431 and now reportedly at 300. Maglev's speed advantage is genuine but has never been demonstrated over a long revenue route.

The decisive commercial objection is not speed but incompatibility. A high-speed train can leave its dedicated line and continue over the existing network into a city center terminal, reusing a century of sunk investment. A maglev cannot: it requires a purpose-built guideway for every kilometer it travels, and every journey must begin and end at a maglev station. That converts what would be an incremental upgrade into a standalone system with no network effects, which is why the Shanghai line, technically successful, terminates at Longyang Road rather than in central Shanghai and why proposed extensions have not been built.

Where maglev is defensible, the case rests on the properties that follow from having no wheels. Steep gradients become acceptable, which can shorten a difficult alignment. Curve radii can be tighter at a given speed on urban lines. Noise at low and moderate speed is very low, since rolling noise is absent, which matters for elevated urban structures. Wear at the vehicle-guideway interface effectively disappears, moving maintenance from consumable mechanical parts toward electronics and structure. Against these, the guideway carries the motor and its power distribution over the whole route, junctions are large slow machines, and the whole system stands outside the existing rail network. Those trade-offs favor dense point-to-point corridors and short urban lines, which is exactly the pattern the operating record shows.

Conclusion

Maglev is a mature technology with a small operating footprint. The suspension, propulsion, and control problems were solved decades ago and have been demonstrated in daily passenger service since 1984: electromagnetic suspension holds a millimeter-scale unstable gap through continuous feedback and works from standstill, while electrodynamic suspension with superconducting magnets earns a hundred-millimeter gap at the cost of wheels below roughly 150 kilometers per hour and cryogenics on board. High-speed designs put a long-stator linear synchronous motor in the guideway, which strips traction equipment off the vehicle, makes speed a wayside command, and turns the segmented propulsion supply into the mechanism that enforces separation between trains in the absence of any track circuit. Urban designs keep a linear induction motor on the vehicle and keep the guideway cheap.

The engineering lessons that generalize are about where the difficulty actually lives. Levitation control is demanding but bounded; the guideway, which must be a bridge, a motor, a power distribution network, and a precision machine surface across every kilometer, is where the cost and the tolerance problems concentrate. Braking and train separation had to be rebuilt from principles other than friction and shunted rails, and the resulting architecture is in several respects safer than conventional practice. The one fatal accident in the technology's history came from a procedural failure that no propulsion architecture could have prevented. And the reason maglev remains rare is neither physics nor reliability but the fact that a maglev line cannot connect to anything that already exists.

Related Topics