Active suspension: from adaptive damping to full body control

“Active suspension” is used for systems with very different authority: some vary damper resistance, change ride height slowly or apply an anti-roll torque, while a smaller group can generate controlled force at each corner. The meaningful question is not whether electronics are present, but which force the system can command, in which direction, over what stroke and bandwidth.

A precise taxonomy

Passive suspension

A passive system has fixed mechanical characteristics during the drive. Its springs and dampers may be nonlinear, frequency-selective or position-sensitive, but no controller commands them in real time.

Manual damper adjustment remains passive after the setting is chosen. A dashboard “sport” mode that changes only steering and accelerator mapping does not make the suspension adaptive.

Self-levelling and ride-height control

Air or hydraulic systems can add or remove fluid to recover a target static height, raise the body for clearance or lower it at speed. This is an active height function, but it is often too slow to control each bump. It should not be confused with fully active vertical suspension.

Semi-active suspension

A semi-active system changes a dissipative property, most commonly damping. In the ideal definition it can vary the force opposing relative suspension motion but cannot command an arbitrary force that adds mechanical energy. Proportional hydraulic valves and magnetorheological fluid are two ways to change that force envelope.

Semi-active control can make the suspension soft for isolation at one instant and add damping for body control at the next. It remains constrained by the current shaft velocity: when body and wheel have no relative velocity, an ideal damper has no useful damping force to vary. suspension dampers

Active roll control

An active anti-roll system applies a controlled torque between the left and right sides of an axle. Electromechanical actuators, hydraulic circuits or split anti-roll bars can oppose body roll while reducing passive cross-coupling on a one-wheel bump.

This is genuine active control in the roll degree of freedom, but it does not necessarily control heave or pitch and may not generate an independent vertical force at all four wheels. Calling it “fully active suspension” would overstate its authority.

Fully active suspension

A fully active system uses powered actuators to create controlled force between body and wheel at individual corners. Depending on the hardware, it can oppose or deliberately produce heave, pitch and roll and can move the suspension even when there was no preceding relative motion.

Springs are normally retained to carry the static load efficiently and provide a mechanical baseline. The active system then supplies dynamic force. ZF's sMOTION, for example, adds an electrically driven hydraulic pump to a two-valve damper so the unit can move the suspension rather than only resist motion. ZF: sMOTION active chassis dampers

Preview control

Preview describes information, not actuator type. A camera, lidar, stored map or the measured response of a front wheel can tell the controller about a disturbance before a rear wheel reaches it. That signal can prepare an adaptive damper, air valve or active actuator. A preview sensor does not by itself create suspension force.

The control problem

The controller estimates body and wheel states from accelerometers, height sensors, wheel-speed sensors, steering angle, brake and drive requests, yaw rate and other vehicle data. It then chooses force targets subject to physical limits.

Typical objectives conflict:

reduce body acceleration for comfort;
reduce heave, pitch and roll for posture control;
keep suspension movement away from bump and rebound limits;
limit variation in tyre load for grip;
maintain ride height and aerodynamic targets;
minimize actuator energy, noise and heat.

A controller cannot minimize every objective at every frequency. “Skyhook” strategies conceptually damp the body relative to an inertial reference; “groundhook” strategies emphasize unsprung motion and tyre load. Production control blends objectives and changes their weighting with speed, drive mode, load, grip and available actuator force.

Software cannot repeal the mechanical limits established in suspension construction and geometry and suspension springs. The tyre still has finite stiffness and grip, the links still define a wheel path, and the actuator still has finite force, velocity and stroke.

What active force can change

During cornering, outward roll is caused by the lateral-acceleration force acting at the body's centre of gravity relative to its roll-force paths. An active system can add a counteracting roll moment or even incline the body inward. During braking and acceleration, it can add pitch-support forces. Over a bump, it can move the wheel relative to the body or move the body to manage occupant acceleration.

The same actuator can also redistribute vertical wheel loads. That can help keep a loaded tyre within a useful operating region, but it cannot create grip independently of the road, tyre and total normal load. Aggressively holding the body level is not automatically the setting that gives the best ride or transient handling.

Fully active control also makes unusual body motions possible. Entry lift, deliberate corner inclination, “dancing” and jumping are different commands applied to the same broad ability to generate vertical force. A spectacular demonstration confirms actuator authority; it does not independently establish ride comfort, durability, energy use or emergency behavior.

Current production approaches

Porsche Active Ride and Audi active suspension

The standard current Porsche Taycan chassis uses two-chamber air springs and continuously controlled two-valve dampers. Porsche Active Ride adds an electrically driven hydraulic pump to each active damper, generating volume flow and wheel-specific force on demand. Porsche says the system controls pitch and roll and can incline the body to reduce the acceleration felt by occupants. Porsche: Taycan Active Ride

ZF identifies its sMOTION damper-and-valve technology as the basis of this active function. Separate external valves regulate compression and rebound, while the pump moves the piston rod actively. This hardware distinction is more informative than the shared word “adaptive.” ZF: sMOTION active chassis dampers

Audi applies the related concept to the e-tron GT family. Its standard two-chamber/two-valve air suspension and optional active suspension are separate configurations; Audi describes wheel-load distribution, pitch and roll compensation, and a 55–77 mm stationary comfort-entry lift depending on starting height. Audi: e-tron GT active suspension

Mercedes-Benz E-ACTIVE BODY CONTROL

Mercedes-Benz E-ACTIVE BODY CONTROL combines an air spring that carries the basic load with 48-volt hydraulic actuation at each corner. Mercedes says it regulates spring and damping forces individually to counter heave, pitch and roll. Its ROAD SURFACE SCAN uses a stereo camera to prepare the struts for detected road undulations, and a CURVE mode inclines the body inward. Mercedes-Benz: E-ACTIVE BODY CONTROL and ROAD SURFACE SCAN

This architecture shows why preview and full activity are separate attributes. The camera improves anticipation when it can identify the road input; the hydraulic actuator supplies the force. The suspension can continue to react from onboard motion sensors when preview information is incomplete.

NIO SkyRide

NIO describes the ET9's SkyRide as an integrated hydraulic fully active suspension. The company claims the system can respond to information within 1 ms and adjust stiffness, damping and height. Those are manufacturer claims for the production system, not an independent measurement of every operating condition. NIO: ET9 SkyRide full active suspension

The glass-tower demonstration is useful for visualizing low-frequency body control over a prepared sequence. It does not reveal tyre-load variation, actuator power, behavior on random broken surfaces, thermal limits or long-term durability. Those require instrumented tests.

BYD DiSus and the Yangwang U9

BYD uses DiSus as an umbrella for several levels of body control: DiSus-C for damping control, DiSus-A for air-body control and DiSus-P for hydraulic body control. This is a clear example of one brand name spanning semi-active and more active hardware. BYD: DiSus body-control system taxonomy

The Yangwang U9's DiSus-X demonstrations show that the vehicle can command large, rapid body and suspension movements, including a jump. The crouch, launch and landing make the presence of powered vertical actuation visible.

The jump does not prove that the tyres remain better connected to an ordinary road, and it should not be described as obstacle detection unless a specific demonstration documents sensing and autonomous timing. Its defensible engineering meaning is narrower: the system has enough controllable force and stroke to add substantial vertical energy to the vehicle.

Preview is useful but conditional

Reactive control begins after sensors detect wheel or body motion. Preview can reduce that delay, but perception creates new limits:

A camera must distinguish road shape from markings, shadows, standing water, snow and low-contrast surfaces.
The controller must transform a measured feature into the path of each tyre.
Vehicle speed, steering and body motion determine when the wheel will arrive.
The actuator needs enough stroke, force and response to execute the command.
A disturbance outside the sensor field, such as a fresh pothole hidden by another vehicle, still needs reactive control.

Wheelbase preview avoids some vision problems by using a front-wheel response to prepare the rear axle, but it cannot help the first axle. Stored-location lift functions are useful for known ramps; they are ride-height automation, not high-bandwidth road preview.

Power, heat and fail-safe behavior

Active force is not free. Electrical input becomes hydraulic work, mechanical work and heat. Pumps, motors, accumulators, valves and cooling add mass and packaging. Continuous high-force operation can be limited by electrical power, fluid temperature, actuator velocity or duty cycle.

An EV's high-voltage or 48-volt system can supply substantial chassis power, but any range penalty or aerodynamic gain is vehicle- and route-specific. Lowering at speed may reduce drag; active body control consumes energy. Neither effect should be generalized without measured data.

Safety also depends on the degraded state. A sensor, pump, valve, communication bus or power supply can fail. The spring and mechanical wheel control must continue to support and locate the vehicle, while the controller detects faults and moves to a defined mode. An active chassis should therefore be assessed by its warnings, limp-home behavior, service procedures and replacement cost as well as its best demonstration.

What buyers should check

Ask what the option actually contains:

fixed, frequency-selective, manually adjustable or electronically variable dampers;
steel or air springs, and whether the air system changes height, volume or both;
active anti-roll control only, or independent force at each corner;
reactive sensing, camera preview, wheelbase preview or stored-location lift;
selectable modes and whether each changes spring, damper, actuator or only steering and powertrain response;
restrictions by wheel size, drivetrain, temperature, towing mode or battery state;
warranty coverage and the cost of a damper, air spring, compressor, pump and calibration.

On a test drive, look beyond flat cornering. Check sharp-edge isolation, long-wave body control, one-wheel inputs, braking pitch, repeated bumps, noise and the transition between modes. A good active system expands the range of usable behavior while remaining predictable. Complexity without a well-developed mechanical baseline and control calibration does not achieve that.