EV Motion Sickness Is a Jerk Problem: Rate-Limiting the Torque Request, With or Without Software

My new video breaks down why switching regen off often doesn't save your rear-seat passengers. Jerk, the time derivative of acceleration, is what the vestibular system struggles to predict. Kill regen and you simply move the jerk to the friction brakes: an EV motor answers in milliseconds, without the torque-converter "sponge" of an ICE drivetrain. Add a heavy, stiffly sprung chassis and no engine sound to warn passengers, and sensory conflict follows.

Three zero-cost fixes from the video:
1. Rubber-band rule: rate-limit your right foot. How fast you change pedal position matters more than how deep.
2. Finish slowing before the corner; hold the pedal steady mid-corner.
3. Give passengers cues: fresh air, eyes on the road ahead, synthetic drive sound on.

Watch: https://youtu.be/QT1mQkBbV_A

Controls folks: would you fix this in the torque-request path, or in the driver?

Bottom line: If your passengers get sick in an EV, the variable to regulate is jerk (da/dt), not whether regenerative braking is on or off. Seen through a controls lens, the fix is a slew-rate limit on the torque command, and until the vehicle software does that well, the rate limiter is the driver's right foot.

What the video covers

My latest video on The Engineering Core YouTube channel goes after a popular piece of forum advice: "turn regen off and the car will feel like a gas car, so nobody gets sick." Its argument:

• The culprit is jerk. The brain is a prediction machine. When acceleration changes abruptly, the otoliths and semicircular canals report motion that the eyes, looking at a still cabin, do not confirm.
• Regen off just relocates the jerk. Lift-and-stab on the friction brakes, on a heavy car, can be rougher than smooth, computer-controlled regen.
• The ICE "sponge" is gone. Torque converters, engine rev-up time and hydraulic lag used to low-pass filter the driver's foot. An EV motor responds in milliseconds.
• Chassis and cues. Two-tonne curb weights push engineers toward stiff springs and low body roll, and there is no engine note to warn passengers what's coming.
• Three habits: rubber-band pedal control, one decel and one accel per corner, and sensory anchoring (airflow, forward view, synthetic sound).

Controls translation: the accelerator pedal is a torque setpoint. The driver closes the loop with vision, preview and proprioception. The rear-seat passenger is an open-loop plant with no preview at all.

1. Jerk is the regulated variable

Peak acceleration is not what makes people sick. A bullet train cruises at 300 km/h and nobody reaches for a bag. What matters is how acceleration changes. Monica Jones of the University of Michigan Transportation Research Institute told ABC News that when her team varied longitudinal jerk in an automated vehicle while holding peak acceleration constant, participants were "very sensitive to the jerk," and the highest-jerk condition produced the earliest onset of sickness. If you write a spec, write it on jerk, not just on g.

2. The rubber-band rule is a slew-rate limiter

The video tells drivers to imagine a thick rubber band between foot and pedal: take about a second to lift off, and when braking, press lightly, let the suspension settle, then build pressure. In block-diagram terms, that is a rate limiter on the torque request, followed by a two-stage ramp that lets pitch dynamics settle before demand rises further. Vehicle software can do the same with pedal-map shaping and torque ramp-rate limits. The trade-off is familiar to anyone who has tuned a drive: a softer ramp is more comfortable, but it adds perceived lag. Aggressive factory pedal maps sit at the other end of that curve.

3. Regen isn't innocent, but the on/off switch is the wrong knob

Here I want to add some nuance to the video's headline. In an on-road study with 16 motion-sickness-susceptible participants, Xie et al. (HKUST, 2025) found that a high regen level did induce more sickness than a low one, and that auditory motion cues reduced symptoms under high regen. So regen strength is a gain that amplifies the problem. The control-relevant question is how the deceleration is built and released: how smoothly the regen/friction blending hands off, and how the lift-off torque is ramped. Switching regen off while keeping binary footwork removes one jerk source and adds another.

4. The driver-in-the-loop is operating in the wrong band

ISO 2631-1 weights motion-sickness exposure (Wf) over roughly 0.1–0.5 Hz. Golding et al. (2001) observed maximum nauseogenicity near 0.2 Hz for fore-aft oscillation, though Griffin and Mills (2002) found no significant difference across 0.2–0.8 Hz when peak velocity was held equal. A period of 2–10 s is exactly the rhythm of a high-gain driver pumping the pedal in traffic, which amounts to a limit cycle parked in the most provocative band. The cornering advice (brake on the straight, hold constant pedal mid-corner, roll on at exit) decouples longitudinal and lateral inputs so the passenger's head isn't driven on two axes at once. The sensory advice adds the missing feedforward: Griffin and Newman (2004) found a forward road view reduced carsickness, and D'Amour et al. (2017) found airflow reduced visually induced motion sickness.

Open questions

• The video blames high-frequency chassis jitter. Most of the literature places motion sickness below 1 Hz, with higher frequencies linked to discomfort. The Guardian cites a 2024 study correlating EV seat vibration with sickness severity; I have not checked the primary paper (unverified).
• The video presents a trigeminal-to-vagal mechanism for why cool air helps. Airflow's effect has experimental support, but I found no direct source for that pathway (unverified).
• OEM torque ramp rates and pedal maps are not public, and nothing here is based on instrumented vehicle data.

Watch

▶ Why EVs Cause Car Sickness (It’s Not Just Regen Braking!) | The Physics of EV Motion Sickness

If you were writing the torque-request path, where would you put the jerk limit: in the pedal map, in a ramp limiter, or in regen blending?

Sources:

• Xie W. et al. (2025), Exploring the Effects of Regenerative Braking and the Auditory Cues for Alleviating Motion Sickness in Electric Vehicles, IJHCI 41(24): doi:10.1080/10447318.2025.2499155
• ABC News (2024), 'It's a real thing': Drivers and passengers report motion sickness in EVs: abcnews.com
• Golding J.F. et al. (2001), A motion sickness maximum around the 0.2 Hz frequency range of horizontal translational oscillation: ResearchGate
• Griffin M.J. & Mills K.L. (2002), Effect of frequency and direction of horizontal oscillation on motion sickness: PubMed
• Griffin M.J. & Newman M.M. (2004), Visual field effects on motion sickness in cars: PubMed
• D'Amour S. et al. (2017), The efficacy of airflow and seat vibration on reducing visually induced motion sickness: doi:10.1007/s00221-017-5009-1
• ISO 2631-1:1997, whole-body vibration (Wf motion-sickness weighting): iTeh catalog
• The Guardian (2025), Do electric vehicles make people more carsick?: theguardian.com

Published on The Automation Core. Video by YouTube @TheEngineeringCore-v (Tech Why). Engineering explainer, not medical advice.