Electric scooter motors: geared hub vs direct-drive hub

“250 W motor” on an electric scooter spec sheet is just a number. Behind it sits a specific architecture that determines whether your scooter will be quiet, whether it will recover energy, whether it will pull up a hill and how much it will actually weigh. This article covers the three main drivetrain configurations found in modern electric scooters: chain drive (historical, kids’ category), the geared hub motor with a planetary reducer (geared hub), and the direct-drive hub motor (direct-drive / gearless hub), on which the overwhelming majority of adult electric scooters today are built — from the Xiaomi M365 to the NAMI Burn-E.

First: BLDC instead of brushed

Almost every modern electric scooter is driven by a brushless DC motor (BLDC). This is a distinct technology that displaced the older brushed DC motors. In a brushed motor, graphite brushes physically rub against the commutator to transmit current to the rotor windings — they wear out, spark and produce heat; efficiency is usually 70–80 %. In a BLDC motor the windings sit on the stator and the rotor is permanent magnets; current in the winding phases is switched by an electronic controller that reads rotor position from Hall sensors (sensored controller) or, less commonly, from the motor’s own back-EMF (sensorless). Per OEM surveys, modern BLDC hub motors hold a steady 85–90 % efficiency, have no rubbing parts and last thousands of hours. A sensored controller is needed because at a standing start a sensorless setup does not know the rotor position and can jerk, especially uphill. (Dewesoft; Greensky Power; Upbeat Geek)

The controller as a separate module (six-step vs sine-wave/FOC, MOSFET set, sensored vs sensorless) is covered in the article on electronics; here we focus on the motor itself.

Architecturally, a BLDC can be mounted next to the wheel and transmit torque through a chain or belt (as in the kids’ Razor E100) or integrated directly into the wheel hub as a “hub motor”. A hub motor, in turn, can be either geared or direct-drive. From this come the three configurations listed below.

1. Chain drive (Razor E100 and the kick-scooter kids’ niche)

This is the oldest and, today, almost exclusively a kids’ configuration. A separate motor is attached to the frame next to the rear wheel and turns it through a chain or toothed belt. In the canonical example — the Razor E100 (since 2003) — this is a 24 V brushed DC motor of 100 W with a 9-tooth sprocket and a #25 chain. The motor housing is about 100 × 68 mm, mass ~3 kg. Razor themselves write: “100-watt high-torque single-speed chain-driven motor”. (Razor; Amazon; MotoTec)

Why this scheme has remained only in the kids’ niche:

• Transmission losses. The chain costs several efficiency points and requires tensioning, lubrication and replacement. A hub motor does without all of this.
• Size and mass. An external motor “eats” space next to the frame, making it harder to build a foldable, compact wheel.
• Noise. A metal chain is louder than a quiet BLDC hub.
• The ASTM F2641 standard (covered in the article on scooter types) for kids’ models allows low-power brushed DC motors: price matters more than efficiency, because a child does not ride tens of kilometres a day. That is why Razor still fits a simpler, cheaper brushed motor with a chain.

In the adult category, chain drive is found only in some retro scooters and home-brew conversions. Integrating the motor into the wheel has become the standard.

2. Geared hub motor

This is a BLDC motor inside the wheel hub with a planetary reducer. A small high-speed rotor in the centre spins 4–5 times faster than the wheel itself; between them sits a planetary gear train (a sun gear plus three planets) that lowers the rpm and, by the same ratio, multiplies the torque. A typical reducer ratio is 5:1: the motor makes five turns per one wheel turn and delivers five times the torque it would have produced directly. (Hentach; Marsantsx)

Strengths:

• High torque at low rpm — pulls away from a standstill, climbs hills and feels confident in stop-and-go traffic.
• Smaller size and lower mass. At the same output power a geared hub is 30–50 % lighter than a comparable direct-drive, because the motor itself can be smaller (its job is to spin fast, not strong). (Levy Electric)
• Freewheel. Most geared hubs put an overrunning clutch between motor and wheel: when the throttle is released, the wheel spins freely and does not drag the magnets and gears with it. There is no “cogging” (the magnetic drag from the stator that passively brakes the wheel).

Weaknesses:

• No regenerative braking. That same overrunning clutch which gives the freewheel mechanically disengages the motor from the wheel when the throttle is released — so the motor cannot brake the wheel or charge the battery. Both the e-bike and e-scooter industries confirm this. (Fluid Free Ride; Electric Bike Report)
• Gear-mesh noise. Plastic or metal planets produce a characteristic “whirr” of 50–60 dB — quieter than conversation, but noticeably louder than a direct-drive hub. (Hentach)
• Gear wear. Nylon gears with reinforcement last thousands of kilometres, but this is still a service item — unlike a direct-drive hub, where there is essentially nothing to wear.

Where you find this in electric scooters. In the e-bike world, geared hubs dominate the moped and cargo categories. In electric scooters, geared hubs today are mostly cheap, ultra-light or high-torque off-road models. One reason is that an e-scooter rides in a cruise mode of 20–40 km/h, where a direct-drive hub is more efficient; and intense standing starts, where a geared hub shines, are not as critical as on a pedal-equipped bicycle. A recent attempt to address the geared hub’s main shortcoming came from Grin Technologies: in 2019 it introduced the GMAC — a geared hub without a clutch, capable of regen precisely because there is no freewheel. In production electric scooters such a scheme is still rare. (Electrek)

3. Direct-drive hub motor (gearless)

This is the dominant configuration in modern adult electric scooters: from the 250-watt Xiaomi M365 to the 8.4-kilowatt NAMI Burn-E 2 Max. A direct-drive BLDC hub is, in effect, “an inside-out motor in the wheel”: the axle is stationary and holds the stator with copper windings, while the rotor with permanent magnets is the wheel shell itself. The electronic controller energises the winding phases in turn; the magnetic field pushes the magnets, and the wheel spins. There are no gears. (Fluid Free Ride; Levy Electric; Unagi)

Strengths:

• Regenerative braking (KERS). Because the motor is always rigidly coupled to the wheel, the controller can “flip” the motor’s role — make the magnets-in-the-wheel induce current into the stator. That current charges the battery while simultaneously braking the wheel. A classic example is the Xiaomi M365: a soft press on the brake lever engages KERS in the front hub motor (3 levels in the Mi Home / Ninebot app — Weak / Medium / Strong), and a harder press adds the mechanical rear disc brake. (Wikipedia; eBike Choices)
• Quiet running. Without gears the motor only hums at the electromagnetic switching frequencies — 40–45 dB, quieter than conversation. This is one reason sharing operators (Lime, Bird, Dott) fit direct-drive: night riding in residential areas does not annoy people. (Hentach)
• Simplicity and reliability. There is nothing to wear: the axle bearings, the stator windings and the rotor magnets. In practice a direct-drive hub runs tens of thousands of kilometres with almost no servicing (the only wear point being the external axle bearings) — for fleet-rotated sharing scooters (Bird Three, Lime Gen 4) this longevity is critical.
• High efficiency at cruising speed. In steady-state riding at 25–40 km/h a direct-drive hub returns 88–90 % efficiency, because there are no reducer losses. (Levy Electric)

Weaknesses:

• Lower torque at low rpm. Without a reducer the motor has to pull “with its own muscle”; from a standstill and on a steep hill a direct-drive hub is noticeably slower than a comparable geared one. Manufacturers compensate with significantly higher nominal power (350–1 500 W instead of 250 W) and dual-motor configurations — see below.
• “Cogging” (magnetic drag). Magnets always pass by the iron stator teeth, creating a weak braking effect even when the motor is off. On the wheel this feels like a faint “pull-and-release”, and on a descent it reduces coast distance.
• Mass and bulk. A direct-drive hub is heavier than a geared hub at the same nominal power. On small 8.5–10″ wheels this is not critical, but on powerful models (NAMI, Dualtron) each rear motor wheel weighs several kilograms (the NAMI Burn-E motor itself is around 5.5 kg, with the tyre and rim on top). (Fluid Free Ride)

Market examples (all direct-drive BLDC hubs)

• Xiaomi M365 / Mi 4 — front motor wheel 250 W nominal / 500 W peak, ~16 N·m of torque, 36 V. Three levels of KERS regen. (Voltride; Wikipedia)
• Segway-Ninebot KickScooter MAX G30 — rear motor wheel 350 W nominal, with IPX7 rating on the motor itself. Sensored controller, regenerative brake. (Segway)
• INOKIM Light 2 — rear 350 W nominal / 650 W peak, 15 N·m, gearless BLDC, 13.5 kg total scooter mass. (Rider Guide; Electrek)
• Apollo City / City Pro — dual-motor 2 × 500 W BLDC hubs in the front and rear wheels, combined peak 2 000 W, independent control of front and rear motors. (Apollo Scooters; Electric Scooter Insider)
• Dualtron Thunder 3 — 2 × 1 500 W BLDC hubs in tubeless 11″ wheels, peak combined output up to 11 000 W, 72 V × 40 Ah LG pack. (Dualtron USA; NYC PEV)
• NAMI Burn-E 2 / Burn-E 2 Max — 2 × 1 000 W (Burn-E 2) with a 5 000 W peak, or 2 × 1 500 W (Burn-E 2 Max) with an 8 400 W peak; 50-amp sinewave controllers with front-rear torque-balance tuning across 5 modes. (Fluid Free Ride; Hyper Rides; Rider Guide)

How to read the “motor” line in a spec

Manufacturers are generous with big numbers, but it pays to know what each figure refers to:

• Nominal (continuous) power — what the motor delivers in sustained operation without overheating. This is the figure regulators care about (eKFV: ≤ 500 W; Ukraine PLET: ≤ 1 000 W — see regulation in 2010–2020 and 2020–2026).
• Peak (max) power — a brief maximum: standing start, overtake, hill. Usually 2–5× the nominal. Do not confuse with nominal in a legal context.
• Torque (N·m) — a more realistic “pull” metric than watts. Most manufacturers do not publish it; when they do, the typical consumer range is roughly 15–50 N·m per wheel (M365 ≈ 16 N·m, INOKIM Light 2 at 15 N·m), with performance models quoting noticeably more.
• Sensored vs sensorless controller — sensored, with Hall sensors, starts smoothly from zero; sensorless is cheaper and lighter, but may jerk at start. For an adult urban scooter this is almost always sensored.
• Sinewave vs square-wave controller — sinewave delivers smoother current into the phases, a quieter motor and lower heating losses; square-wave is cheaper and simpler. Most modern performance models (NAMI, Dualtron, Apollo Pro) explicitly state sinewave.
• Single vs dual motor — dual-motor models carry two BLDC hubs, one in each wheel, with independent control. This gives all-wheel drive (AWD), a better launch, the ability to ride on one motor to save charge, and a higher peak power — at the cost of mass, price, and the fact that such machines fall outside the legal limits of urban-legal consumer classes.

Summary

The next chapters of this guide cover batteries (what real range depends on), brakes (disc, drum, electronic, foot) and suspension and wheels (pneumatic vs solid, IP protection). The motor sets the upper bound of what a scooter can do; the rest of the running gear decides whether it does it safely.

Related topics

Each entry names which §-section of this article it attaches to and which §-section of the target article it points at. Engineering deep-dives provide the full physical/control foundation behind statements in this parts article.

• Motor and controller engineering: BLDC, PMSM, FOC, MOSFET inverter — the full engineering baseline behind the §“First: BLDC instead of brushed” preamble. Park’s 1929 dq0 transform (§5 of the deep-dive), Clarke + sensorless back-EMF estimation (§5), MOSFET inverter sizing and gate drive (§6) — that is the math this parts article compresses into the single phrase “BLDC hubs hold 85–90 % efficiency”. §3 of the deep-dive (copper / iron Steinmetz / eddy losses) explains why direct-drive loses to geared at low rpm and why cogging arises.
• Acceleration and throttle control engineering — §3 (sensored vs sensorless controller) and §“How to read the motor line” (sinewave vs square-wave) are the user-facing surface of §1–§3 of the deep-dive (Hall-sensor angle resolution of 60° vs back-EMF zero-crossing detection), §6 (jerk + soft-start ramping) and §10 (TCS slip-limit). The soft-start difference between a cheap square-wave and a sinewave FOC is described right here; the deep-dive supplies the friction-circle + jerk physics behind it.
• Regenerative braking: KERS physics and control — §3 (direct-drive KERS) is the user-facing surface of §1 of the deep-dive (energy balance), §2 (controller flip — torque vector reversal), §3 (battery charge-acceptance limit at low SoC) and §5 (Xiaomi M365’s three-level regen). Why a geared hub with a freewheel clutch cannot regen — §4 of the deep-dive (mechanical decoupling at the overrunning clutch).
• Controllers, BMS, and IoT for e-scooters — §3 (sensored vs sensorless controller) and §3.6 (sinewave vs square-wave) — a parts-level article covering the controller as a separate module: six-step vs FOC, MOSFET set, BMS interface. This article explicitly defers the controller treatment there in the lead paragraph.
• Batteries and real range — §“How to read the motor line” (nominal vs peak power) is the other side of §3 of the batteries-real-range deep-dive (Coulombic efficiency + Peukert) and §4 (continuous discharge rating). The 2 000 W peak in the Apollo City Pro is only possible because 18650 cells sustain a 30 A burst — that is the battery’s constraint.
• Brake system engineering: hydraulic, drum, regenerative — §3 (KERS in the M365’s front wheel + mechanical disc in the rear) is the canonical dual-system architecture described in §8 of the deep-dive (eABS lever-sensor blending). Regen and mechanical brake are two loops with different response times (KERS ~15 ms, hydraulic ~80–150 ms).
• Anti-lock braking system (ABS) engineering: slip ratio, modulator, control loop — §3 (KERS on direct-drive) overlaps with §6 of the ABS deep-dive (control-loop interaction): when regen and ABS are combined, the controller must shed regen torque quickly to avoid wheel lock-up. Bosch’s eABS single-channel (Niu KQi3 Pro) is exactly that blending.
• Thermal management engineering: heatsink, conduction, convection — §“Sensored vs sensorless” and §“Peak power” — these are the user surface of §2 of the deep-dive (Arrhenius — every +10 °C of winding temperature halves insulation life), §4 (IP-ingress reduces convection — the IPX7 direct-drive hub in a MAX G30 is effectively a sealed thermally-insulated case) and §5 (continuous power ≡ thermal steady state). Why peak power is 2–5× nominal — that is a thermal, not electrical, limit.
• NVH (noise, vibration, harshness) engineering for e-scooters — §2 (gear whirr 50–60 dB) and §3 (EM hum 40–45 dB) — quantitative data from §3 of the deep-dive (A-weighted sound pressure at 1 m), §5 (gear-mesh frequency Z·n/60), §7 (PWM switching harmonics as the standard source of EM hum). Why sharing operators (Lime, Bird) choose direct-drive — §9 of the deep-dive (residential-zone night ordinance budgets typically 40–45 dBA).
• Tire engineering: rolling resistance, grip, standards — §3 (cruising at 25–40 km/h — direct-drive is more efficient) is the user-facing surface of §3 of the tire deep-dive (Crr mode as a function of speed: speed-independent floor + speed² hysteresis term). So “direct-drive is more efficient at cruise” = baseline Crr × motor efficiency × η_controller integrated over the duty cycle.
• Bearing engineering: ISO 281 L₁₀ life, ABEC/ISO 492, lubrication — §3 (axle bearings in a direct-drive hub as the only wear point) is a pointer to §3 of the bearing deep-dive (L₁₀ = (C/P)^p × 10⁶ rev), §6 (a 2RS lip seal with 0.5–2 N preload → IPX5 — that limit is set by the bearing engineer), §11 (false-brinelling on long-term parked vehicles). The >20 000-mile service interval is computed precisely from ISO 281.
• Climbing hills and gradeability — §2 (a geared hub’s high torque at low rpm) is the dependency formalised in §2 of the climbing deep-dive (F_grade = m·g·sin θ), §4 (torque-at-the-wheel requirement → reducer ratio + motor stall torque) and §6 (heat-rate limit on a 6–8 % grade — a direct-drive hub thermally risks derating without power management).
• Real-world range and the energy budget — §“How to read the motor line” (continuous vs peak) is the user-facing surface of §2 of the real-world-range deep-dive (energy budget = Σ (P_dt) integrated over the duty cycle), §4 (motor + controller efficiency map → average η as a function of throttle position) and §5 (regen contribution is typically 5–15 % of consumed energy in urban stop-and-go cycles).
• Display, throttle, and error codes — quick reference — §3 (a sensored Hall sensor failure → an error code) — this is the parts-level reference that maps the controller-side error-code semantics of motor faults: Xiaomi error 14/15 (Hall sensor open), Ninebot error 10 (motor stall), Apollo Pro E1 (phase open). That parts article names the symptom, this one names the cause.
• Xiaomi M365: the platform history — §3 (the canonical KERS-implementation example) is the history of the 250 W + 3-level KERS pattern itself, which all later bottom-tier urban scooters copied. §2 of the historical deep-dive (Ninebot’s 2018 acquisition + the 2018-04 release) and §5 (M365 → 1S → Pro → Pro 2 → 4 Pro motor lineage) — these are the canonical reference on which this parts article builds its baseline.

Sources

ENG-first consolidated bibliography, clustered by article §-section. Each entry is a link plus a 2–3-word context marker for fast scan-finding. Russian-language sources are not used — where a primary source is not available in English, preference is given to the official manufacturer document, a peer-reviewed paper or the relevant UN/ECE/IEC/ISO standards body.

§“First: BLDC instead of brushed” — BLDC architecture + efficiency

1. Dewesoft — Optimizing BLDC motor efficiency in e-scooters — measurement methodology + 85–90 % efficiency claim source.
2. Greensky Power — OEM’s Guide to BLDC for E-Scooters — OEM design parameters: pole count, slot count, magnet grade.
3. Upbeat Geek — Sensored vs sensorless BLDC controllers (practical guide) — Hall vs back-EMF position estimation trade-off.
4. Hanselman, D. C. — Brushless Permanent Magnet Motor Design, 2nd ed., Magna Physics Publishing, 2006, ISBN 978-1-881855-15-7 — canonical BLDC textbook (back-EMF waveform, slot-pole combinations, cogging torque).
5. Miller, T. J. E. — Brushless Permanent-Magnet and Reluctance Motor Drives, Oxford University Press, 1989, ISBN 978-0-19-859369-8 — foundational BLDC vs PMSM distinction (trapezoidal vs sinusoidal back-EMF).
6. Pyrhönen, J., Jokinen, T., Hrabovcová, V. — Design of Rotating Electrical Machines, 2nd ed., Wiley, 2013, ISBN 978-1-118-58157-5 — stator winding theory + iron loss (Steinmetz equation).
7. Wikipedia — Brushless DC electric motor — open reference; cross-checked against Hanselman 2006 §1–2.
8. Park, R. H. — “Two-Reaction Theory of Synchronous Machines — Generalized Method of Analysis — Part I”, AIEE Transactions, vol. 48, pp. 716–727, 1929, DOI 10.1109/T-AIEE.1929.5055275 — seminal dq0 rotating-frame transformation (basis for FOC).
9. Clarke, E. — Circuit Analysis of A-C Power Systems, Wiley, 1943 — αβ stationary-frame transformation (companion to Park).

§1 Chain drive (Razor E100 niche)

10. Razor — E100 Electric Scooter (official product page + spec sheet) — 100 W high-torque chain-driven motor; 24 V SLA pack; ASTM F2641-compliant.
11. Amazon listing — Razor E100 OEM MY6812 100 W chain-drive motor — replacement-part datasheet: 9-tooth sprocket, #25 chain, 100×68 mm housing.
12. MotoTec — Electric motor 24V 100W for Razor E100/E125/E150 (replacement) — cross-OEM spec confirmation.
13. ASTM F2641-08(2015) — Standard Specification for In-Line Skates, Roller Skates, and Skateboards with Powered Locomotion, ASTM International — kids’ powered-rideable safety floor (≤24 V, brushed DC permitted).
14. Wikipedia — Razor USA (company history + product line) — 2000 launch of A-model, 2003 E100 release, chain-drive lineage.

§2 Geared hub motor (planetary reducer)

15. Hentach — The Ultimate Guide to Ebike Hub Motors 2026 (technology, performance) — planetary reducer 5:1 typical ratio; gear-mesh whirr 50–60 dB.
16. Marsantsx — Planetary gears in hub motors (e-bike hub motor guide) — sun + 3 planets + ring topology; nylon-glass-filled vs steel gear life.
17. Levy Electric — Understanding the Mechanics of Electric Scooter Hub Motors — geared 30–50 % lighter than DD at same nominal power.
18. Fluid Free Ride — Electric scooter motors guide (geared vs direct-drive) — overrunning clutch + freewheel as the regen blocker.
19. Electric Bike Report — Direct drive vs geared hub motors (e-bike comparative) — same architectural finding from the sister e-bike industry.
20. Hentach — Geared hub motor vs direct drive (which is better for…) — gear noise quantification + service intervals.
21. Electrek — Grin Technologies unveils GMAC clutchless geared hub motor with regen — 2019 GMAC introduction (geared without freewheel → regen-capable).
22. Norton, R. L. — Design of Machinery, 6th ed., McGraw-Hill, 2019, ISBN 978-1-260-11331-0 — planetary gear-train kinematics, gear ratio derivation, fundamental tooth strength (AGMA equation).
23. Budynas, R. G., Nisbett, J. K. — Shigley’s Mechanical Engineering Design, 11th ed., McGraw-Hill, 2019, ISBN 978-0-07-339820-4 — gear-tooth fatigue (Lewis bending equation), Hertz contact stress.
24. ISO 6336-1:2019 — Calculation of load capacity of spur and helical gears, ISO — formal load-capacity calculation, basis for OEM bearing/gear life prediction.

§3 Direct-drive hub motor (gearless) + KERS

25. Unagi Scooters — What is a BLDC motor on an electric scooter — inside-out motor topology in plain language.
26. Wikipedia — Xiaomi M365 (model history + KERS detail) — front-hub 250 W / 500 W peak; 3-level KERS (Weak/Medium/Strong) via the Mi Home app.
27. eBike Choices — Xiaomi M365 regenerative braking — measured KERS strength + impact on coast distance.
28. Voltride — E-scooter motors catalogue (M365 / 1S / Pro motor lineage) — phase-resistance + torque-constant values across the M365 family.
29. Segway — Ninebot KickScooter MAX G30LP (official spec page) — IPX7 on the motor, 350 W nominal, sensored controller.
30. Rider Guide — INOKIM Light 2 review — 350/650 W rear hub, 15 N·m torque, 13.5 kg total mass.
31. Electrek — INOKIM Light 2 review (2020) — long-term service intervals; brushless gearless reliability.
32. Apollo Scooters — City 2024 (tech specs, dual-motor) — 2×500 W BLDC AWD; independent front/rear control.
33. Electric Scooter Insider — Apollo City Pro review — measured climb performance, regen on both wheels.
34. Dualtron USA — Thunder 3 (official product page) — 2×1 500 W BLDC hubs in 11″ tubeless wheels; 72 V × 40 Ah LG.
35. NYC PEV — Dualtron Thunder spec sheet — phase-current + KMC controller pairing.
36. Fluid Free Ride — NAMI Burn-E 2 (official US distributor product page) — 2×1 000 W with 5 000 W peak; 50 A sinewave controllers.
37. Hyper Rides — NAMI Burn-E 2 Max (NZ distributor) — 2×1 500 W with 8 400 W peak; 5-mode front/rear torque balance.
38. Rider Guide — NAMI Burn-E 2 Max review — measured 0–30 km/h, peak power draw on a hill, thermal de-rate behaviour.
39. Krishnan, R. — Permanent Magnet Synchronous and Brushless DC Motor Drives, CRC Press, 2010, ISBN 978-0-8247-5384-9 — direct-drive vs geared motor design trade-offs (torque density, cogging torque, slot-pole combos).
40. Mohan, N., Undeland, T. M., Robbins, W. P. — Power Electronics: Converters, Applications, and Design, 3rd ed., Wiley, 2003, ISBN 978-0-471-22693-2 — regen power flow: motor → DC bus → battery (four-quadrant inverter operation, §15).
41. Ehsani, M., Gao, Y., Longo, S., Ebrahimi, K. — Modern Electric, Hybrid Electric, and Fuel Cell Vehicles, 3rd ed., CRC Press, 2018, ISBN 978-1-4987-6177-2 — EV-class regen physics (charge acceptance + DC-link voltage rise + battery efficiency limits).
42. Hung, J. Y., Ding, Z. — “Design of Currents to Reduce Torque Ripple in Brushless Permanent Magnet Motors”, IEEE Transactions on Industry Applications, vol. 29, no. 4, pp. 798–804, 1993, DOI 10.1109/41.184826 — cogging-torque minimization technique (relevant to §3 “cogging” weakness).

§“How to read the motor line” — spec interpretation

43. eKFV (Verordnung über die Teilnahme von Elektrokleinstfahrzeugen am Straßenverkehr) — Bundesgesetzblatt, full text — 500 W nominal-power ceiling (§1 (1) Nr. 1c).
44. Wikipedia — eKFV (English summary of German e-scooter regulation) — comparative cross-check.
45. UNECE Regulation No. 79 — Uniform provisions concerning the approval of vehicles with regard to steering equipment, Rev. 4, 2018 — power-mapping + cruise-control safety constraint, relevant to “sensored vs sensorless” at urban speeds.
46. UNECE Regulation No. 10 — Uniform provisions concerning the approval of vehicles with regard to electromagnetic compatibility, Rev. 6, 2019 — sinewave vs square-wave EM emission boundary (PWM switching harmonics).
47. IEC 60034-1:2017 — Rotating electrical machines — Part 1: Rating and performance, IEC — formal definition of “continuous” (S1) vs “short-time” (S2) duty cycles (basis for nominal vs peak distinction).
48. SAE J1939 — Serial Control and Communications Heavy Duty Vehicle Network, SAE International — CAN-bus protocol used in higher-end performance scooters (NAMI, Dualtron) for front-rear torque coordination.
49. TI SPRABZ4 — Sensorless-FOC With Flux-Weakening and MTPA for IPMSM Motor Drives — sensorless control whitepaper; explains start-up jerk on cold sensorless setups.
50. TI SPRABQ2 — Sensorless Field Oriented Control of 3-Phase Permanent Magnet Synchronous Motors — sliding-mode observer + back-EMF estimation.
51. Microchip AN1017 — Sinusoidal Control of PMSM Motors with dsPIC30F DSC — sinewave-drive implementation reference.