User guide

Emergency maneuvering is a discipline distinct from planned braking and from steady-state cornering. There is no time for a second attempt — there is one decision made in 0.5–1.5 seconds and one motor sequence executed in the next 0.3–0.8 seconds. If the decision is wrong (you brake when you should have swerved, or you swerve when you should have stopped), two-wheeled physics with small wheels and a high center of gravity punishes you immediately: 86 million shared trips on e-scooters in 2019 ([NACTO — Shared Micromobility in 2019](https://nacto.org/wp-content/uploads/2020/08/2019sharedmicromobilityreport_final.pdf)) generate 118,485 ED visits in 2024 ([CPSC — E-Scooter and E-Bike Injuries Soar, 2024](https://www.cpsc.gov/Newsroom/News-Releases/2024/E-Scooter-and-E-Bike-Injuries-Soar-2022-Injuries-Increased-Nearly-21)), and CPSC explicitly notes that e-scooters have much higher centers of gravity and smaller wheels with less shock absorption, so pavement quality matters significantly more than it does for bikes or e-bikes. Small wheels and a tall CoG mean that the same patch of damaged pavement that a cyclist will absorb as a transient ride-quality blip will throw an e-scooter rider over the handlebars. This guide covers the two symmetric skills the Motorcycle Safety Foundation (MSF) calls core emergency skills: **threshold braking** (maximum deceleration at the edge of wheel lockup) and **emergency swerve** (rapid line change without braking during the lean phase). Plus — when to use which, and when to combine them sequentially. ENG-first sources: MSF Basic RiderCourse / 'Do I Brake or Swerve' / Quick Video Tips, Wikipedia (Countersteering, Threshold braking, Dooring), CyclingSavvy (Emergency Maneuvers, Door Zone Tragedy), Cycle World and MCrider (target fixation), AASHTO (2.5 s PIEV), CPSC injury reports, IIHS sidewalk speed studies, Nature Communications (projected time-to-collision e-scooter), ScienceDirect (e-scooter vs bicycle crash typology), 99% Invisible (Dutch Reach), Bennetts (brake and swerve), Hupy and URide (emergency drill protocols).

Engineering deep-dive into the load-bearing structure of an e-scooter — parallel to the introductory overview «Frame, handlebar, and folding mechanism» (parts/frame-handlebar-folding): beam mechanics under combined loading (bending stress σ = M·c/I from Euler-Bernoulli + torsional shear τ = T·r/J + axial σ = F/A → von Mises σ_v = √(σ²+3τ²) ≤ σ_y as the yield criterion for 3D stress state; section modulus Z = I/c for a round tube I = π(D⁴−d⁴)/64 — second moment of area is quartic in diameter, so a 2-mm wall in a 50-mm tube has 8× the bending stiffness of the same 2-mm wall in a 25-mm tube); materials (Young's modulus E_6061-T6 = 68.9 GPa + σ_y = 276 MPa + ρ = 2.70 g/cm³ vs E_7075-T6 = 71.7 GPa + σ_y = 503 MPa vs E_7005-T6 = 72 GPa + σ_y = 290 MPa vs E_6082-T6 = 70 GPa + σ_y = 260 MPa vs E_4130_Cr-Mo = 205 GPa + σ_y = 460 MPa with ρ = 7.85 g/cm³ vs E_Mg_AZ91D = 45 GPa with ρ = 1.81 g/cm³ vs CF UD T700S E_long = 135 GPa with ρ = 1.55 g/cm³ → σ_t/ρ ≈ 1645 kPa·m³/kg, the best specific strength; Ashby material selection chart specific stiffness E/ρ vs specific strength σ_y/ρ — why 6061-T6 is the universal choice through the combination of weldability + corrosion resistance + price, not maximum strength); welding metallurgy (GTAW gas tungsten arc welding AC for aluminum — alternating current breaks the Al₂O₃ oxide film with melting point 2050 °C; HAZ overaging T6 precipitation-hardened → T4 solid-solution → annealed with ~50% yield-strength reduction in the heat-affected zone 276 MPa → 138 MPa per AWS and Aluminum Association D1.2; filler 4043 Al-5Si low cracking susceptibility vs 5356 Al-5Mg higher strength with post-weld natural aging vs 4047 Al-12Si no aging response; why 7075 is unweldable in thin-wall frames through precipitation hardening destruction + hot cracking susceptibility — used only locally as a CNC-machined part bolted onto a 6061 frame; why frames have welded gussets — additional reinforcement ribs compensate for the 50% HAZ knockdown); fatigue physics (Basquin equation σ_a = σ'_f · (2N_f)^b with fatigue strength coefficient σ'_f and exponent b = −0.05…−0.12 for metals; high-cycle HCF >10⁴ cycles vs low-cycle LCF <10⁴ cycles; critical difference — aluminum has no endurance limit per ASM Handbook Vol. 19 and ISO 12107: all aluminum alloys keep losing strength linearly on log-log scale as N → ∞, whereas steels 4130 / 4140 have a horizontal endurance limit ≈ 0.5·σ_UTS at N ≥ 10⁷ cycles; Goodman/Soderberg/Gerber diagrams for mean stress correction; Miner's linear damage hypothesis D = Σ(n_i/N_i) → fracture when D ≥ 1 — basis of variable-amplitude life prediction); stress concentration (K_t = 3 for infinite plate with circular hole under tension per Peterson + Pilkey; notch sensitivity factor q = 1/(1+a/r) → K_f = 1 + q(K_t−1); typical hotspots on scooters: stem base weld toe, deck-stem joint, folding hinge pivot pin, fork crown — site of the Xiaomi M365 hook failure); folding-lock kinematics (lever-latch hook moment balance F_lock × a = F_rider × b; multi-point hinge load distribution via 3-bar mechanism; twist-and-fold thread engagement ≥ 5 thread pitches per ISO 5855 and Machinery's Handbook; push-button pin shear F_shear = π/4 · d² · τ_y; secondary safety pin as defense-in-depth single-point failure mitigation); steering geometry (headset 36°/45° angular contact bearings; mechanical trail t = R·cosα − r_offset/sinα → 30–80 mm on scooters, ~60 mm on MTBs; wheel flop for low-speed handling); full comparison matrix of 8 safety standards (EN 17128:2020 § 6.4 frame impact 22 kg × 180 mm drop test + § 6.5 frame fatigue 50 000 cycles × 1.3 dynamic factor / ISO 4210-3:2014 bicycle frame+fork 100 000 cycles vertical 1 200 N + horizontal forward 600 N / EN 14781:2005 racing bicycle / ASTM F2641-15 Recreational Powered Scooters ≤ 32 km/h / ASTM F2711-08 Trick Scooters / DIN 79014:2014 City Bike additional German requirements / JIS D 9301:2024 Bicycle Frame Strength / UL 2272:2016 e-mobility structural integrity + battery + electrical); engineering ↔ symptoms diagnostic matrix; 8-point recap.

Engineering deep-dive into the upper rider interface of an electric scooter (handgrip, brake-lever, throttle) — parallel to other engineering-axis articles on [deck and anti-slip surface](@/guide/deck-and-footboard-engineering.md) as the lower rider interface, [brake system](@/guide/brake-system-engineering.md) as the executor of brake-lever commands, and [motor and controller](@/guide/motor-and-controller-engineering.md) as the executor of throttle commands: anatomy of the upper interface (8 components — handlebar tube, handgrip, brake lever, brake cable assembly, throttle housing, Hall-sensor PCB, magnet rotor, connector pigtail); typical form-factor geometry (handgrip dia 28-34 mm, length 120-145 mm, brake-lever reach 60-100 mm, lever pivot-to-pad distance 60-90 mm, throttle travel 25-35° for twist-grip + 8-12 mm for thumb-trigger); 10-row safety standards matrix (EN 17128:2020 § 6.3 controls + § 6.4 handlebar + § 6.5 fatigue, BS EN 14764:2005 § 4.6 brake-system + § 4.10 hand controls, BS EN ISO 4210-5:2014/-8:2014 handlebar/handlebar stem fatigue, ASTM F2641-23 § 7 PMD handles, ASTM F2272 throttle dimensional, ISO 5349-1:2001 hand-arm vibration measurement + ISO 5349-2:2001 workplace application, EU Directive 2002/44/EC physical agents vibration, EN ISO 8662 hand-held power tools vibration, BS 6841/EN ISO 2631 mechanical vibration human exposure, IEC 60068-2 environmental thermal cycling); biomechanics — Chang/Hwang/Moon/Freivalds 2011 optimal grip span study via 2D biomechanical hand model + power grip 30-50 mm cylindrical diameter optimum + sustained grip force 70-100 N intermittent vs 200-300 N peak vs 50-65 N max sustained (Mital/Kumar 1998); HAVS — EU Directive 2002/44/EC daily exposure action value DEAV 2.5 m/s² + daily exposure limit value DELV 5 m/s² over 8-hour A(8) reference period (rms frequency-weighted), Stockholm Workshop scale stages 1V-4V, Raynaud's phenomenon and white finger; materials — grip rubber compounds (TPE Shore A 60-80 vs EPDM Shore A 70 vs silicone Shore A 50-60 vs PVC stretch-fit Shore A 80-90), lever forged Al 6061-T6 σ_y 276 MPa / AZ91D Mg-alloy die-cast σ_y 160 MPa / nylon 6,6+30 % glass-fibre 145 MPa; throttle types (3 — thumb-trigger 8-12 mm travel, twist-grip 25-35° rotation, finger-trigger 5-8 mm); Hall-effect sensor engineering — Honeywell SS49E linear ratiometric 1-1.75 mV/G + Allegro A1324/A1325/A1326 5/3.125/2.5 mV/G factory-programmed sensitivities, 50 % quiescent output, supply 2.7-5 V, current 6-9 mA, temp range -40…+85 °C (SS49E) vs -40…+150 °C (A132x automotive AEC-Q100), bandwidth 10-30 kHz, ratiometric transfer function V_out = (V_cc / 2) + k · B; brake-lever mechanics — lever ratio MA 6:1-8:1 for disc mechanical, modulation curve (linear vs progressive vs digressive), pivot pin friction loss, dual-pull splitter, cable retention barrel-nut; brake cable engineering — inner cable 1×19 stainless 304/316 dia 1.5 mm tensile ≥1700 MPa, housing liner PTFE / nylon, ferrule 6 mm OD, recommended replacement 2-3 years or 5000 km; failure modes — 10-row diagnostic matrix (grip slippage / grip rotation on bar / lever bend after crash / lever pivot rust / cable fray inner-wire / housing kink / barrel-end pull-out / Hall-sensor magnet demagnetisation / Hall-sensor stuck-open ASW failure / throttle housing crack); CPSC recall case studies — Razor Dirt Quad 2008 throttle controller stuck-open 60 reports/2 injuries, Razor Icon 2024 downtube/floorboard separation 7300 units/34 reports/2 injuries; 4-step DIY upper-interface check (grip-twist test, lever-pull span measurement, throttle return-to-zero test, cable tension free-play measurement); 6-step DIY remediation (grip replacement, lever bleeding/pad-gap adjustment, throttle Hall-sensor swap, cable replacement, housing trim/cap install, end-of-life criteria); 8-point recap and conclusion.

Engineering deep-dive into impact physics and certification mechanics for protective gear — parallel to the general regulatory overview in «Safety gear, traffic rules». Linear acceleration vs rotational velocity — HIC15 (NHTSA: 700 = 5 % risk of severe injury, 1000 = original 1972 FMVSS 208 threshold) and BrIC; the trade-off between peak force (kN) and duration (ms) as the central engineering parameter. Full standards comparison matrix: EN 1078:2012+A1 (1.5 m flat / 1.06 m curb, 5.42 m/s, 250 g max, single-impact), NTA 8776:2016 (~150 J, ≈ 6.2 m/s, written specifically for speed pedelecs up to 45 km/h), ASTM F1492 (multi-impact, flat + cylindrical + triangular anvils — a distinct skateboarding discipline), CPSC 16 CFR Part 1203 (2 m flat at 6.2 m/s / 1.2 m curb+hemispheric at 4.85 m/s, 300 g max), DOT FMVSS 218 (5.0–5.4 m/s, 400 g peak), ECE 22.06 (slow ≈ 6.0 m/s allows 180 g / fast ≈ 8.2 m/s allows 275 g), Snell B-95 (lower max acceleration, voluntary premium). Rotational mitigation technologies with physical explanation: MIPS (von Holst + Halldin 1996, 10–15 mm slip plane, up to −50 % rotational acceleration), WaveCel (inverted-V cell crumple, −16–26 % linear + up to 5× rotational reduction vs EPS), KOROYD (welded co-polymer tube structure, mostly linear, often paired with MIPS), SPIN. Virginia Tech STAR rating: 24 impact tests × 6 positions × 2 speeds, biofidelic linear + rotational combination. FOOSH biomechanics: the distal radius = 80 % of the wrist joint surface, Colles (pronation) vs Smith (supination) fracture patterns, Frykman classification; ASTM F2040 wrist guard splint design + prevalence (25 % of bone injuries in children / 18 % in the elderly / 8–15 % in adults). D3O dilatant shear-thickening polymer mechanism (Richard Palmer 1999) and EN 1621-1 Level 1 (≤ 18 kN mean / 24 kN peak — limb protector) vs Level 2 (≤ 9 kN / 12 kN) with a 5 kg striker at 4.47 m/s = 50 J. Back protectors EN 1621-2, eyewear ANSI Z87.1 / EN 166, retention test ECE 10 kg drop 0.75 m max 25 mm displacement. Fit protocol: two-finger above brow, Y-junction strap geometry under the ear, shake test, expiration 3–5 years (CPSC) / 5–10 years (Snell). The engineering source matrix runs parallel to existing applied-physics guides — braking, acceleration, cornering, climbing, descending, emergency maneuvers.

Mirror of the winter-operation guide, only the opposite end of the scale. Four independent scooter subsystems hold the summer temperature budget, and each fails at its own threshold: (1) Li-ion chemistry — calendar aging accelerates exponentially above 30 °C, Battery University BU-808 records up to 35 % capacity loss per year at 40 °C + full SoC; BU-410 and OEM BMS block charging above 45–50 °C; Xiaomi 4 Pro warns at >45 °C, Segway-Ninebot trips a warning at battery ≥55 °C; (2) brakes — organic pads begin to fade at 150–200 °C, glaze from 300–400 °F (≈150–200 °C), rotors warp at 250–300 °C; (3) tyres and hot asphalt — pavement reaches +60–70 °C while air is +35 °C (ScienceDirect, UGA Extension), tyre pressure rises ≈1 psi per 10 °F (Tire Rack); (4) IP protection — IP54/IP66/IP67 are lab-certified, not against UV aging of gaskets plus a summer downpour; FDNY/FSRI 2024–2025: NYC 18 deaths in 2023, 6 in 2024 (NFPA Journal); (5) rider — CDC NIOSH: heat stroke can raise core temperature to 41 °C in 10–15 min, heat exhaustion + dehydration are silent risks; (6) thermal runaway — FSRI experiment: an e-bike engulfs a room in <20 s.

Engineering deep-dive into the systemic environmental-protection layer of an electric scooter — the two-digit IP code per IEC 60529:1989+AMD2:2013 / EN 60529 decodes precisely without marketing interpretation: first digit (0-6) is solid-particle protection with tests IP1X (50 mm object), IP2X (12.5 mm finger probe), IP3X (2.5 mm tool), IP4X (1.0 mm wire), IP5X (dust chamber 2 kg/m³ × 8 h under 20 mbar vacuum), IP6X (full dust-tight); second digit (0-8 plus 9K in ISO 20653) is water protection with tests IPX1 (1 mm/min drip 10 min), IPX2 (3 mm/min drip at 15° tilt), IPX3 (oscillating spray 60° / 10 L/min), IPX4 (splash 360°), IPX5 (jet 6.3 mm nozzle / 12.5 L/min at 2.5-3 m), IPX6 (powerful jet 12.5 mm / 100 L/min), IPX7 (immersion 1 m for 30 min), IPX8 (continuous immersion at manufacturer-declared depth), IPX9K (high-pressure hot water 80 °C / 100 bar / 14-16 L/min per ISO 20653:2013). Why the letter 'X' means 'not tested' rather than 'zero', and why IPX5 is formally 'worse than zero' against dust. Why additional letters A/B/C/D (back-of-hand / finger / tool / wire access) and supplementary H/M/S/W are practically absent on consumer scooters. How sealing is physically built — labyrinth seal (Xiaomi Mi 4 Pro deck cap), gasket-gland design (Parker Hannifin O-Ring Handbook), durometer 50-70 Shore A NBR for maintenance access, 70-90 Shore A FKM for permanent seal. How gasket compounds are selected: NBR (Buna-N) cheapest, oil/fuel-resistant -40…+100 °C; EPDM ozone/UV/water-resistant -50…+150 °C; silicone (VMQ) wide thermal -60…+230 °C but low chemical resistance; FKM (Viton) premium -20…+200 °C with full chemical resistance. Why a scooter controller PCB gets conformal coating per IPC-CC-830C: acrylic (AR) cheap and repairable, urethane (UR) abrasion-resistant, silicone (SR) wide thermal high-flex, parylene (XY) thinnest CVD coating 12-50 μm but non-repairable. Why any sealed enclosure needs a vent membrane: pressure equalization during temperature swing (+50 °C ride → -10 °C overnight) otherwise the gasket gets sucked inward and loses sealing. W.L. Gore PolyVent VE series — PTFE membrane 5 μm pore, water-tight to 1 m head, air-flow 100-1000 ml/min/cm². Model-by-model audit of IP ratings: Xiaomi M365 / Mi 4 Pro / Mi 4 Pro 2nd gen IP54-IP55; Segway-Ninebot Max G30 dual IPX5 body + IPX7 battery; Apollo City Pro IP54 / Apollo Phantom V3 IP56; Dualtron Thunder 3 / Dualtron X II IP55; NAMI Burn-E 2 IPX7; Kaabo Mantis 10 IP54; Inokim OX / OXO IP54. Real-world failure modes — gasket compression set after 1000 insertion cycles plus 12 months UV reduces seal integrity from IP67 to IP54 equivalent; salt-fog corrosion per ASTM B117-19 and IEC 60068-2-11 (5% NaCl mist at 35 °C) — IP-test is fresh water only, sidewalk salt and calcium chloride DOT spray for winter de-icing destroy tin plating and aluminum frame faster than rain. Why EN 17128:2020 nor eKFV nor UK rental trial regulations fix a minimum IP — it is left to manufacturer discretion. Why IP rating is a **delivery-state property**, not a **lifetime guarantee**: degrades linearly with gasket aging (Arrhenius 10 °C rule). 12-step post-rain inspection and replacement schedule.

An engineering deep-dive into the lighting and signaling subsystem of an e-scooter — parallel to the introductory overview at parts/lights-signaling: photometry as a distinct discipline from radiometry (luminous flux Φᵥ in lumens via CIE 1924 V(λ) photopic + 1951 V'(λ) scotopic luminous-efficiency functions; K_m = 683 lm/W peak sensitivity at 555 nm; lumens vs candela vs lux vs cd/m²; Lambertian source I = I_0 · cosθ vs isotropic; inverse-square law E = I / d² for a point source), the headlamp beam pattern (ECE R113 Annex 4 photometric zones — B50L oncoming-glare 0.4 lx max @ 25 m, 75R road-illumination 12 lx min, HV horizon-point 0.7 cd min, vertical test point 50V, cut-off line with 1 % gradient by G = log(E_above / E_below); why asymmetric beam distinguishes the «transmitting» side from the «oncoming» side), LED thermal physics (Rθjc 5–15 K/W chip-to-package + Rθcb 1–5 K/W board + Rθba 10–30 K/W ambient via the electrical-thermal equivalent-circuit model; chromaticity shift Duv at high Tj > 105 °C from phosphor degradation; lumen-maintenance L70/L80/L90 lifetime in hours per IES TM-21-19 extrapolation method with Arrhenius equation k = A · exp(−E_a / kT); chromaticity shift Δuv ≤ 0.007 by TM-21 limit; IES TM-28-22 luminaire-level testing), optical design (TIR total-internal-reflection lenses with polycarbonate n = 1.586 vs PMMA n = 1.491 vs glass n = 1.52; reflector parabolic axis-of-revolution with focal length f; projector lens focal point + shield for cut-off; optical efficiency η_o = Φ_out / Φ_chip = 70–90 % for glass vs 60–80 % for polycarbonate; UV photodegradation via E_UV = hc/λ → polycarbonate ester-bond cleavage over 5–7 years outdoor exposure; chromatic aberration short-wavelength shift), retroreflectivity physics (RA coefficient in cd/(lx·m²) per CIE 54.2-2001 Standard Reflectance Geometry; observation angle α = 0.2° / 0.33° / 1° test values; entrance angle β = ±5° / ±30°; glass-bead n = 1.9–2.1 spherical optics with double refraction + back-reflection vs micro-prismatic full-cube triangular face refraction with theoretical 100 % efficiency; EN 471:2003 + EN ISO 20471:2013 class 2/3 minimum RA 100/500 cd/(lx·m²) for high-visibility apparel; ASTM E810-22 portable retroreflectometer + ASTM E811 hand-held test methods; CIE Photometric Geometry), photometric specifications for signal lamps (SAE J586 stop lamp 80 cd min center / 300 cd max; SAE J588 turn-signal lamp 80–700 cd front / 50–350 cd rear; ECE R7 brake lamp 60 cd min center / 18 cd at ±45°; ECE R6 direction indicator front 175–700 cd / rear 50–500 cd; IEC 60809 flash rate 60–120/min ±5 % deviation per cycle; ramp-up time < 200 ms), audible signaling acoustics (Lp dB(A) with 20 µPa reference; A-weighting curve attenuates < 500 Hz and > 5 kHz, reflecting equal-loudness contours per Fletcher-Munson 1933 + Robinson-Dadson 1956 + ISO 226:2023 equal-loudness contours; EN 17128:2020 § 5.6 minimum 70 dB(A) @ 2 m peak frequency 1–4 kHz; piezo speaker resonant frequency f_r 2.5–4 kHz via RLC equivalent circuit), and a full comparative matrix of 14 standards (IEC 60809:2015 + Amendments / SAE J583 Front Fog Lamp / SAE J586 Stop Lamp / SAE J588 Turn Signal Lamp / ECE R113 Rev 3:2014 Headlamps emitting symmetrical passing beam / ECE R148:2023 consolidated signal lamp / ECE R149:2023 consolidated road illumination / ECE R6 Direction Indicators / ECE R7 Position+Stop+End-outline Lamps / EN 17128:2020 PLEV § 5.5 lights + § 5.6 audible warning / FMVSS 108 49 CFR § 571.108 Lamps, Reflective Devices, and Associated Equipment / StVZO § 67 Germany Bundes-Ministerium für Verkehr / eKFV § 5 German Elektrokleinstfahrzeuge / CIE 54.2-2001 Retroreflection — Definition and Specification of Materials / EN 13356:2001 Visibility accessories); engineering ↔ symptom diagnostic matrix; 8-point recap.

Engineering deep-dive into the e-scooter powertrain — parallel to the introductory overviews «Motors: geared vs direct-drive hub» and «Controller, BMS, display, IoT»: BLDC electromagnetic physics (Lorentz force F=BIL, Faraday EMF ε=-dΦ/dt, Lenz law), KV constant in RPM/V as winding characteristic, torque constant Kt=60/(2π·KV) — why KV 10 on 48 V gives a theoretical 480 RPM/V × 0,95 = 22 N·m/A through mirror symmetry; stator/rotor topology (12-slot 14-pole inrunner vs hub-mount outrunner, NdFeB N42/N48/N52 remanence Br 1.28–1.44 T, ferrite Y30 Br 0.4 T, samarium-cobalt SmCo for high temperatures); three loss types — copper I²R (`P_cu = 3·I²·R_phase`), iron/hysteresis via Steinmetz (`P_h = k_h · f · B^n`, n≈1.6–2.2), eddy currents (`P_e = k_e · f² · B² · t²`); efficiency 85–92 % and why peak efficiency is always near ~50–75 % rated load; thermal management — IEC 60085 insulation class B (130 °C), F (155 °C), H (180 °C), IEC 60529 IP54/65/67 sealing for hub-mounted motors; FOC (Field-Oriented Control) — Clarke transform abc→αβ, Park transform αβ→dq with rotor angle θ, PI controllers for i_d=0 + i_q as torque command, SVPWM (space-vector PWM) modulation; MOSFET inverter — six-MOSFET three-phase bridge, IRFB3077/IPB019N08N3 with RDS(on) 1–5 mΩ, switching losses `0.5·V·I·(t_r+t_f)·f_sw` at 16–32 kHz, dead time 200–500 ns, gate driver 10–15 A peak; DC-link capacitor — ripple current 10–30 A, low-ESR aluminum-electrolytic 1000–2200 μF or polypropylene film; regenerative braking physics — motor as generator, inverter as rectifier, BMS-limited charge acceptance; engineering ↔ symptom diagnostic matrix; full matrix of 9 standards — IEC 60034-1:2022 rotating electrical machines, IEC 60034-30-1 efficiency classes IE1-IE5, UL 1004-1 motors general, UL 1310 Class 2 power units, ISO 21434:2021 road vehicles cybersecurity, IEC 61508 functional safety SIL 1-4, ECE R10 rev 6 EMC + CISPR 14-1, FMVSS 305 high-voltage powertrain, UN ECE R136 L-category propulsion.

76% of US pedestrian and 56% of US bicyclist fatalities happen in darkness, dusk or dawn (NHTSA / FARS), and the Austin Public Health / CDC e-scooter injury study found the typical injured rider is a male aged 18–29 riding on the street at night. This guide moves night risk from the «hope they see me» bucket into the managed-risk bucket: visibility as a **three-component system** (active lights + passive retroreflectors + conspicuous clothing), the physiology of dark adaptation (5–10 min for cones, up to 30 min for full rod adaptation — Webvision NCBI), **biomotion configuration of retroreflectors** (Wood et al., QUT Vision and Everyday Function: retro material on ankles/knees/wrists increases driver recognition distance 3× vs a vest with the same area and 26× vs all-black clothing), the difference between detection and recognition in driver perception, front-light modes by lumens and context (Cycling UK: 50–200 lm for lit streets, 600+ lm for unlit roads, 1000+ for high speed), German StVZO § 67 and UK Highway Code rule 60 as the two regulatory poles, route planning with lit streets vs dark cut-throughs in mind, protocol for losing your front light mid-ride, the alcohol + night risk (PMC: 63% of nighttime riders alcohol-involved vs 22% daytime, 77% head/face injuries with alcohol vs 57% without). ENG-first sources: NHTSA Pedestrian Safety + Bicycle Safety countermeasures, FHWA EDC-7 Nighttime Visibility, Webvision (NCBI), Wood et al. biomotion studies, UK Highway Code rule 60, German StVZO § 67, Cycling UK light guide, PMC e-scooter alcohol/nighttime studies.

Step-by-step roadside protocol after an e-scooter crash: first 60 seconds for self-medical assessment and clearing the carriageway, fixed inspection order for the frame (deck, stem, fork, handlebar) — anchored to the Xiaomi M365 June 2019 recall (10,257 units, serial ranges 21074/00000316–21074/00015107 and 16133/00541209–16133/00544518, stem could fracture from a loose folding-mechanism screw under load), wheel free-spin test and brake-lever verification, **battery after mechanical impact as the central safety pillar** — Battery University BU-304a (mechanical abuse → potential heating, hiss, bulge; modern high-density 3,400 mAh cells with ≤24 µm separator films are more vulnerable than older 1,350 mAh designs); pre-vent signs (solvent smell, visible dent, swelling, popping, localized heat), 24–72 hour delayed thermal-runaway window (NFPA and fire-service monitoring of EV crash scenes 24–48 hours after initial signs), FSRI 2024 e-scooter freeburn test — ignition **13 seconds** after first visible smoke, fireball with 6–7 ft jet flame; folding mechanism and motor-cable routing checks, low-power 50–100 m safe-area test ride, **STOP-conditions** (bent stem, battery dent, brake-fluid loss), **single-impact helmet rule** (CPSC 16 CFR 1203.6(a)(4) warning-label mandate, EN1078:2012+A1 single-impact design, Snell B-95 5-year replacement window; PMC 8735878 — concussion-threshold impacts at 90–100g often leave no visible external damage, hence safer-to-replace policy), insurance claim photo documentation (Velosurance / Markel — 8 mandatory photos plus repair estimate plus written account plus receipts), 24–72 hour delayed checks (battery puffing, hairline frame cracks, brake-fluid contamination), psychological return-to-riding protocol. Sources, ENG-first: CPSC 16 CFR 1203.6(a)(4) via BHSI, PMC 8735878 (Williams et al., bicycle helmet damage visibility study), FSRI 2024–2025 e-scooter freeburn tests, Battery University BU-304a, Velosurance claims process, Xiaomi M365 recall portal + TechCrunch.