Hub motors are the defining engineering choice for bolt-on electric skateboard drives such as the Mellow system: torque originates inside the wheel rather than at a central gearbox linked by belts. Academic literature rarely studies longboards directly, yet a 2022 EVS35 paper by Wonders and colleagues compares central-drive and in-wheel architectures on a modelled city bus, quantifying energy demand differences large enough to inform how we think about drivetrain losses at any scale. Our dedicated in-wheel motors page walks through product history; this essay translates peer-reviewed bus modelling into careful skateboard context without claiming identical percentages on a four-kilogram deck.
What the EVS35 study compared
The authors presented at the 35th International Electric Vehicle Symposium (11 June 2022), modelling a representative urban bus duty cycle rather than a consumer board. Their question was structural: given the same battery and route profile, how much energy does a classical central drive with mechanical transmission consume relative to in-wheel motors that eliminate long shafts and gear pairs? The paper isolates components—motors, inverters, tyres—so readers can see where joules disappear as heat instead of propelling mass. That decomposition is what makes the work relevant to skateboard hub drives even though mass, speed, and aerodynamics differ by orders of magnitude.
City buses and electric skateboards share the abstract problem of stop-start friction and the need to minimise unsprung weight, but they do not share duty cycles. A bus may haul dozens of passengers for sixteen hours a day; a Mellow rider might draw high peak current for hill starts then coast. The EVS35 contribution is therefore conceptual: in-wheel layouts remove mechanical paths that inherently dissipate energy. Translating bus percentages to board range requires humility. We cite the eight-to-eleven per cent band as bus-model output, not as a measured Mellow range boost.
Peer review at a symposium proceedings level still beats marketing superlatives. When crowdfunded hardware teams promise impossible efficiency, a bus study gives engineers a sanity check on whether hub integration is merely aesthetic. Archive readers evaluating belt drives versus hubs should weigh maintenance, noise, and packaging alongside the thermodynamic argument Wonders et al. articulate for large vehicles.
Mechanical losses the central drive pays
Central drives route power through differentials, shafts, and reduction gears before torque reaches the road. Each interface exhibits friction, lubricant churning, and alignment sensitivity. In-wheel motors mount torque at the rim, shortening the path electrons travel before becoming rotation. The EVS35 modelling attributes part of the in-wheel advantage to shedding those mechanical branches, though it still accounts for inverter losses and tyre rolling resistance common to both layouts. Skateboard belt drives replicate a miniature central-drive story: pulley wrap, belt tension, and bearing drag consume watts riders feel as warmth after a long session.
On a bus, mechanical losses scale with torque and operating hours, so percentage savings compound into meaningful kilowatt-hours per day. On a board, absolute watt savings may be small yet still influence heat in the motor can and the duration before thermal limits cap output. Hub designs also free deck space for batteries—packaging interacts with range as much as efficiency. The paper does not discuss battery placement, reminding us not to merge distinct engineering axes when reading one study.
Maintenance culture differs: transit agencies log failures; riders swap tyres at home. Hub motors complicate tyre changes but remove belt alignment checks. Those trade-offs do not appear in the EVS35 energy tally yet matter for ownership cost. Connecting research to product experience keeps the archive honest: cite the numbers the authors publish, narrate the skateboard parallels cautiously.
The eight-to-eleven per cent energy-demand gap
Across the scenarios Wonders et al. model, in-wheel configurations show roughly eight to eleven per cent lower energy demand than the comparable central-drive bus. The interval reflects duty-cycle assumptions and component efficiency maps, not a single universal constant. Readers should treat it as an order-of-magnitude illustration that drivetrain topology matters, not as a promise that swapping a board's architecture automatically adds ten per cent range. Aerodynamic drag dominates at high road speeds; at skateboard velocities, rolling resistance and repeated acceleration pulses weigh more heavily, potentially changing which loss channel leads.
Even if a board captured only a fraction of bus-scale savings, the direction of travel supports hub integration for efficiency-minded designs. Regenerative braking, when implemented, returns energy through the same inverters the study models—central versus in-wheel comparisons may shift slightly depending on control algorithms. Mellow's historical emphasis on a self-contained drive module aligns with the engineering logic the paper quantifies, independent of any claim that the company validated bus percentages on Munich streets.
Policy makers reading EVS35 may focus on fleet procurement; riders might notice smoother acceleration when inertia sits outboard. Both perspectives hinge on the same physics. When discussing efficiency in forums, link to primary papers instead of recycling vague "hubs are greener" slogans. The DOI trail through Citelec's EVS35 repository preserves the PDF for anyone auditing our summary.
Lessons for electric skateboard design debates
First, efficiency arguments belong beside thermal and reliability data—an eight per cent bus saving does not excuse inadequate cooling on a compact hub. Second, scale matters: unsprung mass penalties that worry bus chassis designers appear differently under a rider's feet. Third, legislative categories ignore drivetrain theology entirely; German light-electric-vehicle rules care about handlebars and watts, not whether torque is centralised. Cross-linking motor research with history shows how Mellow positioned hubs as premium engineering while regulators still debated whether boards were toys or transport.
Future studies may model electric kick scooters or cargo bikes with similar rigour. Until then, EVS35 remains a respectable anchor for in-wheel advocacy. Teachers in STEM outreach can use the bus narrative to explain why Mellow chose integrated wheel motors without claiming transit-scale statistics for leisure products. Intellectual honesty preserves trust more than overstated range figures ever could.
If you experiment with custom firmware or aftermarket wheels, remember the paper's boundary conditions: tyre rolling resistance parameters swamped small drivetrain gains in some bus runs. Your choice of urethane may dwarf motor topology in real-world skateboard range tests. Empirical logging on your own routes still beats any single academic model, yet models clarify which experiments are worth running. That interplay of lab and pavement is where the archive intends to sit—between marketing decks and unread PDFs.