Most shared-scooter deployments are decided before a single vehicle is ordered. The operator signs a permit that fixes a maximum speed, a docking rule and a service obligation, then goes looking for a scooter that satisfies all three. When the vehicle catalogue and the permit disagree, the difference is paid for in retrofits, grounded units and, in a few documented cases, a suspended operating licence.
This guide is about that translation step: taking a city permit and a depot layout, and turning them into a vehicle specification a factory can quote against. It does not model revenue per ride. That arithmetic depends on fares, trip density and season in a way no manufacturer can supply honestly. What a manufacturer can supply is the four vehicle numbers that decide whether a fleet can meet its permit, and the eight-line build spec that produces them consistently across a thousand units.
The platform discussed here is the sharing-fleet configuration described on our shared mobility fleet page: a 500W sealed hub motor, a swappable 48V pack, solid or puncture-resistant tyres, tamper-proof fasteners and a welded IoT module. The fleet procurement process itself, including permit mapping and the three build tiers, is covered in the shared mobility fleet procurement guide, and this article assumes you have already chosen a tier.
Start With the Permit, Not the Catalogue
Permits for shared light electric vehicles converge on the same short list of technical conditions. The exact thresholds vary by city, but the categories do not: a governed maximum speed, an audible warning device, front and rear lighting visible at a defined distance, geofenced speed reduction in designated zones, and a vehicle identification method that the operator can demonstrate to an inspector.
Two of those conditions are firmware problems, not hardware problems. A 20 km/h governed ceiling and a school-zone geofence are both implemented in the motor controller, which means the vehicle you buy must have a controller that accepts a speed map rather than a single fixed limit. Our electric scooter OEM guide covers how that configuration travels through an order. The practical procurement consequence is blunt: if the factory cannot document how the speed ceiling is set and how it is locked against rider modification, the vehicle cannot be permitted in most regulated markets regardless of how good the hardware is.
The third condition, lighting, is easier to verify and harder to get right. A front light and a rear light that meet a visibility distance at specification are not the same as lights that still meet it after a season of curb strikes and connector corrosion. Ask for the lighting circuit to be on its own fused path with a sealed connector, not spliced into the main harness.
The Four Numbers That Decide Whether a Fleet Can Comply
Beyond the permit line items, four vehicle numbers determine whether a fleet can be operated at all. They are worth pinning down in writing before price is discussed, because every one of them is expensive to change after tooling.
Real range, not catalogue range
Catalogue range is measured on a flat course in economy mode with a nominal rider. A shared vehicle in a dense zone does the opposite of that: repeated short trips, frequent stops, a rider plus a load, and cold mornings. The city-scooter platform is rated 20–40 km depending on pack, from 36V 10Ah up to 48V 10Ah, measured in ECO mode with a 75 kg rider. Plan a sharing deployment on the low end of the band and treat the high end as a best case, then size the swap or dock cycle against the low number.
Payload and rider weight spread
Sharing vehicles carry a wider weight spread than commuter vehicles, and the load capacity figure matters for frame and tyre selection. The city-scooter platform is rated 100–120 kg. A deployment that attracts two-adult trips on a single deck, which is common on leisure routes, needs the upper figure and a deck that is rated for it, not just a frame that survives it.
Charge time versus operating window
Charge time is 4–6 hours on the standard 2A charger for the city-scooter packs, depending on capacity. That single number decides the whole charging architecture. If a vehicle works a daytime window and can be docked overnight, 4–6 hours is comfortable and dock charging needs no special provision. If the vehicle is expected to work both a morning and an evening peak, 4–6 hours does not fit inside the gap, and the fleet either needs a second pack per vehicle or a swap process. This is the arithmetic that sets fleet size far more often than range does.
Service interval versus tyre and brake wear
Solid and honeycomb tyres are standard on sharing builds for one reason: a flat is not a repair, it is a service visit that removes a vehicle from the street for hours. The same logic applies to brakes. A rear disc plus front electronic brake is the city-scooter configuration; on a sharing build the brake wear items should be specified as field-replaceable without removing the wheel.
Dock, Swap, or Both: What Each One Costs in Vehicles
The charging model is usually presented as an operating-cost question. In practice it is a fleet-size question, and the two answers differ by a large margin at the same service level.
| Model | Vehicle time lost per day | What it demands of the vehicle | What it demands of the site |
|---|---|---|---|
| Dock charging | 4–6 h, inside the operating window unless overnight | Sealed charge port, captive cable retention, charge-state reporting | Power, weatherproofing and cable management at every dock position |
| Pack swap | Under a minute, at the depot | One-person removable pack with a captive slide-in connector, pack-level ID | A charging cabinet for spare packs, plus transport for them |
| Hybrid | None if spare packs cover both peaks | Both of the above on the same vehicle | Both of the above |
The trade is straightforward once it is written down. Dock charging buys simplicity at the site and pays for it in vehicle count, because every vehicle is unavailable for the length of its charge. Swap buys vehicle utilisation and pays for it in packs, cabinets and logistics. A fleet that runs one operating window per day is almost always better served by docks. A fleet that runs two is almost never served by docks alone.
The pack itself is where swap fleets fail quietly. A pack that one person cannot remove and refit in under a minute, in the dark, wearing gloves, is not a swappable pack. The practical swappable size on this platform class is the 48V 15Ah pack used on the Core tier, with a slide-in rail and a connector that seats without alignment. Our battery pack guide covers cell grade, BMS tiers and cycle life in more depth than a fleet specification needs, but the cycle-count question is worth importing: a pack rated for 800 or more cycles is what makes the swap model hold up over a multi-year service life.
Telematics Is a Vehicle Component, Not an Accessory
The most common mistake in a sharing specification is treating the IoT module as something fitted after delivery. It cannot be. A module bolted to a bracket is a module that leaves with the first determined thief, and a tracker that can be unplugged without consequence is worse than no tracker, because it produces a false sense of coverage.
The sharing-fleet configuration addresses this in hardware: the IoT module is welded on, and the motor is disabled when the tracker is removed. Specify that behaviour explicitly in the RFQ, because it is a controller function that has to be built in, not configured later. The module itself should provide 4G positioning at roughly 2.5 m accuracy with BLE unlock, and the firmware should support over-the-air updates, since every fleet has at least one controller fix in its first year.
Tamper resistance on the rest of the vehicle follows the same principle. Torx security fasteners throughout, pin-lock axles, and no exposed cable runs that can be cut to disable lighting or the lock. Exposed cabling is the single cheapest way to make a vehicle worth stealing for parts.
What a Thousand-Unit Order Actually Changes
Unit economics at scale are an operator's problem, not a factory's. What does change at a thousand units is process control, and that is where a fleet buyer should spend their diligence.
- Consistency. At a thousand units, a one-percent variance in anything is ten vehicles. Weld quality, torque values and connector seating have to be controlled by fixture and documented per unit, not by operator judgement.
- Spare parts as a deliverable. Ask for a module-level spare parts price list with the RFQ. If the deck, pack, controller and motor are individually replaceable, an incident costs a module, not a vehicle. If the factory cannot produce that list, the vehicle was not designed for fleet service.
- Firmware control. The speed ceiling, geofence behaviour and lock logic must be settable per market and lockable against rider changes, with the settings documented in writing for the permit file.
- Certification as a package, not a claim. EN 17128 for the vehicle, UL 2272 for the electrical system, UN 38.3 for the pack, IP67 for electronics and ISO 4210 for the frame are the standard set on the sharing-fleet page. Each should arrive as a document with the shipment, matched to the serial range in the order.
- Lead time honesty. Standard fleet models run 15–30 days after deposit. A build with custom branding, a different pack or a market-specific controller map runs 30–45 days. Neither number includes sea freight, and a permit deadline is not a reason for a factory to compress a first-article check.
The order quantities and price bands for the three sharing tiers are laid out on the shared mobility fleet page, with MOQ at 50 units on the base city-scooter platform. Custom branding and colour are where the MOQ moves, and that is worth confirming before announcing a launch date.
Where the Deployment Plan Usually Breaks
Three failure modes account for most sharing deployments that miss their launch window, and none of them are vehicle faults.
The dock count was sized on range. Range tells you how far a vehicle can go. It says nothing about how many vehicles are needed to keep a given number on the street while the rest charge. Size the fleet on charge time and operating windows first, then check that range covers the longest trip in the zone. Doing it in the other order produces a fleet that looks adequate on paper and runs short at midday.
The permit was signed before the vehicle was specified. A speed map, a geofence and an identification method each require a specific controller capability. Signing a permit that commits to a behaviour the chosen platform cannot implement leaves the operator retrofitting a hardware function in software, which is not possible. Read the permit against the controller documentation before ordering.
Spares were treated as a later purchase. A fleet with no spare decks, packs and controllers has a service level that degrades from day one. The correct quantity is a judgement call, but the correct time to negotiate it is in the same RFQ that sets the unit price, because a spare parts commitment negotiated later has no commercial leverage behind it.
If you are also running cargo or delivery vehicles alongside a passenger fleet, the charging infrastructure question is shared between them and is covered in the charging dock and station guide. For maintenance planning across a mixed fleet, the maintenance guide sets out service intervals in the same terms a sharing operator needs. And if the deployment involves a rider-facing safety programme, the safety features guide covers the hardware that a permit inspection will look for.
Building the Specification
What follows is the same eight-line build spec that any factory quote can be checked against. It is deliberately short, because a specification that runs to several pages does not survive contact with a purchasing department.
| Line | Specification | Why it is in the spec |
|---|---|---|
| Frame and deck | 6061 aluminium, reinforced stem weld, 6-point reinforced welds on the sharing build, internal cable routing | Deck impacts and rider weight spread; internal routing removes the easiest vandalism target |
| Motor | 500W sealed hub, single motor, aluminium housing | Single motor halves powertrain failure exposure; 500W is the sharing sweet spot |
| Pack | 48V 15Ah swappable, slide-in rail, captive connector, pack-level ID, 800+ cycles | Makes the swap model physically possible and lets the operator track pack health by serial |
| Tyres | Solid or puncture-resistant, rated for high mileage | A flat is a vehicle off the street, not a roadside fix |
| Brakes | Rear disc or drum plus front electronic brake, field-replaceable wear items | Weather and heavy use tolerance, plus a service cost that does not require wheel removal |
| Electronics protection | IP67 for electronics, sealed connectors on the lighting circuit | Water ingress is the leading premature failure mode on shared vehicles |
| Telematics | Welded IoT module, 4G with BLE unlock, motor cut on removal, OTA firmware | Makes the tracker part of the vehicle rather than a removable accessory |
| Compliance | EN 17128 / UL 2272 / UN 38.3 / ISO 4210, documents matched to serial range | A permit file needs documents, not a claim on a catalogue page |
Order against that spec and the rest of the deployment plan has a fixed foundation. Order against a price and the specification drifts toward whatever the factory had in stock, which is the point at which the permit, the depot and the service contract all start to disagree with the vehicle.
Get a Sharing Fleet Build Spec
Send us the permit conditions and the depot layout. We will map them onto the vehicle specification, confirm which of the three sharing tiers fits, and return the build spec with certification documents and a module-level spare parts list for review.
