How scooter sharing actually works

From a rider’s side, scooter sharing is four actions: find, unlock, ride, park. Underneath each one is a piece of machinery that most riders never see. Here is what is happening.

The scooter is a connected device

A shared scooter is not a rental object with a padlock. It carries a controller with a mobile data connection and a GPS receiver. It reports where it is and what state it is in, and it listens for commands.

That connection is the whole basis of the service. Without it, an operator cannot tell you where a scooter is, cannot unlock it remotely, cannot stop it being ridden somewhere it should not be, and cannot tell whether it has been knocked over.

Finding a scooter

The map in the app is built from position reports coming off the vehicles. A good app shows more than dots: it draws the riding area, the parking areas and any restricted zones on the same map, so that you can see before you start a ride where you will be allowed to end it.

Unlocking

Scanning the code identifies the specific vehicle. The app asks the server to start a ride; the server sends an unlock command to that scooter. Then, and this part matters, it waits.

A well-built system does not start charging you the moment you scan. It starts when the scooter confirms it has actually unlocked. The difference shows up on the day a vehicle has no signal or a flat battery: with confirmation, nothing happens and you walk to another one. Without it, you have been charged for a scooter you never rode.

Riding, and the rules of the street

This is where operators differ most. Cities need certain streets to be off limits, certain areas to be slower, and certain places to be unavailable for parking. There are two ways to implement that.

The weak way is to show the rider a warning in the app. The rider may be looking at the road, may have the phone in a pocket, and may simply ignore it. The rule exists as a notification.

The strong way is to apply the rule to the vehicle. A speed limit is pushed to the scooter’s controller, which enforces it whether or not the app is even open. A no-ride zone stops the throttle delivering power at the boundary, so the scooter coasts to a stop and cannot accelerate again until it leaves.

The test of a geofence is simple: does anything happen if the rider ignores it?

Ending the ride

Parking is where shared scooters earn or lose their welcome. Systems that let a ride end anywhere produce the scattered vehicles that cities complain about. Systems that refuse to end a ride outside a permitted area do not, because ending the ride is the thing the rider wants.

The refusal has to be helpful rather than just obstructive: tell the rider where the nearest permitted area is, and keep the ride open until they reach it.

Paying for it

Most services charge an unlock fee and then a rate per minute. Topping that balance up should take seconds: choose an amount, confirm with a saved card, and ride. Payment is the part of the experience a rider should never have to think about mid-journey.

The part nobody sees

Behind the app is an operation that is mostly physical:

• Charging: either swapping batteries in the field or collecting vehicles to a depot.
• Redistribution: moving scooters from where rides end to where rides start.
• Maintenance: brakes, tyres, stems and electronics, on a schedule and on report.
• Recovery: retrieving vehicles that end up somewhere they cannot be ridden out of.
• Support: riders with a problem, and residents with a complaint.

This is the real business. The software makes it possible; the operation makes it work. Any assessment of a scooter company that looks only at the app is looking at the smaller half.