E-scooter data maps wheelchair accessibility

Bolt is exploring how data collected by shared e-scooters could help cities identify poor street surfaces and prioritise accessibility improvements.

Its Wheels4Wheels project uses information from sensors already fitted to scooters to assess conditions affecting wheelchair users and others with limited mobility.

The vehicles use accelerometers, gyroscopes, GPS receivers and cameras for safety, maintenance and product development. Wheels4Wheels processes data from these sensors and assigns each street segment a score on a five-point scale, ranging from smooth and passable to terrain that a wheelchair user cannot cross.

Ardo Reinsalu, Director of Vehicles at Bolt, told Cities Today that the technology could help authorities identify parts of their street networks requiring attention.

“Wheels4Wheels can give cities a more detailed picture of street conditions and help identify areas where poor surfaces may warrant closer inspection or improvement,” he said. “Combined with other data available from shared scooters, such as where falls, skids or other recurring issues happen, this could help cities build a better picture of problematic areas across their street network.”

The information is contributed to OpenStreetMap, where it can support accessible route planning and be used by cities, developers and navigation services.

A pilot in Tallinn, Estonia, generated 40.5 million data points and more than 66,000 surface ratings across eight districts. Volunteers also surveyed accessibility conditions manually, allowing Bolt to compare their findings with the scooter-generated results.

A single scooter journey matched the verified surface data around eight times out of ten. The accuracy increased to approximately nine times out of ten when several scooters travelled along the same street.

“Accuracy is a core part of how we are developing Wheels4Wheels,” said Reinsalu. “During the Tallinn pilot, we compared scooter-derived measurements with verified mapping data and carried out manual surveys alongside the scooter data to validate the methodology.”

He said the system considers GPS accuracy, travelling speed, street coverage and consistency between measurements. Streets with limited scooter traffic will have less information available, but confidence in the rating increases as more journeys are recorded.

The current methodology concentrates on surface quality rather than identifying temporary obstacles in real time. However, a barrier preventing a scooter from passing could also indicate a potential accessibility problem that may not be detected through mapping by car or on foot.

Bolt is now considering whether Wheels4Wheels could be introduced in other locations, although no cities have yet been announced.

“We are currently evaluating how the initiative could be expanded to other cities,” Reinsalu said. “Any future expansion would depend on local conditions and operational considerations.”

The methodology is open source, allowing local authorities, developers and other transport operators to adapt it independently.

Image: Doron Rosendorff | Dreamstime.com