Lisbon gave 350 passengers with reduced mobility a ride they can rely on
With city operator Carris, door-to-door transport across 116 km² — where a vehicle that does not arrive is not an inconvenience but a missed appointment.
- Setting
- City-wide, 116 km²
- Service model
- Door-to-door
- Operator
- Carris, Lisbon City Council
- Live since
- 2019, via Smart Open Lisbon

Results
Registered passengers with reduced mobility using the service.
Carried door-to-door by two wheelchair-adapted buses across 116 km².
All of it reached within the first six months of service.
The challenge
For a PRM passenger, a late vehicle is a missed appointment
Passengers with reduced mobility cannot absorb uncertainty the way other riders can. If a vehicle does not show, there is rarely a walk, a different stop or a later service to fall back on — the appointment is simply missed.
The service ran on phone bookings and hand-built schedules. Every change meant a call, and the operator carried the whole burden of getting complex, wheelchair-accessible journeys right.
Carris — 651 buses, 140 million passengers a year — had run dedicated vehicles for this community for years. It wanted higher occupancy and a better channel to passengers, and selected Shotl through Beta-i’s Smart Open Lisbon challenge. With populations ageing, demand for PRM transport is only set to rise — and with it the cost of providing it.

What we deployed
Accessible vehicles, digitised booking, and schedules built by the system
- 01Two adapted buses
Wheelchair-accessible vehicles, drivers and tablets provided by Carris; the service funded by the City Council.
- 02Door-to-door zone
203 virtual stops across 116 km² — placed right outside the door, no construction required.
- 03Booking that includes everyone
App, web and phone — nobody loses access by not owning a smartphone.
- 04Bulk pre-booking
Recurring trips programmed in one go; live booking, change and cancellation for everything else.
Punctuality first, travel time second
For PRM passengers, respecting the pick-up and drop-off time matters more than the shortest trip, so the algorithm was weighted that way. Boarding and alighting take longer and vary by passenger; those times were built into the routing rather than assumed.
Repeat rides — hospital, social services, work — are pre-programmed in bulk. Dispatchers now monitor and review optimised routes instead of building them by hand; passengers track the vehicle in the app instead of phoning to ask where it is.
How the assignment rules workOn site
João VieiraDirector for Strategy and Innovation, Carris“Accessibility is a key issue for us because, as a public transport operator, it is strongly related to our core business, which is to move people effectively across the city.”
“We focus on global solutions rather than very specific ones — by doing this you can still address the needs of people with more acute needs, but also develop products that are easier to use by anyone.”
Similar challenges
Every city with an ageing population has this service to fix
Lisbon proved that digital on-demand tools apply to any form of mobility — including the passengers who most depend on them. It also taught us that PRM boarding times and requirements differ sharply from other riders, and those lessons are now part of the algorithm.
Every city with an ageing population and a phone-and-spreadsheet PRM service faces the same pressure: rising demand, rising cost, and passengers for whom a late vehicle is a missed appointment. The same model — virtual door-to-door stops, punctuality-weighted routing, bulk pre-booking — carries over without new infrastructure.
Demand responsive7–9 minCan Sunyer
SpainA neighbourhood 45 minutes from its own town hall, now waiting 7–9 minutes for a ride to the town next door.
Group commuting27,600 peak weekRabat & Benguerir
MoroccoDaily track-and-board transport plus code-controlled event mobility for two university campuses, on one platform.
Demand responsive−85% waitVallirana
SpainWaits fell from 75–90 minutes to 13, and cost per passenger dropped 29.5%.


