What is the minimum number of vehicles needed to serve
all the trips in New
York without delaying passengers’ pick up times?
The Minimum Fleet Network model developed by the Senseable City Lab could reduce the taxi fleet size by 40%.
Currently, active taxis are circling to look for someone to pick up when they are not driving a passenger to destination. While each driver seeks to reduce this empty time individually, taxis still travel without serving passengers for about 40% of their time. In the Minimum Fleet Network model, this circling is substantially reduced, and each active taxi operates in one of two possible modes:
Current taxi situation
Minimum Fleet Network model
By relying on the algorithm to identify passengers,
taxis optimize their
work hours and reduce empty travel time.
This way, taxi fleet size can be minimized
Working hour optimization
Monday Tuesday Wednesday Thursday Friday Saturday Sunday
Types of car modes
Cumulated time with and without passenger in a typical working day
▬ ▬ With passenger
▬ Without passenger
The Minimum Fleet Network model proposes efficient vehicle-trip assignments that reduce the empty time between trips with a passenger onboard. This way, a taxi can serve more passengers in the same time, and overall fleet size is minimized.
Currently, taxi operation is mostly demanded to individual driver’s decisions about where to roam once a passenger is dropped.
1 hour
● taxi trips
With coordinated operation, assignment of taxi to passengers is optimized and empty time reduced. Passengers are guaranteed to have a taxi at the desired location and time, and fleet size is minimized.
1 hour
● fleet optimization trips
By minimizing empty travel times, the Minimum Fleet Network model reduces the number of cabs needed to serve NY of about 40% compared to current operation.
Total fleet on the road
▬ Current taxi situation
▬ Minimum Fleet Network model
Activity of the fleet in the model
■ With passenger
■ Driving to pick up
■ Waiting to pick up