New York City could make do with a taxi fleet as much as 40 per cent smaller than the one currently used, reports a paper published this week in Nature. This finding could become even more relevant in the years ahead, as fleets of networked, self-driving cars become commonplace.

Information and communication technologies have enabled new solutions for urban mobility, including on-demand vehicles. However, a fundamental unsolved problem is how best to size and operate a fleet of vehicles without causing delays to the passengers or requiring them to share taxis.

Mohammad Vazifeh and colleagues developed a network-based approach for analysing this issue, in which individual vehicles (not individual rides) are shared across the transport network. The authors test their scheme against the 150 million taxi trips taken in New York City over the course of one year and find that the city could indeed reduce cab use with prior knowledge of the trips to be taken. They calculate that the same service could have been provided with a fleet 40 per cent smaller than the one currently used. When the scheme is modified to work with real-time, rather than historical information, the authors find that to serve more than 90 per cent of trips, a reduction in fleet size of around 30 per cent is still possible.