Determining the minimum urban fleet for a valet style autonomous mobility service using real trip data

Area of Science:

• Urban planning and transportation systems analysis.
• Autonomous vehicle deployment strategies.
• Sustainable mobility solutions.

Background:

• Over 1.5 billion private cars worldwide contribute to urban congestion, land use, and emissions.
• Fully driverless robotaxi fleets face regulatory, cost, and technological hurdles for large-scale European deployment.
• An intermediate solution is needed to bridge private car ownership and fully autonomous mobility.

Purpose of the Study:

• To evaluate the feasibility of a valet-style Autonomous Mobility on Demand (AMoD) service.
• To quantify the required fleet size for such a service under specific operational constraints.
• To assess its potential to reduce private car dependency and optimize urban space.

Main Methods:

• Development of a dynamic simulator integrating vehicle assignment, rebalancing, and charging strategies.
• Replay of one week of real private car trip data from 28,369 vehicles in Milan.
• Simulation of a valet AMoD service with quality of service constraints (20-minute maximum wait time, <2% unmet requests).

Main Results:

• Only 2,300 shared autonomous vehicles are needed, a twelve-fold reduction compared to the current private fleet.
• This represents a significant efficiency gain over conventional free-floating car sharing, which would require 7,750 vehicles.
• The model demonstrates feasibility within an urban context, incorporating real trip data and operational heuristics.

Conclusions:

• Valet AMoD offers a practical intermediate step towards fully autonomous mobility.
• It can enable households to reduce car ownership, free urban curb space, and lower emissions.
• The service preserves the individual driving experience while paving the way for future autonomous transportation.