ロボタクシーを都市交通に統合するための課題を分析 (How cities can make room for robotaxis)

2026-10-30 スイス連邦工科大学ローザンヌ校（EPFL）

スイス連邦工科大学ローザンヌ校（EPFL）の研究チームは、自動運転タクシー（ロボタクシー）を都市交通に導入する際、交通渋滞を増加させず、公共交通機関との共存を実現するための車両運用・配置最適化手法を研究した。研究では、人工知能と強化学習を活用し、乗客需要の予測に基づいて多数の自動運転車両を都市内に再配置する車両管理アルゴリズムを開発。中国・深圳市の実データを用いた分析では、車両の待機場所や配車を集中管理することで、利用者の待ち時間短縮、車両稼働率の向上、郊外や交通サービスが不足する地域への移動手段の提供が期待できることを示した。一方、無乗客車両の無用な走行による交通量増加を防ぐため、空車走行の制限、課金制度、専用待機場所の整備などの政策的対応も提案した。また、ロボタクシーが既存の公共交通を代替するのではなく、鉄道やバスの利用圏を補完する交通体系の構築が重要であると指摘した。本研究は、自動運転技術の実用化に向け、AIによる車両運用最適化と都市交通政策を統合するアプローチを提示している。

＜関連情報＞

サービスアクセス保証を備えたスケーラブルなライドソーシング車両再配分：制約付き平均場強化学習アプローチ Scalable ride-sourcing vehicle rebalancing with service accessibility guarantee: A constrained mean-field reinforcement learning approach

Matej Jusup,Kenan Zhang,Zhiyuan Hu,Barna Pásztor,Andreas Krause, Francesco Corman
Transportation Research Part C: Emerging Technologies Available online: 21 April 2026
DOI:https://doi.org/10.1016/j.trc.2026.105705

Highlights

Mean-field RL rebalances thousands of vehicles within a fraction of a second.
Guaranteed equitable service access by integrating accessibility constraints.
Model-based RL achieves higher sample efficiency than the previous studies.
Matching process learned or approximated by the optimal transport.
Evaluations on real-network simulations.

Abstract

The rapid expansion of ride-sourcing services such as Uber, Lyft, and Didi Chuxing has fundamentally reshaped urban transportation by offering flexible, on-demand mobility via mobile applications. Despite their convenience, these platforms confront significant operational challenges, particularly vehicle rebalancing—the strategic repositioning of a large group of vehicles to address spatiotemporal mismatches in supply and demand. Inadequate rebalancing not only results in prolonged rider waiting times and inefficient vehicle utilization but also leads to fairness issues, such as the inequitable distribution of service quality and disparities in driver income. To tackle these complexities, we introduce continuous-state mean-field control (MFC) and mean-field reinforcement learning (MFRL) models that employ continuous vehicle repositioning actions. MFC and MFRL offer scalable solutions by modeling each vehicle’s behavior through interaction with the vehicle distribution, rather than with individual vehicles. This mitigates the curse of dimensionality with respect to the number of agents, enabling coordination across large fleets with significantly reduced computational complexity and eliminating the need to retrain the model when fleet size changes. To ensure equitable service access across geographic regions, we integrate an accessibility constraint into both models and derive rebalancing policies that strike a balance between high fulfillment of passenger demand and fair coverage of vehicle supply. Extensive empirical evaluation using real-world data-driven simulation of Shenzhen demonstrates the real-time efficiency and robustness of our approach. Remarkably, it scales to tens of thousands of vehicles, with training times comparable to the decision time of a linear programming rebalancing. Besides, policies generated by our approach effectively explore the efficiency-equity Pareto front, outperforming conventional benchmarks across key metrics like fleet utilization, fulfilled requests, and pickup distance, while ensuring equitable service access.

共有型自動運転車システムにおける車両管理および駐車管理に影響を与える要因の探求：エージェントベースシミュレーションフレームワーク Exploring influential factors of fleet and parking management in shared autonomous vehicle systems: An agent-based simulation framework

Yuqian Lin,Kenan Zhang,Daniel Kondor, Zhan Zhao, Carlo Ratti, Yang Xu
Transportation Research Part A: Policy and Practice
DOI:https://doi.org/10.1016/j.tra.2025.104762

Abstract

Shared autonomous vehicles (SAVs) are expected to enhance urban transportation efficiency through innovative mobility resource management. By developing a comprehensive agent-based simulation framework, this study investigates several key factors influencing fleet size and parking demand for the adoption of SAVs in future urban mobility systems. The framework evaluates how both operational factors (e.g., reservation time and maximum waiting time) and demand-side characteristics (e.g., demand rate and the balance between trip origins and destinations) jointly affect the performance of the SAV system. It uses a two-stage simulation process that includes capacity estimation and performance evaluation. In the initial warm-up stage, the simulation estimates the fleet size and parking spaces needed to serve specific travel demand. These initial estimates are then used in the second stage to run further simulations and assess additional performance indicators, including final required parking spaces, empty meters traveled, and trip rejection rate. To obtain a holistic understanding of the studied factors, we construct various simulation scenarios based on historical taxi data in central areas of Chengdu, Shanghai (China), and Manhattan of New York City (USA), and build structural regression models on the simulation outcomes. The results reveal a general mechanism by which operational characteristics and demand patterns influence SAV fleet and parking sizes. We find that a 1 % increase in overall travel demand results in about a 1 % increase in the number of SAVs needed and required parking spaces. Meanwhile, a 1 % improvement in the balance of origin-destination (OD) trips, which reduces spatial mismatches between vehicle supply and trip requests, can help offset the need for additional vehicles and parking spaces. These findings offer critical policy implications, emphasizing the need for integrating SAV deployment with land-use strategies, balancing fleet investment, environmental costs, and service quality (e.g., lower waiting time) in SAV planning and operations.

月	火	水	木	金	土	日
						1
2	3	4	5	6	7	8
9	10	11	12	13	14	15
16	17	18	19	20	21	22
23	24	25	26	27	28	29
30