Control of robotic mobility-on-demand systems: A queueing-theoretical perspective

By Rick Zhang, Marco Pavone. Int. J. Robotics Res., 2014

425 citations recorded by the source corpus.

In this paper we present queueing-theoretical methods for the modeling, analysis, and control of autonomous mobility-on-demand (MOD) systems wherein robotic, self-driving vehicles transport customers within an urban environment and rebalance themselves to ensure acceptable quality of service throughout the network. We first cast an autonomous MOD system within a closed Jackson network model with passenger loss. It is shown that an optimal rebalancing algorithm minimizing the number of (autonomously) rebalancing vehicles while keeping vehicle availabilities balanced throughout the network can be found by solving a linear program. The theoretical insights are used to design a robust, real-time rebalancing algorithm, which is applied to a case study of New York City and implemented on an eight-vehicle mobile robot testbed. The case study of New York shows that the current taxi demand in Manhattan can be met with about 8,000 robotic vehicles (roughly 70% of the size of the current taxi fleet operating in Manhattan). Finally, we extend our queueing-theoretical setup to include congestion effects, and study the impact of autonomously rebalancing vehicles on overall congestion. Using a simple heuristic algorithm, we show that additional congestion due to autonomous rebalancing can be effectively avoided on a road network. Collectively, this paper provides a rigorous approach to the problem of system-wide coordination of autonomously driving vehicles, and provides one of the first characterizations of the sustainability benefits of robotic transportation networks.

Subfields: Automotive Engineering, Control and Systems Engineering, Transportation

View Paper: https://www.semanticscholar.org/paper/6d11a91b9399381365eb952000f498e4ce430ecf

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Automotive Engineering

In Control and Systems Engineering

In Transportation

Citations in the corpus

Platoons of connected vehicles can double throughput in urban roads (2015, methodology): “So we use three models to predict the average queue size and delay of vehicles, and throughput, beginning with the simplest model for an isolated intersection, namely the M/M/1 queue (Soh et al. (2012); Anokye et al. (2013); Zhang and Pavone (2015)).”
Flow: A Modular Learning Framework for Mixed Autonomy Traffic (2017, background): “On a large-scale network, ﬂeets of AVs have been studied for shared-mobility systems, such as autonomous mobility-on-demand [59], [5], [60], which abstracts out the low-level vehicle dynamics and considers a queuing theoretic model.”
Flow: Architecture and Benchmarking for Reinforcement Learning in Traffic Control (2017, background): “a such as stopand-go waves [59]. On a large-scale network, ﬂeets of AVs have recently been explored in the context of shared-mobility systems, such as autonomous mobility-on-demand systems [60], [5], [61], which abstracts out the low-level vehicle dynamics and considers a queuing theoretic model. Lo”
Ridesourcing systems: A framework and review (2019, background): “…and algorithms for empty vehicle routing and repositioning for both taxi and ridesourcing systems has arisen; see Braverman et al. (2016), Zhang and Pavone (2016), Wallar et al. (2018), Iglesias et al. (2018), Iglesias et al. (2019) , and Yu et al. (2019b) ; machine learning techniques,…”
Autonomous cars: Past, present and future a review of the developments in the last century, the present scenario and the expected future of autonomous vehicle technology (2015, background): “It is being successfully tested at various universities across Switzerland, England and Singapore (Zhang, 2014).”
Planning and Decision-Making for Autonomous Vehicles (2018, background): “Much of the fleet management literature for mobility-on-demand systems considers the case of ride sharing without pooling requests, focusing on fluid approximations (136), queuing-based formulations (137), and case studies in specific regions.”
On-demand high-capacity ride-sharing via dynamic trip-vehicle assignment (2017, background): “Muchof the ﬂeet management literature for MoD systems considers the case of ride-sharing without pooling requests, focusing on ﬂuid approximations (6), queuing based formulations (7), case studies in speciﬁc regions [e.g., Singapore (8)], and operational considerations for ﬂeet managers (9).”