Parts | Days | Selection | Search | Updates | Downloads | Help

SOE: Fachverband Physik sozio-ökonomischer Systeme

SOE 10: Traffic Dynamics, Urban and Regional Systems I

SOE 10.2: Talk

Wednesday, March 29, 2023, 15:15–15:30, ZEU 260

Scaling laws of ridepooling random routes — Philip Marszal, Marc Timme, and •Malte Schröder — Chair for Network Dynamics, Institute of Theoretical Physics & Center for Advancing Electronics Dresden (cfaed), TU Dresden

On-demand ridepooling services are emerging as a new form of urban mobility across the globe. Users request a trip, are assigned to one of the available ridepooling vehicles, and then transported in a shared ride. Despite the complex dynamics of these systems, their efficiency and service quality is characterized by universal macroscopic scaling laws. However, the scaling laws are typically derived from mean-field calculations neglecting the details of the microscopic dynamics of individual vehicles. Here, we analyze these microscopic dynamics in terms of the ridepooling random routes of the vehicles as a new type of random walk process. In contrast to standard random walks, random routes are characterized by a pre-planned route of scheduled stops evolving as new requests are added to the route and users are picked up and delivered. We identify multiple timescales emerging in the routing dynamics and quantify the resulting structure of the random routes in terms of their persistence and diffusion properties. Importantly, microscopic scaling laws of the random route properties directly give rise to macroscopic scaling laws of the overall system efficiency. This perspective of microscopic random routes in ridepooling services may help to better understand their complex dynamics and scaling laws and optimize these services by identifying recurrent route patterns or adjusting allowed stop locations to guide the self-organized routing dynamics.