Data-Driven Optimization for E-Scooter System Design

Jianqiang Cheng, University of Arizona


The objective of this project is to develop data-driven, decision-making models and computational methods for shared-mobility system design and operation. Specifically, we will use shared e-scooters as a representative system, with the ultimate goal of facilitating an electric shared-mobility revolution that promises a more sustainable future.
In the past few years, shared e-scooter systems have gained increased popularity around the world because of their benefits to health, traffic congestion, the environment, and accessibility. As of 2018, approximately 100 U.S. cities have launched shared e-scooter programs, accounting for 38.5 million trips. However, the business model to manage e-scooter sharing remains nascent, with many challenges still poorly addressed and outstanding. As a result, we propose to solve several urgent questions that arise at the company and policy-maker levels for e-scooter sharing (e.g., planning, operations), by: (i) developing a data-driven robust optimization model to provide the decision makers with a robust solution that enables low cost and high service quality, and explicitly captures endogenous uncertainty in demand in the case of limited demand information; and (ii) designing computationally efficient methods with solution quality guarantees to solve the e-scooter sharing system design and operation problems. In line with the NITC themes, these research results have the potential to provide e-scooter companies with new decision-making tools and methodologies to effectively design and operate shared e-scooter systems, and thus help to ensure system reliability and cost-effectiveness.

Project Details

Project Type:
Project Status:
In Progress
End Date:
January 31,2022
UTC Grant Cycle:
NITC 16 Round 4
UTC Funding: