Dr. Sheng Liu, Assistant Professor of Operations Management and Statistics at the Rotman School of Management, presented “Data-Driven Approaches in Smart City Operations” on February 19, 2021. In this UT-ITE seminar, Liu discussed two projects around data-driven solutions for city operations planning and management.
Food delivery operations
The first project presented by Liu was on emerging food delivery operations. Liu worked with a major food delivery service provider in China to develop a data-driven framework for optimizing practical order assignment in delivery services. In addition, a machine learning approach predicted travel time with drivers’ routing behaviours while accounting for uncertain and heterogenous service times.
The solution technique used Branch-and-Price optimization and showed improved performance compared to Vehicle Routing Problem (VRP)-based models.
Bike lane network design
Liu’s second project used real-world bike trajectory data for bike lane network design. Bike-sharing systems are increasingly popular, exposing challenges with bike lane networks and prompting bike lane initiatives.
Liu created a general framework for urban bike lane planning using real-world bike trajectory data. This framework has two main objectives:
- Coverage. Bike lanes should cover as many cyclists and rides as possible.
- Continuity. Bike lanes should be connected to each other for better riding experience and improved safety.
This optimization used general cyclist utility functions and efficient algorithms by exploiting the problem structure.
Resources
- Liu, Sheng and He, Long and Shen, Zuo-Jun Max, On-Time Last Mile Delivery: Order Assignment with Travel Time Predictors (May 15, 2018). Forthcoming in Management Science.
- Liu, Sheng and Shen, Zuo-Jun Max and Ji, Xiang, Urban Bike Lane Planning with Bike Trajectories: Models, Algorithms, and a Real-World Case Study (December 1, 2018).
Abstract
This talk presents two projects that build data-driven solutions for city operations planning and management.
The first part is devoted to the emerging food delivery operations. Working with a major food delivery service provider in China, we develop a data-driven optimization framework to minimize customers’ expected delivery delay. To capture the driver’s routing behaviors, we propose a machine learning approach that predicts travel time with covariate information. Combined with the travel time prediction, our optimization framework is robust and yields significant reductions in delay times.
In the second part, we present how the real-world bike trajectory data can be leveraged for bike lane network design.
We integrate the two objectives, coverage and continuity, into the bike lane planning model in view of the cyclists’ utility functions. Efficient model formulations and algorithms have been proposed to solve this large-scale planning problem.
About the speaker
Dr. Sheng Liu
Dr. Sheng Liu is an Assistant Professor of Operations Management and Statistics at the Rotman School of Management. His research interests lie in smart city operations, data analytics, and optimization. His research has been published in Management Science, Operations Research, Manufacturing & Service Operations Management, and INFORMS Journal on Computing.
He received a PhD in Operations Research from UC Berkeley in 2019, and a BSc in Industrial Engineering from Tsinghua University in 2014.
He has been working for leading tech companies such as Lyft and Amazon (as a Data/Research Scientist intern).
Presented by University of Toronto ITE Student Chapter, UT-ITE.