UM Principle Investigator: Siqian Shen
Project title: Learning-Based Distributionally Robust Optimization Method for Route Planning
Funding Source: Didi Chuxing
Route planning is fundamental to most transportation services. To design and operate a route planning model, decision makers face two main challenges: (i) diverse objectives of the platform, drivers, customers and societal benefits; (ii) highly uncertain, dynamic, and correlated traffic conditions. The purpose of this collaborative research is to design, implement, and evaluate a novel route planning model via the combination of robust optimization and machine learning with extracted knowledge from historical transportation trajectory data. The work will lead to the invention of new methodologies and practical impacts.