Optimization of shared on-demand transportation

Abstract

Urban growth globally increases urban commuting, causing congestion, pollution, and health risks. Technology-driven transportation innovations, like on-demand and shared mobility services, aim to address these challenges. Integrating these with public transit could revolutionize transportation. This thesis explores optimizing on-demand transportation in cities through three methods:

1. Analyzing historical travel time data improves accuracy, benefiting traffic optimization and congestion identification.
2. Implementing a real-time shuttle ridesharing algorithm reduces waiting times and in-car delays for taxi rides in Manhattan.
3. Developing precise short-term demand forecasting models, particularly using deep learning techniques, enhances prediction accuracy.

This research provides insights for optimizing urban traffic operations, improving ridesharing services, and efficiently planning fleet dispatching.