Identifying Priority HSR Routes

An exploratory AI search problem designed to find an optimal High-Speed Rail (HSR) network for the US. This is a final project for Northeastern's CS4100 - Artificial Intelligence.

This project links census population data with city airport codes to analyze population served by flight routes and identify high-traffic and high-emissions corridors, then visualizes input flight data, census data, and outputted solutions using GeoPandas and Matplotlib.

Description of the Search Problem

This project explores solutions to two search problems, numbered Problem 1 and Problem 2:

Problem 1

Given a fixed construction budget, what is the optimal selection of intercity rail corridors to build?

• The cost of building a certain rail segment is the negative of the "benefit" gained from the segment, as defined by the metrics we are interested in.
• The goal state is any state in which the total monetary costs of constructing all present rail segments is equal to the predetermined budget.

Problem 2

Given a goal of improving national metrics (emissions, population served by rail, etc.) by a specified amount, what is the least-cost investment necessary to achieve this, and which arrangement of intercity rail corridors would satisfy this?

• The cost of building a certain rail segment is the construction cost.
• The goal state is any state in which the total benefit gained by all rail segments is at least the predetermined threshold.

To run:

Navigate to the project root and execute: python search/hsr.py

To generate the data needed for hsr.py:

1. Execute python preprocessing/generate_intercity_passenger_vol.py
2. Run the notebook preprocessing/generate_algo_testing_date.ipynb

To visualize the outputted solution:

Run the notebook search/explore-solution.ipynb

Authors

Jonah Stadtmauer, Vinit Patel