Artificial Intelligence-Aided Rail Transit Infrastructure Data Mining

Download Final Report

CAIT project no.: CAIT-UTC-REG43

Fiscal Year: 2020/2021

Status: Final

Principal investigator(s): Xiang Liu, Ph.D. (PI), Rutgers

Performing organization(s): Rutgers CAIT

Managing organization: Rutgers CAIT

In cooperation with: Metropolitan Transportation Authority
Partner project manager: David Kraft Sr. Director, Enterprise Asset Management Program Administration

Supported by: USDOT-OST-R

UTC, grant, or agreement no.: 69A3551847102


The primary goal of this proposal is to develop a pilot, proof-of-concept study in collaboration with the MTA to explore the use of Artificial Intelligence (AI) for analyzing infrastructure big data to predict track degradation and future condition.

The intended outcome of the project is a novel AI model for track infrastructure data mining in rail transit. This tool aims to forecast track infrastructure condition and to address track degradation before failures occur. This will lead to better life cycle performance and save the total life cycle cost while ensuring infrastructure safety and durability. MTA and other agencies could use the approach to prioritize their infrastructure capital planning, inspection and maintenance.