Autonomous Metro Rail Sensor Data Analytics for Anomaly Detection and Correlation Analysis with Environmental Variables

Anomaly Detection (AD) is a key research problem in the fast-evolving area of Intelligent Transportation Systems. To proactively detect faults, AD methods are applied to detect deviation from regular behavior, which can also be caused by environmental factors such as weather conditions, neighbouring trains, and track curvature. Thales Canada has unlabeled data from their Next-Generation Positioning System (NGPS) on-board their Train Autonomy Platform which contains speed and positioning readings from sensors onboard multiple trains collected over multiple years. The project will develop an AD tool to identify anomalous readings in the NGPS data by modeling the regular behavior of the train for different source and destination, number of stops, and time of the day, and thereby, identify anomalous behavior, and compute correlations of the run data with external factors. The AD tool will allow Thales to automate the metro-rail system with minimal human supervision and ensure a safer metro-rail system.

Faculty Supervisor:

Farhana Zulkernine

Student:

Partner:

Thales Canada Inc (North York, ON)

Discipline:

Computer science

Sector:

Professional, scientific and technical services; Transportation and warehousing

University:

Queen's University

Program: