Railway Smart Passenger Information System (PIS)

This project focuses on the design and development of a smart passenger information system (PIS) for railway environments, combining real-time data ingestion with intelligent analytics.

The system aims to improve passenger experience by delivering accurate, timely, and contextual information while also supporting operational decision-making.

Key Contributions

• Design of an integrated passenger information architecture
• Use of analytics to process live and historical railway data
• Focus on scalability and real-world deployment constraints

Research to Practice

While my blog discusses the motivation and architectural thinking behind smart PIS systems, this project demonstrates how those ideas are converted into deployable, passenger-facing solutions aligned with modern railway operations.

Integration and objectives

The project integrates real-time passenger loading information into TrainFX Smart-PIS and enhances the PIS using big‑data analytics, reporting, and forecasting of dynamic, connected passenger information. The AI‑empowered Smart‑PIS will provide dynamic data acquisition and real‑time, personalized information to travelers, railway operators, and service providers. The system targets improved service efficiency, higher passenger satisfaction, and the enablement of new services through affordable, scalable technologies. It is a first‑of‑its‑kind solution in rail that supports bespoke customization aligned with industry objectives.

Key objectives:

• Integrate live passenger loading metrics into TrainFX Smart‑PIS
• Apply big‑data analytics for reporting and short‑/mid‑term passenger forecasting
• Deliver real‑time, personalized passenger information and alerts
• Build a scalable, low‑cost architecture for practical deployment
• Provide APIs and interfaces for operators and third‑party service providers
• Support operational decision‑making and service optimization

Knowledge base

Knowledge base lead:

• Professor Baihua Li — Professor of Computer Vision and Machine Learning, Department of Computer Science

Knowledge base supervisors:

• Professor Qinggang
• Dr Mohamad Saada
• Dr Haibin Cai

Photo: Katie Pea at TrainFX LMC4, 13 March 2025.

Photo: Katie Pea at TrainFX LMC5.