This project deployed a tool designed to identify, track, and visualise disruption data related to accessibility across the Transport for London (TfL) network. I wanted to come up with a project that tested my use of APIs, live reporting, and dashboard by creating an accessibility solution after hearing the issues users face across public transport from a Policy Connect Ressearch Fellow.

User Side:

For many passengers with accessibility needs, being informed about a station's step-free status is a hassle. All too often people only find out lifts are out of service when they reach a station, delaying their journey and causing undue stress.

While Transport for London (TfL) does publish accessibility and disruption information, it is often:

• Hard to navigate in real-time
• Not centralised for accessibility concerns
• Lacking historical insight into how often disruptions affect certain stations

Technical:

Aside from the user case, the project involved working with several tools I was previously unfamiliar with.

These included:

• Navigating the TfL Unified API for the first time
• Working with live data
• Utilising Airflow for data processing
• Ensuring low-latency response times to notify users live

I aimed to address this issue through the following implementations:

• Providing a filterable, visual dashboard of accessibility-related disruptions (Streamlit)
• Automating the collection of live data from the TfL Unified API
• Making real-time and historical insights available in one place
• Notifying users as soon as their station requires steps, and rapidly rerouting them

Tech Stack:

• Airflow: DAG scheduling and automation
• Streamlit: Interactive dashboard frontend
• Asyncio + REST APIs: Efficient data collection and backend services
• Pandas: Data processing and transformation
• PostgreSQL: Backend storage