UK - Railway Data Studies

The "UK Studies" project involved conducting in-depth analyses of the operational and event data of mobility and infrastructure systems in England. The main objective was to provide relevant and actionable insights, delivered through customized reports (XLSX, PDF), to support strategic decision-making for the improvement and optimization of these infrastructures.

Problem:

How to effectively analyze complex mobility and infrastructure data for the UK market and extract strategic insights? It was necessary to design an automated system capable of generating customizable reports (types, periods, assets) and distributing them regularly to stakeholders.

Solution:

The solution involved the design and deployment of a robust Python pipeline, integrated with the client's HealthHub (HH) platform. This pipeline is responsible for processing operational and event data, generating reports in various formats (XLSX, PDF, Excel) based on defined input parameters (period, asset/train list, area of interest, type of reports), and their automated email distribution. The scheduling of these script executions is managed by specific Cron tasks, allowing for daily, weekly, or other frequencies of distribution.

Achievements:

• Requirements Gathering and Clarification: I actively collaborated with key stakeholders and partners abroad (notably in the UK) to gather, clarify, and integrate their specific needs regarding analysis and reporting.

• Management of Multiple Projects and Various Study Topics: I contributed to the execution of studies for about ten UK projects, covering diverse fleets of trains and UK clients. This allowed for the analysis of over thirty different and valued study topics, such as:

  • Event latency statistics

  • Analysis of KPI alerts

  • Event distribution (number/duration)

  • Event counting by area of interest (AOI)

  • Event delay reports

  • Reports on mileage and load of the Electrostar fleet

• Design and Development of Python Pipelines Independently: I designed and developed a complete Python pipeline for processing operational and event data, including the use of libraries such as openpyxl and fpdf2 for generating XLSX, PDF, and Excel reports. The alstompy library was also used to interact with the data.

• Automation and Customization of Reporting: I implemented a highly customizable report generation system, taking into account input parameters such as:

  • The data period

  • The specific asset/train list

  • The area of interest (AOI)

  • The desired type of reports (PDF, XLSX, etc.)

• Deployment and Integration: The script was successfully deployed within the client's associated HealthHub (HH) platform, ensuring its smooth integration into the existing ecosystem and the automated sending of reports via email.

• Management of Improvements: I submitted requests to the HH platform support teams for the addition of additional features, contributing to the evolution of the tool.

• Support: I also provided ongoing support on validated topics and deliverables once the reports were deployed.

• Workshop Organization: I organized and facilitated workshops with clients and users around the designed solutions, facilitating the adoption and continuous optimization of reports.

Technical Stack:

• Data Types: Events (alerts, KPI, incidents), Operational data (notably mileage and load)

• Language & Python Libraries: Python (openpyxl, alstompy, fpdf2, Requests, BeautifulSoup, Pandas, Numpy)

• Visualization / Reporting Tools: Tableau Software, Microsoft Excel, PDF

• Platform: HealthHub (HH)

• Orchestration / Scheduling: Cron

• Development Environments (IDE): PyCharm, Docker

• Language: English (for communication and reports)