Alstom Drives Operational Excellence with Predictive Maintenance on Cloud 

About Alstom

Alstom, a global leader in sustainable and smart mobility solutions, offers a comprehensive range of rail transport systems, including high-speed trains and advanced signalling solutions. For Alstom and its customers, maintaining operational excellence and minimizing disruptions are critical to ensuring reliable and safe train travel. In partnership with Devoteam, Alstom implemented a custom-built, cloud-powered predictive maintenance solution on Amazon Web Services (AWS), enhancing the reliability and safety of train operations.

The Challenge

Traditionally, Alstom’s operators and analysts manually sifted through vast amounts of train log data to identify error patterns that could indicate potential failures. This manual approach was time-consuming and often insufficient in proactively addressing issues that could lead to costly disruptions. Alstom sought a solution to:

• Proactively Identify Potential Issues: Move from reactive to proactive maintenance to optimize train uptime.
• Enhance Customer Satisfaction: Reduce unexpected delays and improve the overall reliability of train services.
• Streamline Data Analysis: Efficiently analyze massive datasets to uncover actionable insights.

Our Goal: Develop a user-friendly application supported by a robust infrastructure that gathers and analyzes diverse data types, identifies the root causes of failures in near real-time, and enables operators and analysts to predict future issues with high accuracy.

The Solution

In collaboration with Devoteam, Alstom embarked on a digital transformation journey to automate the analysis of onboard data and detect critical error patterns. Leveraging AWS, the EBISense Onboard Platform was developed, offering a proactive, data-driven approach to maintenance. Key components of the solution include:

1. Seamless Data Integration
   • Sensor Data: Real-time and historical data from various sensors, including temperature, speed, humidity, vibration, and pressure sensors, are securely collected from Alstom’s train fleets.
   • Performance Metrics: Continuous monitoring of performance indicators such as engine performance, brake efficiency, and energy consumption.
   • Hardware Health Data: Detailed information on the health and status of critical hardware components to detect wear and tear or impending failures.
   • Integration with Existing Systems: Direct integration with Alstom’s existing maintenance and operational systems, ensuring technicians have seamless access to all necessary information.
2. Advanced Analytics Engine
   • Scalable Infrastructure: AWS provides the robust and scalable infrastructure needed to support sophisticated predictive models.
   • Machine Learning Models: Continuous analysis of data from multiple sources using machine learning algorithms to detect patterns that indicate potential component failures.
   • Real-Time Processing: Near real-time data processing ensures timely identification of issues, enabling swift preventive actions.
3. Visualized Insights and Alerts
   • User-Friendly Interface: An intuitive dashboard presents clear, actionable insights and alerts to Alstom’s maintenance teams.
   • Actionable Alerts: Automated notifications highlight potential equipment issues, allowing for prompt intervention.
   • Customizable Views: Maintenance teams can tailor the interface to focus on specific data points relevant to their roles.
4. Predictive Maintenance Package
   • Predictive Maintenance Data Assessment: Comprehensive evaluation of existing data sources and identification of additional data requirements (e.g., temperature, speed, humidity, vibration) to enhance predictive capabilities.
   • Predictive Maintenance Data Platform: A robust platform built on AWS that consolidates and manages diverse data streams, ensuring data integrity and accessibility.
   • Predictive Maintenance MLOps Platform: Streamlined machine learning operations that facilitate the deployment, monitoring, and maintenance of predictive models, ensuring continuous improvement and scalability.

The Results

The implementation of the Alstom Onboard Platform has revolutionized Alstom’s train maintenance operations, delivering significant benefits:

• Reduced Downtime and Delays
  • Preventive Maintenance: Predictive insights enable scheduled maintenance and precautionary measures, preventing disruptive breakdowns.
  • Cost Savings: Minimizing unexpected downtimes reduces costly repairs and enhances on-time performance.
• Optimized Resource Allocation
  • Proactive Deployment: Maintenance teams are strategically deployed based on data-driven predictions, increasing operational efficiency.
  • Timely Upkeep: Ensures that maintenance activities are performed precisely when needed, avoiding overuse or underuse of resources.
• Enhanced System Reliability
  • Fewer Unscheduled Interruptions: Predictive maintenance leads to a significant reduction in unexpected service interruptions.
  • Improved Passenger Experience: Increased train reliability enhances overall passenger satisfaction and trust in Alstom’s services.
• Data-Driven Innovation
  • Continuous Improvement: Ongoing data collection and analysis foster a culture of continuous improvement and innovation within Alstom’s operations.
  • Scalable Solutions: The platform’s scalability allows Alstom to extend predictive maintenance practices to other aspects of their operations, driving further efficiencies.