Case Study | CAF: Maintenance Processes Optimized by 16-21% Using Massive Daily Sensor Data
CAF is a multinational group with more than 100 years of experience offering integral transport systems being at the forefront of technology and adding high value in sustainable mobility. As a reference in the railway sector, it offers to its clients one of the widest and most flexible ranges on the market in rolling stock, components, infrastructures, signalling and services (maintenance, rehabilitation and financial services).
Within this framework, CAF launched several years ago an initiative called “Tren Digital”, which led to the creation of the LeadMind platform.
LeadMind provides a new generation of connected trains and more competitive services for railway industry operators and maintainers through the collection, storage, processing and advanced analysis to support real-time decision making and move towards condition/predictive based maintenance.
The transport sector has adopted industry standards 4.0.
Industry 4.0 standards characterised by intelligent systems and Internet-based industrial solutions, has been adopted by the transport sector, especially the railway sector.
The use of new technologies is leading to improvements in the quality of services and business models, based on the analytical capabilities of large data and their potential to transform current platforms into a network of collaborative communities that move the transport of goods and passengers. The current trend in automation and data exchange is towards the adoption of new and emerging technologies to achieve greater levels of efficiency and effectiveness.
Solution on AWS
CAF relies on Keepler Data Tech for the integration of AWS technological solutions pursuing two main technological objectives:
Implement LeadMind Analytics solution in a cloud architecture to ingest, process and storage, both batch and real time, if needed, from train data.
Generate reports and dashboards with KPIs identified and categorized by CAF.
In order to avoid over-engineering and supported by agile methodologies, Keepler’s approach is the construction of minimum viable products using Big Data processing services and AWS analytics, which allow to validate
technologies and approaches used to solve specific problems, with a sufficient scope and measure the effectiveness in a simple way.
The result is a comprehensive solution that receives data from trains and processes the information so that it is properly stored in a Data Lake. The solution allows to ingest data from a limited set of vehicles equipped with diagnostic units (sDiag) and scale to any number of vehicles in the future.
The solution is based on the use of managed services, which achieves a serverless implementation easy to maintain, robust, secure and scalable.
AWS services used were as follows
AWS S3 as main storage repository.
AWS Athena to query Data Lake using SQL.
AWS Glue as an ETL tool and Data Catalog.
AWS EC2 for BI services with TIBCO Spotfire.
AWS Glacier as backup of old files.
AWS SageMaker to launch iPython Notebooks, used by CAF data scientists to develop new models.
AWS Redshift automatically loaded with a subset of data processed from source data to optimize Business Intelligence processes.
Amazon DynamoDB as metadata storage.
AWS RDS (with MySQL) as master data storage for field transformations.
AWS Batch for FTP synchronization.
AWS Lambda to execute detection application logic in the ETL and near-real time alarms.
AWS SNS and AWS SES to process errors and near-real time notifications.
Benefits for the client
Pay Per Use Model
The pay per use model of the public cloud has allowed CAF to have a solution that reduces considerably costs of investment.
Cost Reduction
As it is a solution implemented entirely through managed services, the operating cost is reduced.
Horizontal Scaling for Sensor Integration
All parts of the solution scale horizontally, so the integration of more sensors or the increase in train fleet does not represent a bottleneck and allows agile and automatic escalation.
Open System
It is an open system that allows the integration of any tool, that can be deployed on AWS, for exploiting the information.
Cost-Efficient Data Storage and Processing
The storage and processing cost is significantly lower. For example, the processing of all the historical data presents a time reduction of more than 90% compared to previous on-premise solutions.
Keepler is a full-stack analytics services company specialized in the design, construction, deployment and operation of advanced public cloud analytics custom-made solutions. We bring to the market the Data Product concept, which is a fully automated, public cloud services-based, tailored software that adds advanced analytics, data engineering, massive data processing, and monitoring features. In addition, we help our customers transition to using public cloud services securely and improve data governance to make the organization more data-centric.
Would you like to talk about your business?
We can help you leverage the power of data to enhance your operations.