Case Study | Smart Trains
Smart technology is driving safer and more reliable transit.
The new breed of smart trains generate vast data sources from the train’s on-board telemetry which in turn feeds big data and BI systems to alert on key events such as engineering failures and geo-located track faults and to allow later analysis of trends including maintenance issues and customer activity.
Spike automation and performance services accelerated delivery and ensured performance which enabled the launch of some of the UK’s first smart trains.
- Solved technical challenges in testing highly complex data flows and analytics
- Accelerated testing of new features and bug fixes
- Integrated testing to Continuous Integration builds allowing daily feedback on build quality
- Ensured client could go-live with initial train deliveries
- Gave confidence on capacity for additional train rollouts
- Proved reliability of telemetry feeds used in critical decision making
- Spike led the design and implementation of an automated regression and smoke test pack for High and Medium risk test cases, reducing manual effort and accelerating testing of new features.
- We also performed a deep analysis of the technical platform, defined the performance test strategy and executed tests to identify key issues that were resolved prior to launch of the first UK smart trains.
Our Automation specialists developed a test automation framework taking Cucumber BDD tests and using open-source tools, principally Selenium, as the front-end driver and a number of add-ons to provide ETL, logging and reporting with a custom integration written to use HP ALM.
- Bespoke code was written to handle the high complexities involved in the Web client front end which used AngularJS framework, Java script, HTML, the Google Map API and HighCharts library as well as Scalable Vector Graphics.
- The final pack was also made “CI ready” to allow it to run on demand for each new build.
- We successfully delivered 250 automated regression tests in under 4 months to cover the core user journeys across the application.
- The test pack was executed nightly on a schedule to provide feedback to the offshore development team who made regular changes to the application due to the fast paced environment.
- The pack was used to sanity test environments immediately after build/refresh to understand quickly if any issues or problems had been introduced.
Our Performance experts drove the test strategy for performance testing including:
- Defining the Non-Functional Requirements
- Agreeing the performance testing planned to provide the confidence to allow the solution to be deployed to the Production environment
- Obtaining sign-off from key stakeholders recipients that the testing planned is achievable and the responsibilities of the test activities were understood and identified performance risks are mitigated
- Ensuring a common base for understanding of the test activities, terms and responsibilities across the project
- Performance test execution included:
- Assessing performance of key components including data lake, operations database, rules engine, analytics database, core web applications and BI analytics
- Developing test scripts against both onboard train simulator and live train data
- Executing testing at go-live and extended capacity levels enabling client to meet release targetstest
- Smart technology can provide huge advances in safety and reliability
- Data flows are complex and the volume of data is vast
- Automated testing is proven to accelerate the testing of complex flows and presentation screens and provide critical feedback to developers
- Performance testing is essential to ensure systems can process large volumes of data speedily and accurately
Download as a PDF
No contact details needed, click below to download a copy of our case study with our compliments. Please consider the environment and read online or print responsibly.