Siemens UK RailXplore: Migrating Railway Systems Without Guesswork

Rail infrastructure produces diagnostic data without pause: signalling, track vacancy detection, point machines, each reporting on its own terms. Collecting it has never been the hard part. The hard part is making streams from separate systems answer one question about the health of the network, quickly enough to act on.

Estates like this rarely fail because a team cannot write code. They fail when migration is treated as a rewrite race: unclear event boundaries, weak observability, and no clean path to hand the system on.

Siemens UK's RailXplore programme needed a migration path that could carry real railway system concerns — not a prototype that looked modern in a slide deck.

Approach

We designed and delivered a serverless, event-driven architecture on AWS, with a React frontend and Node.js services, so the migration could move in controlled increments.

Event-driven was not a stylistic choice. Diagnostic equipment emits a continuous stream of independent events and does not wait for a request-response cycle, so the messaging layer is where the architecture is decided. Kafka carried the durable, replayable streams, where losing an event is not acceptable and consumers need to read at their own pace; RabbitMQ and Amazon MQ handled work that was genuinely queue-shaped. Getting that split right is what keeps a diagnostics platform from silently dropping the reading that mattered.

Engineering

• AWS Lambda and API Gateway as the event-driven edge

• React frontend with Node.js backend services

• DynamoDB for high-volume diagnostic and sensor telemetry

• Kafka for durable event streams, with RabbitMQ and Amazon MQ where the operating model called for queues

• S3, EC2 and Elastic Beanstalk across the surrounding estate, with an Oracle NetSuite integration

• Infrastructure expressed with CloudFormation; monitoring via CloudWatch

• Documentation and training so the system could be owned after delivery

Outcome

A migration architecture that could be operated, monitored and handed over — not a one-off build that only the original team understood.

RailXplore has since grown into Siemens Mobility's integrated rail diagnostics platform, unifying diagnostic modules behind one API-first interface. Unglamorous groundwork — clean event boundaries, telemetry that holds up under volume, infrastructure described in code — is what makes a platform still extensible years later.