Use Case Intelligence RSS Search use cases Search
IBM Manufacturing Spain

CAF Signalling

Enterprise Knowledge Search

CAF Signalling's onboard systems team had to repeatedly search, interpret and adapt dense ERTMS documentation, making it hard to access the right standards quickly and keep rail signaling implementations compliant as requirements evolved.

Read the original story Explore all use cases
3 weeks duration of proof-of-concept (POC)

Reported impact

Business outcome

CAF Signalling created a centralized, AI-assisted way to consume complex standards information in a three-week proof of concept, reducing documentation friction and helping engineering teams work more efficiently on interoperable rail-signaling projects.

  • Compliance
  • Productivity Improvement

Solution

CAF Signalling worked with IBM Client Engineering to build a generative AI knowledge assistant for European Rail Traffic Management System documentation. The implementation combined IBM Watson Discovery, IBM watsonx.ai and IBM watsonx Assistant so engineering teams could query a centralized corpus of manuals and standards instead of manually hunting through fragmented documents. Discovery ingests and indexes the ERTMS material and returns relevant passages for a user question; watsonx.ai uses those retrieved passages to generate a grounded natural-language answer; and watsonx Assistant provides the conversational interface that exposes the workflow to engineers. The result is a governed retrieval flow for technical standards that reduces search friction, helps teams find the right requirement faster and supports interoperability and compliance work across international rail projects.

Data flow

ERTMS manuals and standards are ingested into Watson Discovery, engineers submit questions through watsonx Assistant, Discovery returns matching passages, and watsonx.ai turns that evidence into a natural-language answer that supports signaling design and compliance work.

Solution architecture

3 components · 3 layers
  1. Ingest
    • IBM Watson Discovery Indexes ERTMS manuals and technical documents, then retrieves the passages most relevant to each engineering query.
  2. Compute
    • IBM watsonx.ai Uses retrieved standards content to generate grounded answers that simplify interpretation of complex rail-signaling requirements.
  3. Serving
    • IBM watsonx Assistant Provides the conversational front end engineers use to ask questions and consume AI-generated guidance.

Architecture clues

  • Discovery retrieval supplies the evidence used by the generative-answer layer.
  • The solution was built to support access to ERTMS standards rather than replace engineering decision-making.
  • The workflow was aimed at improving interoperability across international rail networks.

Evidence from the source

3 weeks
IBM Watson Discovery
IBM watsonx Assistant
IBM watsonx.ai
to facilitate access to the ERTMS standards
Transparency Evidence & methodology 85/100 reference value · 89% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value85/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact15/25
    1 metric describe measurable impact.
  • Deployment scale5/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence89%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability21/25
    10 of 12 key fields link to a named source.
  • Direct evidence20/20
    5 direct source quotes retained.
  • Date provenance10/10
    The publication date comes from the source page or publisher.
  • Source transparency8/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. IBM

    CAF Signalling customer story

    First-party
    https://www.ibm.com/case-studies/caf-signalling
    Source type
    Vendor-published customer story
    Published
    Apr 3, 2024
    Supports
    • Customer
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes
    • Publication date

Date evidence

Not separately sourced Industry, Region, Company size. Treat these normalized attributes as lower-confidence until a dedicated source is added.

Related

Similar use cases

See all like this
IBM May 22, 2024 MIKOPTIK Manufacturing China MIKOPTIK needed to reduce the time engineers spent finding R&D knowledge and coordinating across siloed systems because access to design expertise depended too heavily on a small number of key specialists. 40% boost in access to engineering information Enterprise Knowledge Search IBM Cloud Pak for Business Automation · IBM Cloud Pak for Integration · IBM Watson Discovery IBM Apr 22, 2024 Enztec Manufacturing New Zealand Enztec researchers spent nearly 90% of their work week gathering, filtering and reviewing medical publications, slowing the discovery phase of new orthopedic device projects. nearly 90% of their work week Enterprise Knowledge Search IBM Watson Discovery · IBM watsonx.ai Microsoft Sep 24, 2026 Rockwell Automation Manufacturing Singapore Rockwell needed to reduce the time and inconsistency involved in diagnosing machine faults, preserve retiring workers' tacit shop-floor knowledge, and shorten training time for new technicians supporting hundreds of… costs are down by about 25 percent Enterprise Knowledge Search Azure AI Search · Azure OpenAI Service · Azure security, identity and governance services