Standardized Developer-Friendly Data Modeling with LinkML and the CIM

Bart Kleijngeld | June 20th, 2024

Hello 👋

  • My name is Bart Kleijngeld...

  • 📐 Data Architect

  • 👨‍💻 Software Developer

  • I work at Alliander, one of the Dutch electrical grid operators.

Data Challenges

Data Management within Alliander

We have a lot of data...

  • and a lot of teams.

Data and solutions emerge from everywhere, and managing it is a challenge:

  • What does the data mean?
  • How can the data be used (responsibly)?
  • Who owns the data?
  • Etc.

How can we deal with this?

  • We need to have a scalable way to produce data in a standardized manner.

Sharing Data with Other Organisations

Not just within the organisation do we face data challenges.

As a grid operator, we face a variety of challenges:

  • Sustainability deadlines
  • Network congestion
  • Personnel scarcity
  • Etc.

Collaboration with other organisations is becoming increasingly important to tackle these challenges.

  • Data sharing is a vital part of that.
  • Formal data models and standardization are essential to enable this.

We want FAIR data!

The IEC Common Information Model (CIM)

The CIM is an information model that's the result of decades long international collaboration.

It's a very large object oriented UML model in the Sparx Enterprise Architect application:

  • Classes represent entities
  • Classes have attributes
  • Between classes there exist various types of relationships

It provides a great vocabulary for speaking the same 'data language' between grid operators.

  • But Enterprise Architect poses issues.
  • Vendor lock-in
  • Top-down way of working
  • No Linked Data compatibility
  • Hard to interoperate with other tools
  • Hard to do version management and CI/CD automation
  • Heavily disliked by developers

How can we utilize this great standard such that it will be useful and adopted by developers?

First attempt: SHACL and OWL

There is also an OWL version of the CIM.

  • It is generated from the UML.

It's not perfect, but at least we have URIs which identify CIM model elements.

The Idea

  • Teams develop their own data models in SHACL.
  • Standardization is achieved through SHACL targeting CIM URIs.
  • We would write code generators...
    • And by "we", I mean "I".

Challenges

Many challenges arose:

  • Business minded people failed to understand the necessity of formal, machine readable models.
  • Developers often struggle to appreciate the need for information standardization.
  • SHACL and RDF were a hard sell on developers in terms of syntax and complexity.

Second attempt: LinkML

Facing these hardships, I reviewed my notes and uncovered a bookmark I wanted to take a closer look at.

  • It was the LinkML website.

Turns out LinkML is literally what I was aiming to build...

  • But better!

The Idea, Revised

  • Teams develop their own data models in LinkML.
  • Standardization is achieved through mapping to CIM URIs.
  • We can leverage existing generators, as well as bake our own...
    • And by "we", I mean "we".

This works better with LinkML:

  • One language, and accessible to developers as well!
  • Standardization and Linked Data are baked into the language.
  • It's code, so automation, transformations, CI/CD and versioning are all a lot easier.

Work Done

The simple, machine readable nature of LinkML makes it very easy to do cool things.

The CGMES LinkML Schemas

CGMES is an exchange standard based on the CIM.

  • I've written a script that has generated a LinkML schema for each CGMES profile.

Resources:

The CIM LinkML Schemas

Similarly, I've generated LinkML schemas for the entire CIM.

  • It is based on the official Sparx Enterprise Architect project.

Resources:

  • The script that generates the LinkML schemas.
  • The LinkML schemas.
  • The documentation hasn't been generated yet.

Profiler

Creating data models based on enormous standard models such as the CIM can be very hard and tricky.

Ritger Teunissen has built a CLI profiler that makes it easy to create schemas based on a list of classes.

  • The profiler will recursively fetch all dependencies.

Road Ahead

  • Dutch grid operators are currently using LinkML to define data models for data sharing.
  • At the European level LinkML is being considered as an interesting tool as well.
  • Within Alliander teams are starting to work with LinkML too.

*

  • Thanks for listening!