SysML v2 Deep Dive 3 min read

Lesson 2: Understanding KerML

KerML: The Engine Powering SysML v2

For over a decade, systems engineers built complex models on a foundation that was never actually designed for them. Because SysML v1 was built on top of UML which is a software modeling language, engineers often had to force physical hardware into software concepts.

With the arrival of SysML v2, that legacy engine has been replaced. The new foundation is KerML (Kernel Modeling Language). To understand why SysML v2 is a major shift for MBSE, you have to understand the engine that drives it.

Why SysML v2 Changed Its UML Core

The primary limitation of SysML v1 was its software centric roots. In the past, modeling a physical pump or a mechanical sensor required using UML classes and adding stereotypes to define material or weight. While it worked in practice, it was conceptually awkward.

SysML v2 replaces that legacy engine with KerML. This is a foundation built from the ground up specifically for systems engineering rather than software development.

The New Architecture: The KerML in SysML v2

Before going into the specific parts of KerML, we need to clear up one common confusion that many people have. Are SysML v2 and KerML the same thing?

It is important to recognize that KerML and SysML v2 are distinct layers within the same ecosystem:

  • KerML is the platform independent meta language (the engine).
  • SysML v2 is the systems engineering language built on top of that engine.

The KerML Foundation: KerML consists of three logical layers that provide mathematical rigor:

  • Root (Syntactic Foundation): The structural base. It organizes elements and names but contains no specific engineering meaning yet.
  • Core (Logical Foundation): The mathematical heart. This layer uses formal logic to ensure the model can be checked for consistency and valid inheritance.
  • Kernel (Modeling Capabilities): This defines the Universe of Discourse, introducing the specific types of data and classes used to represent reality.

The Result: Unifying Entities in Space and Time

In SysML v1, Structure and Behavior were treated as two separate worlds. In KerML, these are unified under the concept of Occurrences. An Occurrence represents anything that exists in both space and time.

This 4D approach bridges the gap between what a thing is and what a thing does:

  • Objects: Occurrences that are spatial and persist over time (e.g., a physical tire).
  • Performances: Occurrences that are temporal and happen over time (e.g., the rotation of that tire).

Because a physical part and a behavior process are now fundamentally the same kind of thing, you can model the lifecycle of a part simply as a behavior. This finally resolves the traditional disconnect between structure and function.

The Payoff: Automated Reasoning

Because KerML is grounded in formal logic, your model is a database of facts rather than just a drawing. This enables inference, allowing the engine to answer complex questions mathematically. Instead of manually checking diagrams, the KerML foundation allows for automated validation, ensuring your system design is consistent and logically sound from the start.

Moving Forward

Understanding the engine powering SysML v2 is the first step toward mastering the next generation of systems modeling. By moving away from software based legacy structures, KerML provides the mathematical certainty required for modern engineering challenges.

SysML v2

SysModeler Team

Engineering & Product