AI-Assisted MBSE (AIM): Course Outline
SysML v2 AI workflows SEM anchor example Templates

AI-Assisted MBSE (AIM) teaches practical, repeatable workflows for using AI to dramatically accelerate SysML v2 modeling - without sacrificing rigor, structure, or traceability.

The course is organized around a concrete anchor example (a Scanning Electron Microscope) to demonstrate how models evolve from early concepts through architecture, simulation, implementation, and verification.

Back to AIM Training Open Enrollment Back to Training

This page provides the detailed module-by-module course outline. The module details below are presented in an expandable format so you can quickly skim the flow or drill into each module.

Template-Driven Acceleration with AI

A distinguishing feature of AIM is its emphasis on template-driven generation. Participants learn how to define reusable code and model generation templates that encode architectural decisions, modeling structure, and quality rules. AI is then used to populate those templates consistently and repeatedly.

This approach changes the economics of MBSE: engineers update intent or templates and regenerate artifacts, instead of manually reworking dozens of diagrams and specifications.

High-level generation strategy: generate system spec, then subsystem specs one at a time

At the highest level, AIM generates a system specification first, followed by detailed specifications for subsystems - one subsystem at a time. This staged approach establishes system-level intent before detailed architectural elaboration begins.

System specification template: domain model, high-level requirements, use cases, list of subsystems

Engineers begin by defining a system specification template. AI is then used to populate that template to generate a domain model, a set of high-level requirements, a set of use cases, and an initial list of subsystems.

Logical architecture template: subsystem requirements, parts decomposition, state machine, IBD

Once the system structure is in place, engineers apply logical architecture templates to each subsystem. These templates generate subsystem requirements, parts decomposition, subsystem state machines, and subsystem internal block diagrams. The result is dramatic acceleration: update intent or templates and regenerate artifacts, instead of manually reworking dozens of diagrams and specifications.

Who this course is for
  • Cameo users learning SysML v2 textual notation
  • Systems engineers exploring the model-as-code paradigm
  • Teams that want to start SysML v2 models much faster
  • Engineers using AI to accelerate MBSE without becoming programmers
  • Engineers who need SysML v2 but are confused by the textual language
What you will learn
  • Use SysML v2 textual notation without becoming a programmer
  • Generate subsystem specs a subsystem at a time with a single prompt
  • Use Cameo-compliant SysML v2 templates to avoid syntax errors
  • Import generated SysML v2 textual notation into Cameo Systems Modeler
  • Get SysML v2 models started dramatically faster
Detailed course modules

Click a module to expand the details.

  • Template-driven generation as a core AIM principle
  • Human responsibility for structure and intent
  • AI as an execution engine, not a system designer
  • Regeneration as a first-class engineering activity
  • Why software-first thinking changes architecture
  • Aligning SysML models with object-oriented design
  • Avoiding architectures that software teams ignore
  • Domain-driven thinking as a unifying approach
  • Domain modeling via system specification templates
  • Generating domain objects, relationships, and vocabulary
  • Ensuring consistency across systems and variants
  • Regenerating domain artifacts as understanding evolves
  • Structured requirements templates
  • Detecting ambiguity and testability issues
  • Maintaining traceability without excessive overhead
  • Using AI to assist engineering judgment
  • Use case templates that drive consistent narratives
  • Discovering alternate and exception paths
  • Bridging stakeholder intent and system behavior
  • Avoiding brittle, over-templated behavior models
  • Subsystem-level architectural templates
  • Generating decomposition, behavior, and interfaces
  • Avoiding functional-decomposition traps
  • Preserving architectural coherence through regeneration
  • Mapping logical elements to physical components
  • Preserving intent during architectural refinement
  • Managing constraints and implementation detail
  • Understanding where AI helps - and where it does not
  • Constraint and parametric templates
  • Common simulation pitfalls: units, bindings, context
  • Accelerating setup while preserving correctness
  • Integrating behavior and parametrics coherently
  • Code generation templates derived from system models
  • Using AI to populate templates across subsystems
  • Regenerating code as models evolve
  • Achieving significant reductions in development effort
  • Schema templates based on domain models
  • Aligning data design with behavior and requirements
  • Maintaining consistency through regeneration
  • Using AI within clearly defined structural guardrails
  • UI flow templates derived from use cases
  • Maintaining consistency between UI and system behavior
  • Accelerating UI specification with AI
  • Preserving traceability and testability
  • Test templates derived from behavior and requirements
  • AI-assisted generation of test suites
  • Regenerating tests as systems evolve
  • Coverage across nominal and off-nominal scenarios
Open Enrollment Training Available

AI-Assisted MBSE is available in an open enrollment format through Caltech CTME.

Free Downloads!!!

Exploring SysML v2? This free Lunar Lander example provides a complete SysML v2 model overview and diagram set.

Interested in Software?

Interested in using software specs with coding agents? Our two-day AIM for Software course focuses on spec-based coding with Codex, Claude Code, and ChatGPT.

Parallel Agile