1 of 75

Making MBSE

More Rigorous and Agile with openCAESAR

Maged Elaasar, Ph.D.

Senior Computer Scientist

Jet Propulsion Laboratory

elaasar@jpl.nasa.gov

Yuta Nakajima

Visiting Researcher at Jet Propulsion Laboratory

(Engineering Test Satellite-9 Project)�Japan Aerospace Exploration Agency

nakajima.yuta@jaxa.jp

Aug 23, 2024

Copyright 2024 California Institute of Technology. Government sponsorship acknowledged

2 of 75

Disclaimer

Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not constitute or imply its endorsement by the United States Government or the Jet Propulsion Laboratory, California Institute of Technology.

2

jpl.nasa.gov

3 of 75

Outline

  1. History of openCAESAR
  2. Overview of openCAESAR
  3. Overview of OML
  4. openCAESAR Infusion at JPL
  5. openCAESAR outside JPL
  6. Case Study: openCAESAR at JAXA
  7. openCAESAR with UML/SysML
  8. openCAESAR Roadmap

3

jpl.nasa.gov

4 of 75

  1. History of openCAESAR

4

jpl.nasa.gov

5 of 75

Jet Propulsion Laboratory

  • Federally Funded R&D Center (FFRDC) in California
  • Managed by Caltech for NASA
  • 85 years of engineering complex space systems
  • Leader in the field of Systems Engineering

5

jpl.nasa.gov

6 of 75

Systems Engineering is an Information-Centric Domain

credit: Robert Karban, JPL

7 of 75

Systems Engineering Journey at

7

Integrated Model

Centric Engineering

Model Based

Systems Engineering

Document Based

Systems Engineering

jpl.nasa.gov

8 of 75

Document Based Systems Engineering (Europa Clipper)

8

Siemens CAPITAL Logic

logical

topological

wiring

Electrical Flight System Requirements

Systems Engineering

Subsystems Engineering

Concerns

  • Process inefficiency & cost (duplicated information)
  • Limited capability for detecting errors

jpl.nasa.gov

9 of 75

Model Based Systems Engineering (Europa Clipper)

9

Requirements Models

System Models

Simulation Models

System Views

Concerns:

  • No well defined Methodology
  • Poor interfaces between Tools

jpl.nasa.gov

10 of 75

MBSE Anecdotes at Europa Clipper

  1. Steep SysML v1 learning curve, hard to customize for methodology
  2. API changes in tool revisions forced revising analysis scripts
  3. Hard to integrate with other tools, needed to be triggered manually
  4. Breaking model in fragments made maintaining cross links fragile
  5. Working with branches was hard, model locking helped but was not ideal
  6. Consistency was only checked with explicitly defined rules
  7. Interfaces to get data in and out of the tool were very limited

10

[1] Elaasar, et. al: “The case for integrated model centric engineering”, Proc. of 10th Model Based Enterprise, NIST, MD, 2019.

jpl.nasa.gov

11 of 75

Integrated Model Centric Engineering (IMCE) Project

Mission: Developing Systems Engineering Methodology for JPL

  1. Using SysML profile directly (2012)
    1. Complex mapping to SysML
    2. Syntax only, no semantics

  • Using OWL ontologies then translating to SysML profile (2014)
    • Syntax and semantics using Description Logic (DL)
    • Developed API to translate between OWL and SysML

  • Using OWL ontologies only (2016)
    • Developing Ontological Modeling Language (OML)
    • Developing CAESAR modeling workbench (UI)
    • Developing openCAESAR framework (open source)

11

jpl.nasa.gov

12 of 75

: Computer Aided Engineering for System ARchitecture

12

2003

SysML

MBSE

Limited Adoption

2016

CAESAR

SysML Adapter

MBSE

Increasing Adoption

2018

2024

MBSE

Wider Adoption

CAESAR

Workbench

2014

2012

MBSE Exploration

2010

jpl.nasa.gov

13 of 75

2. Overview of openCAESAR

13

jpl.nasa.gov

14 of 75

openCAESAR Enables Digital Systems Engineering

14

RAISE: Rigorous, Agile, and Intelligent Systems Engineering

  • Using formal methods (precise semantic vocabularies, logical reasoners, model checkers and finders)

Rigor

  • Using DevOps methods (customizable vocabularies, flexible model componentization and federation, continuous model analysis and reporting)

Agility

  • Using AI methods (nature language progressing, machine learning, chatbot, copilot)

Intelligence

jpl.nasa.gov

15 of 75

openCAESAR Framework

15

  • Open-source domain-specific modeling and analysis framework

  • Uses Ontological Modeling Language (OML) as a lingua franca

  • Realizes the RAISE vision by integrating various technologies

  • Developed and maintained by the Jet Propulsion Laboratory

jpl.nasa.gov

16 of 75

openCAESAR Project

16

jpl.nasa.gov

17 of 75

openCAESAR Use Cases

  • UC 1: use ontologies as a native modeling formalism
  • UC 2: use ontologies to analyze models in existing tools
  • UC 3: use ontologies to create a digital thread between tools

17

UC1

UC2

UC3

jpl.nasa.gov

18 of 75

openCAESAR Addresses Modeling Pain Points

  • Models are frequently authored informally without a methodology
  • Models are hard to author based on a defined methodology
  • Models are hard to collaborate on within a team and between teams
  • Models are not rigorously checked
  • Models are not infrequently checked
  • Hard to do the creative modeling activities (requires expertise / training)
  • Hard to tell what/where is the model baseline is
  • Hard to propose a model change and have it reviewed and integrated
  • Hard to generate gate products from models
  • Hard to find the latest or historical gate products
  • Hard to trace the provenance of gate products

18

jpl.nasa.gov

19 of 75

openCAESAR Features (supporting RAISE)

  1. Model Representation

Using precise semantic web ontologies

  • Model Authoring

Using easily customizable viewpoints in modern IDEs

  • Model Configuration Management

Using Git repositories with branches, versions, and releases

  • Model Federation

Based on concern and authority with Maven dependencies

  1. Model Integration

Continuous integration of federated datasets with analyses

  • Model Analysis

Scalable analysis of consistency, completeness and correctness

  • Model Reporting

Using updated webviews framing different stakeholder concerns

19

jpl.nasa.gov

20 of 75

Ontological Modeling Language

A language for descriptive modeling with ontological semantics

  • Two levels of modeling: vocabularies and descriptions
  • Expressivity based on subset of description logic (DL)
  • Consistency checking w/explanations (DL reasoners)
  • Entailment generation & custom rules (simpler queries)
  • Open world assumption (extensibility)
  • Bundle closure algorithm (verifiability)
  • Axiom level model organization (modularity)
  • Abstract, textual, graphical syntaxes (accessibility)
  • Language server protocol (LSP), Java API (tooling)

20

2

DL

OML Luxor (VS Code Extension)

jpl.nasa.gov

21 of 75

openCAESAR Architecture

21

Workbench

Adapter

CI CD

Artifacts

Analysis Pipeline

DL Reasoner

Reports

Data

Other

Tools

Author / Configure / Federate

Integrate / Analyze

Report

Knowledge

Base

OWL

CoPilot

CoPilot

CoPilot

jpl.nasa.gov

22 of 75

3. Overview of OML

22

jpl.nasa.gov

23 of 75

OML: Enables a Methodological Use of OWL

OML is defined as profile of OWL2-DL with several benefits

23

  1. Improved authoring via higher abstractions
    1. By defining design patterns

  • Enabling closed-world verification
    • By using a world-closing algorithm

  • Scalable reasoning performance
    • By avoiding unions, negations, and recursions

  • Improved reasoning explanations
    • By minimizing anonymous entities

OML Abstraction

OWL Abstraction

Vocabulary, Vocabulary Bundle, Description, Description Bundle

Ontology

Extends, Uses, Includes

Import

Concept, Aspect, Relation Entity

Class

Scalar

Datatype

Scalar Property

Datatype Property

Relation

Object Property

Concept Instance, Relation Instance

Named Individual

Anonymous Concept Instance,

Anonymous Relation Instance

Anonymous Individual

Rule

Rule

translates into

jpl.nasa.gov

24 of 75

OML Abstract Syntax

Vocabulary

VocabularyBundle

Description

DescriptionBundle

Rule

Term

Instance

uses

uses

extends

extends

extends

extends

includes

includes

*

*

*

*

*

*

*

*

*

references

uses

*

*

*

Vocabularies

Descriptions

jpl.nasa.gov

25 of 75

OML Concrete Syntax

25

vocabulary <http://imce.jpl.nasa.gov/foundation/mission#> as mission {

extends <http://imce.jpl.nasa.gov/foundation/base#> as base

extends <http://www.w3.org/2000/01/rdf-schema#> as rdfs

concept Component < base:Container [

restricts all relation base:contains to Component

restricts all relation base:isContainedBy to Component

]

relation entity Presents [

from Component to Interface

forward presents reverse isPresentedBy

inverse functional

asymmetric

irreflexive

]

}

description <http://opencaesar.io/kepler16b/components#> as components {

uses <http://imce.jpl.nasa.gov/foundation/base#> as base

uses <http://imce.jpl.nasa.gov/foundation/mission#> as mission

instance orbiter-spacecraft : mission:Component [

base:hasIdentifier "C.02"

base:hasCanonicalName "Orbiter Spacecraft"

base:contains orbiter-power-subsystem

]

instance orbiter-power-subsystem : mission:Component [

base:hasIdentifier "C.02.01"

base:hasCanonicalName "Orbiter Power Subsystem"

]

}

Vocabularies

Descriptions

jpl.nasa.gov

26 of 75

OML Workbenches

26

Rosetta Workbench in Eclipse

(PRODUCTION)

Vision Workbench in VS Code (Web Based)

(NEXT GEN)

jpl.nasa.gov

27 of 75

OML Configuration Management

27

Version Control Repository

Artifact Repository

publish

jpl.nasa.gov

28 of 75

OML Ontology Stack

Define an OML vocabulary stack for a modeling methodology and use it to describe a system

28

Core Vocabularies

Upper Vocabularies

Vocabulary Bundle

Discipline Vocabularies

Methodology

Definition

Systems

Description

[xsd, rdf, rdfs, owl, dc, metrology, iso, etc.]

[imce, ufo, bfo, uml, sysml, capella, etc.]

[mechanical, electrical, scenarios, etc.]

[methodology bundle]

Descriptions

[system models]

Description Bundles

[system model bundles]

Vocabulary Bundle

[project’s extended methodology bundle]

jpl.nasa.gov

29 of 75

OML Stack Federation

29

Mechanical Vocabularies

Methodology Vocabularies

Kepler16b

Foundation Vocabularies

Electrical

Vocabularies

imports

Kepler16b

Mechanical

Kepler16b

Architecture

Kepler16b

Electrical

v5.0.0

v7.2.0

v3.2.8

v4.2.1

v1.0.3

v2.2.0

v3.0.0

v2.0.0

4.+

7.+

5.+

5.+

3.+

+

1.+

2.2.0

3.0.+

Methodology

Definition

Systems

Description

jpl.nasa.gov

30 of 75

OML Analysis Pipeline

30

Knowledge Base

SPARQL Query

Query Results

Analysis Notebooks

Published Reports

DL

Reasonor

Entailments

Ontologies

consistency report

jpl.nasa.gov

31 of 75

OML Tutorials

31

jpl.nasa.gov

32 of 75

4. openCAESAR Infusion at JPL

32

jpl.nasa.gov

33 of 75

openCAESAR Deployment at JPL

33

CAESAR Workbench

(Authoring)

CAESAR Cloud

(Reporting)

CI/CD

(Analyses)

Repositories

(Version Control)

jpl.nasa.gov

34 of 75

SE Methodology Development at JPL

34

Technologists

(implement SE methodology)

Collaborate

Subject Matter Experts

(specify SE methodology)

Project Stakeholders (analyze, review)

System Engineers

(author, link, reconcile)

Lead System Engineers

(configure, integrate, deliver)

App

MBSE Methodology Team

(IMCE)

Systems Engineering Teams

(NASA Missions)

jpl.nasa.gov

35 of 75

IMCE Foundation Vocabularies (Excerpt)

Defines foundational systems engineering patterns

35

Mission

mission:Junction

mission:Interface

mission:Component

joins

presents

Base

base:Container

base:ContainedElement

contains

mission:Function

performs

Project

project:Organization

project:Authority

project:Product

isResponsibleFor

produces

project:Role

approves

Analysis

analysis:Requirement

analysis:Constraint

analysis:CharacterizedElement

requires

constrains

jpl.nasa.gov

36 of 75

Discipline Application Development at JPL

36

Discipline Vocabulary

Authoring

using CAESSR Workbench

Analysis

Using CI/CD

Reporting

Using CAESAR Cloud

Consistency

Completeness

Correctness

Continuous Delivery

jpl.nasa.gov

37 of 75

openCAESAR Based Discipline Applications at JPL

  • System decomposition
  • Electrical and harness requirements
  • Power management
  • Mass management
  • Interface management

37

jpl.nasa.gov

38 of 75

Improved Harness Design at JPL with openCAESAR [4]

38

IMCE CAESAR

Review

Reports

Siemens CAPITAL Logic

logical

topological

wiring

architecture

report views

architecture

vocabularies

Continuous

Automated

analyses

[4] Wagner, D., Kim, S., Jimemez, A., Elaasar, M., Rouquette, N., Jenkins, S. “CAESAR Model-Based Approach to Harness Design,” Proceedings of IEEE Aerospace Conference, Big Sky, MT, USA, March, 2020.

jpl.nasa.gov

39 of 75

5. openCAESAR outside JPL

39

jpl.nasa.gov

40 of 75

openCAESAR Community

  • openCAESAR Project on Github
    • Examples (8)
    • Contributors (20+)
  • openCAESAR Blog
    • Posts (16)
    • Papers (9)
    • Tutorials (6)
    • Workshops (125 attendees)
  • onto:Nexus Group on LinkedIn
    • Members (153)
    • Posts (50+)
  • Collaborations
    • Pilots (10)
    • Invited Talks (50+)

40

jpl.nasa.gov

41 of 75

Some Members Exploring openCAESAR

  • Government Labs
    • NASA (US)
    • MITRE (US)
    • Army, Navy, Air Force (US)
    • ESA (Netherland)
    • JAXA (Japan)
    • DLR (Germay)
  • Corporate Labs
    • Ford Motors (US)
    • BAE Systems (UK)
    • Leonardo (UK)
    • Boeing (US)
    • Airbus (France, Germany)
  • Tool Vendors
    • CAE-List (France)
    • Autodesk (UK)
    • Vitech (US)
    • Mantech (US)
  • University Labs
    • Stevens Institute (US)
    • Arizona University (US)
    • University of Alabama (US)
    • Auburn University (US)
    • Johns Hopkins APL (US)
    • Carnegie Mellon University (US)
    • California State Northridge (US)
    • Polytechnique Montreal (Canada)
    • Concordia University (Canada)
    • Twente University (Netherlands)
    • ISAE Supaero (France)

41

jpl.nasa.gov

42 of 75

onto:Nexus Workshop

Int’l Workshop on Ontological Modeling and Analysis

  • 2024 edition: Infusing Rigor and Agility in MBSE Practice
    • Reporting on Results of Pre-Event Survey
    • MBSE@JPL: the CAESAR Journey
    • MBSE@Ford: Applying Ontological Approach
    • OML Review
    • OntoUML: How Semantics Help Description Rigor
    • Demo: System Model DevOps with OML Vision
    • Short Presentations by Community Members
    • R&D Directions for Onto MBSE
    • Panel: How Should We Collaborate?
    • Virtual Posters

  • 2025 edition: Jan 29th (we invite you!)

42

16 presentations

125 attendees

jpl.nasa.gov

43 of 75

6. Case Study

openCAESAR at JAXA

43

[1] Nakajima, Yuta. "MBSE-Driven Exploratory System Architecture Analysis Framework for the Engineering Test Satellite-9 Project." AIAA AVIATION FORUM AND ASCEND 2024. 2024.�[2] https://www.youtube.com/watch?v=rM2sJ5-mOrQ

jpl.nasa.gov

44 of 75

OML/openCAESAR Application to JAXA’s ETS-9 Project

OML + Automation Web Base Engineering Dashboard

44

Precise Language

CI/CD Technologies (Automation)

>

jpl.nasa.gov

45 of 75

Systems Engineering Methodology using openCAESAR

Role of “Methodologists” is a key for effective MBSE deployment.

Modeling

Identify Business Questions

Reasoning�+�Analysis

Reporting

visualize data

to answer Business questions

First Rule of Modeling

�Build the model to complement the analysis you intend to use it for.

What is the intent for the analysis?

What is the current workflow of target users?

How we describe things using OML?

Prototyping and user testing of

- Vocabularies

- Descriptions

- Analysis queries

- User Interactions

Define Methodology and guides systems engineers

Deploy products

Sustain products

Evolve products

Role of Methodologists

45

46 of 75

Information gathered by the project team is unstructured and decision-making based on data and analytics is limited.

Step.1: Identify Business Questions

Problem

Business Questions

  • How many panels are there in total?
  • How many of them complete production?
  • Which are the current AIT bottlenecks?
  • How do delays in power equipment affect the integration schedule?

Project Systems Engineers

Domain Experts�Manufactures

Local Server

Project Manager

Project Team

Project meeting

System Level Assembly and Integration phase

46

47 of 75

Step.2: Modeling

Methodologist ≠ System Modeler. Design process, vocabularies, tools and environment for automation.

Language

Tools

Environment

Framework

Describe knowledge

Business questions

Define a semantic business vocabulary

Check consistency

OML Vocabulary

owlReason

OML Description

Write queries

Run queries

SPARQL code

to answer Business questions

Gradle Task / JSON

Modeling

Reasoning/Analysis

Reporting

Import Data

Transform Data

Visualize

Publish

47

48 of 75

JPL IMCE Vocabulary

Systems Engineering Vocabularies

48

49 of 75

Modeling System Decomposition Using IMCE Vocabularies

mission:Component + Extensions

49

50 of 75

Step.3: Reasoning + Analysis

We use an openCAESAR as a framework for integrated digital engineering process

  1. A System Model in OML describes the structured data using a precise vocabulary and generates a spreadsheet.

  • Users input component status and attribute parameters such as scheduling information using their familiar tool(Excel).

  • Automatically generate and deploy web-based dashboards triggered by spreadsheet changes, leveraging GitLab's CI/CD.

50

51 of 75

Step.4: Reporting

The dashboard presents the most highly-digested information derived from CI analysis.

51

52 of 75

User Interactions: Input data through spreadsheets instead of OML

Spreadsheet user interface works well for just updating properties like development status, scheduling information, mass, vendor information, etc… (Not good for creating new structure but easy to adopt)

User Input Data

Instance Name

Contains

52

53 of 75

7. openCAESAR with UML/SysML

53

1:00 - 1:10

jpl.nasa.gov

54 of 75

UML/SysML Adapter Used in Europa Clipper

Analyzing a UML model defined with an ontologically-defined DSL

54

DSL

Vocabulary

uses

System Model

UML

Profile

2. Adapter

3. Adapter

System

Description

UML/SysML

Vocabulary

extends

uses

UML/SysML

Metamodel/Profile

extends

1. Adapter

Analysis Tools

applies profile

jpl.nasa.gov

55 of 75

Experiment for integration with OpenCAESAR and OpenMBEE

Adding strong semantic web capabilities to the SysML community

55

jpl.nasa.gov

56 of 75

Motivations

As a SysML user, we want to validate our model using ontological reasonings.

56

In OML, the reasoner tells us the inconsistency.

jpl.nasa.gov

57 of 75

Power of OML with Semantic Web Technologies

OML Model ( Vocabularies + Descriptions)

57

Logical Reasoner helps Systems Engineers to detect errors and to query the model

jpl.nasa.gov

58 of 75

OML bridges the gap between SE and Semantic Web technologies.

Ontology Modeling Language(OML) is a user-friendly way to describe information in OWL 2 DL, to improve the speed and quality of modeling and analysis.

58

jpl.nasa.gov

59 of 75

Prototype using OpenMBEE Execubot

59

Setup:

OpenCAESAR

  • OML-Luxor
  • OML-Vision

OpenMBEE

  • MMS-Execubot
  • MDK
  • View Editor
  • Cameo Systems Modeler

jpl.nasa.gov

60 of 75

Detect Ontology is inconsistent!

SysML model is transformed into oml and model is validated using ”owlReason”

60

jpl.nasa.gov

61 of 75

Experiment for Integration of OpenCAESAR and OpenMBEE Workflow Detail

61

jpl.nasa.gov

62 of 75

Data Exchange Architecture

Using OpenMBEE as a hub to integrate Cameo and OpenCAESAR

62

jpl.nasa.gov

63 of 75

Detail of prototype

1. OML -> SysML : Transform OML Vocabulary to SysML Profile in CSM

63

jpl.nasa.gov

64 of 75

Detail of prototype

2. SysML -> Json : Using MMS to extract SysML data as JSON ( transformed to df in R)

64

jpl.nasa.gov

65 of 75

Detail of prototype

3. SysML -> OML : Using R to transform SysML Json data into OML Descriptions

65

jpl.nasa.gov

66 of 75

8. openCAESAR Roadmap

66

jpl.nasa.gov

67 of 75

  • Carries R&D Projects related to openCAESAR
  • Provides internships to undergrad/grad students

Rigorous, Agile, and Intelligent Systems Engineering Lab

67

jpl.nasa.gov

68 of 75

Representing SysML2 in

68

KerML 2.0 Metamodel

SysML 2.0 Metamodel

System Model

SysML 2.0 Libraries

Vocabulary

Description

KerML 2.0 Vocabulary

SysML 2.0 Vocabulary

System Description

SysML 2.0 Libraries

full

full

partial

Work in progress

jpl.nasa.gov

69 of 75

Developing CoPilots with LLMs

69

Workbench

Vocabularies

DL Reasoner

Reports

Graph DB

OWL

Ontologies

CoPilot

CoPilot

CoPilot

govern

R&TD Proposal

R&TD Proposal

SEAL CoPilot

Work in progress

jpl.nasa.gov

70 of 75

Integrating with FRET Tool

  • Formal Requirements Elicitation Tool (FRET) by NASA Ames

  • Formalizes requirements in a specialized natural language mapped to Linear Temporal Logic (LTL)

  • Validates requirements via text, diagrams and simulation

  • Checks requirement realizability using SMT reasoner

  • Verifies requirements in simulation via runtime monitors

70

Work in progress

ontologies

jpl.nasa.gov

71 of 75

Summary

71

jpl.nasa.gov

72 of 75

Learning from OpenCAESAR

MBSE, Digital Engineering, Ontology, Architecting, VnV, State Analysis, Formal Requirement… Essentially, the main purpose is to leverage computers to improve human collaboration.

72

jpl.nasa.gov

73 of 75

IMCE Vocabularies helps Abstraction Thinking

Ontology for Systems Engineering

73

jpl.nasa.gov

74 of 75

74

Thanks

Any questions?

Yuta Nakajima

Maged Elaasar

jpl.nasa.gov

75 of 75

75

jpl.nasa.gov