This document discusses model-based system engineering (MBSE) and its application to the design of electro-optical sensors. It describes how MBSE differs from traditional engineering approaches by enabling integrated modeling across disciplines. Examples are given that show how MBSE allows designs to be evaluated earlier and problems to be identified and resolved more quickly. Both advantages and disadvantages of the MBSE approach are outlined. The author's goal of promoting wider adoption of MBSE methods through consulting is also mentioned.
2. Introduction
• What is Model Based System Engineering (MBSE)?
• How does it differ from current practice?
• A MBSE environment for Electro-Optical (EO) sensors
– Software interface and features
– Useful design and analysis features
• Pros and cons of the MBSE approach
• My goal as a consultant in concurrent design and MBSE
• References
3. Two Different Paradigms for making Complex Products
Assembly Line for Model T Fords
(Source: www.shadetreemechanic.com) Antonio Stradivari making violins
(Source: www.altontobey.com)
4. Traditional Product Organization
- Division of labor into specialized
tasks
- Documentation ties tasks together
- Hierarchical control
- Prone to delays and errors at each of
the many manual handoffs needed to
execute a project
-Each engineering discipline works
to its own requirements
- CAD/CAE tools and models unique
to each discipline
- Little interdisciplinary interaction
unless requirements are violated
-No integrated, high-fidelity model of
overall product function
- Error budget promotes overdesign
(design in excess of requirements)
Does not promote
ownership for overall
product success
among team members
5. Model-Based System Engineering Organization
-Work is organized around a whole product
- One lead engineer/decision maker per
discipline area
- The team of leads is collectively responsible
for system engineering and product success
- Engineering data and models shared across discipline
boundaries with MBSE software tools
-Project tree file structure ensures version control, model
consistency as design evolves
- All discipline models derive from a common CAD model
- Engineers use the CAD/CAE tools they prefer
- PP chart snapshots of design status not needed for reviews
-Concurrent design sessions alternate with off-line
detailed design work
- All models projected to common screen for
design assessment and problem resolution
- Design errors and conflicts caught early and often
- Design cycle time dramatically reduced (2X to 3X)
by integrating discipline models for re-use
6. Files
Typical Federated Software Integration Approach
Excel
Thermal
Structural
FEA
Optics
SigFit
SigFit
MATLAB
Results Extraction
CAD Mesh
CAD Mesh
Optics: Nominal Design
Federated System:
Metadata & Results
Comet MBE Framework
Abstract Engineering (Functional) Model
Adaptors to External Tools
CAD
Optics:
Nominal Design
Systems Model
(e.g., Excel, SysM
L)
Comet Unified Data Model & Process
Excel
Thermal
Structural
FEA
SigFit Optics
MATLAB
Results Extraction
Mesher
CAD Mesh
CAD Mesh
Federated System:
Metadata & Results
Version 1
7. Comet’s Performance Engineering Workspace
Project Tree:
Stages
Simulation Process
Geometry/Mesh/Results
Viewers
• Ensures version control for all models
• Access to all data and history
• Manage all CAE model configurations
and simulation results
• Capture integrated simulations
• Developed by team of SME’s
• Automates iterations for different
design forms, boundary conditions,
material properties, etc.
• Manage Performance Requirements
• Compare simulation results with Requirements
• Manage high-level Constants and Variables
• Access CAD geometry of all formats
• Create complex meshes
• Visualize results from all CAE codes
Project Dashboard
Constants Variables Requirements Metrics
8. MBSE Software Provides Direct Access to all Principal
CAD/CAE Data
CAD and
Structures
Thermal
Optics
9. MBSE Enables Integrated Design – Example 1
CAD and Structures models are integrated in
a Comet Simulation Process
CAD and structures models viewed in
real time, side-by-side
Mass and maximum Von Mises
stress levels viewed in dashboard
CAD engineer sees effects of design changes on structural
response in real time as he tweaks design to meet launch load
requirements – 6 design iterations were made in 3 hours
10. MBSE and Integrated Design – Example 2:
Integrated Structural/Thermal/Optical Analysis
Computes telescope Wavefront Errors
(WFE) resulting from thermally induced
changes in the telescope optical
components and metering structure
11. MBSE and Integrated Design – Example 3:
Integration of Models at Different Levels of Fidelity
The CAD design for an EO sensor payload can
be input to a higher level spreadsheet based
integrated design for the entire space segment.
The space segment environment generates top-
level designs for:
Systems
Attitude determination and control
Astrodynamics
Command and data handling
Communications
Power and data processing
Propulsion
Software
Telemetry and tracking
Overall mass and power
Space segment design outputs (bus size and
configuration, solar array size and placement)
are, in turn, used as inputs for payload thermal
design.
12. MBSE Facilitates Discovery and Diagnosis of Design
Problems
Telescope WFE is
very large – about
200 waves of
spherical aberration
Almost all of the
WFE is coming
from the primary
mirror
Structures model shows WFE to be due to bending of
the mirror due to excess clamping force at cold
temperatures. Remedy = refining thermal design or
more compliant primary mirror mount.
Design issues and conflicts can be diagnosed
directly within the MBSE design environment.
Design changes can be made and analyses
repeated 2X to 3X faster than is possible
when discipline designs are not integrated in
this fashion.
13. Pros and Cons of MBSE
• Disadvantages of MBSE
– Logistically more complex and expensive to implement
– Additional cost of MBSE software and a concurrent design facility
– Requires regular assembly of the engineering team and customer representatives
for half-day concurrent design sessions
• Advantages of MBSE
– Enables higher fidelity instrument designs earlier in the program cycle at a fraction
of the current cost
– Enables discovery and root cause determination of design problems and conflicts
early and often throughout the instrument development process.
– Enables dramatic (2X to 3X) reduction in design cycle times when integrated
simulation processes can be utilized
– Enables models to be integrated at appropriate level of fidelity at each stage of
design evolution
– It’s just a lot more fun to work this way
14. My Consultancy Goals
• I’ve been involved with concurrent design of EO sensors for the past 15 years
– First as the cognizant optical engineer on the concurrent design team that designed and built
the Miniature Integrated Camera and Spectrometer (MICAS) instrument during my days at
JPL
– Culminating in the development and application of a MBSE design environment that has been
the subject of this presentation
• My goal now is to use my experience as both a member and a leader of MBSE teams to help
spread the application to MBSE methods and tools to a wider range of civil and commercial
organizations.
15. References
• Jason Geis, Jeff Lang, Leslie Peterson, Francisco Roybal, Jenny Tanzillo, David
Thomas, David Warren, “Concurrent engineering of an infrared telescope system,”
Proc. SPIE, Vol. 8127, 81270L, 2011
– Demonstrates use of MBSE environment to generate a pre-Phase A level design of a low cost
IR telescope payload at a high level of engineering fidelity and for only 400 hours of
engineering labor for the payload design portion of the task.
• David A. Thomas, “Causes of catastrophic failure in complex systems,” Proc.
SPIE, Vol. 7796, 77960K, 2010
– Summarizes root causes of four famous catastrophic failures in complex systems as revealed
by post-mortem analyses of those failures and shows how concurrent design can be used to
mitigate similar failure mechanisms in future systems
• Jason Geis, Jeff Lang, Leslie Peterson, Francisco Roybal, David
Thomas, “Collaborative design and analysis of Electro-Optical Sensors,” Proc.
SPIE, Vol. 7427, 74270H, 2009
– Demonstrates use of MBSE environment to determine root cause of a focus anomaly in a
critical lens subassembly on a flight payload that was discovered during TVAC testing of the
hardware and not revealed by instrument contractor models. Saved an estimated 1 month of
schedule during instrument I&T.