The Executable Model Integration Bridge (EMIB): An Integration of Descriptive and Analytical Models April 3rd , 2014 Tamara Valinoto Systems Architect/Model Driven Engineering (MDE) Community of Practice(CoP) Chair tamara.valinoto@ngc.com Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Acknowledgements • Co-Authors: – Sola Olaode – Jessica Carleton • Previous Chair and Visionary – Sean McGervey 2 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. What is Model Based Engineering? MBE = MBSE + MDD + MBI&T Framework Collaboration Support MDE MDE COP Model Based SE COP Descriptive MBSE MBI&T Model Based I&T Perf Verification Model Driven Development Analytical MBSW/ MBHW (MDD) 3 3 Common MDE Framework Views Tools & Processes Artifacts / Products MBE includes Model-Based Systems Engineering, Model Driven Development, and Model Based Integration and Test Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Why Model Based Engineering? “One Fact, One Place” 4 4 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. What’s in a Name? Different Kinds of Models for Different Purposes Descriptive“Model” (as in “Model Airplane”) Blueprints, Schematics, Diagrams… Analytical “Model” (as in “Flight Model”) Computational Models, Simulations… 5 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. What is EMIB? Leveraging Phx Integration’s ModelCenter’s ® Out-ofthe-Box Integrations for MBSE • Enables Integration of Individual Analyses – In-House Code Descriptive Models Analytical Models – Commercial Programs Requirements • Analysis Server Client System – Reside on Computer or on Server – UNIX or Windows XP/7 Platform Parametric Modeling (SysML) Cost Bridge Analysis ModelCenter® • Create & Manipulate Models using drag and drop facility • Performs Trade & Optimization Studies, Probabilistic Analyses, Design Snapshots 6 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Matlab ® /Excel – Applications within MC Artisan Studio®/DOORS® – Excel Specifying Your System’s KSAs Isn’t Enough! How good is your design, and how do you know? bdd [Package] Demo Radar • Parametric models allow engineers to specify how their system’s attributes and constraints are related to one another • System models can relate: – KPPs/KSAs (e.g. size of the aircraft engine), to… – MOPs (e.g. maximum range of aircraft per sortie), to… – MOEs (e.g area under surveillance per day) «block» Radar X values frequency : Hz = 5.6e9 loss : dB = 6.5 noiseFigure : dB = 3.0 11 1 1 «block» Antenna values gain : dB = 45 mass : kg 1 1 1 1 «block» Transmitter values mass : kg = 5 power : W = 1.5e6 1 «block» Receiver values filterBandwidth : Hz = 5.0e6 mass : kg = 21 1 3 «block» LRU {Abstract} 2 7 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. «block» Processor 1 controller «block» Controller EMIB Facilitates Exploration of Design Space Common Set of Data Drives Analysis for Refining Design • With a tool-supported integrated model, engineers can vary the design and instantly see whether it will meet requirements • With the push of a button, KSAs, KPPs, and other design constraints can be sent from the system model in ArtisanStudio ® to PHX ModelCenter® for analysis • Results from the analytical models are then captured in the descriptive model as updated values and constraints for all the affected technical measures 8 2 1 Key Parameters for System are pulled from Artisan into ModelCenter ModelCenter orchestrates analytical models to generate trade study result Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Let’s Look at a Use Case to Analyze Impact of Constrain Requirement Change Nominal Activity Diagram 9 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Parametric Data Flow through Models Internal Block Diagram DOORS ® (Performance Requirement Model) Microsoft Excel ® (Cost Model) Output Cost Constraints Phx ModelCenter ® Performance Requirements Output Target Cost/ Design Constraints Input Design Parameters Output Target Design Constraints Matlab ® (Performance Model) Output Target Design Constraints ArtisanStudio ® (Parametric Model) Microsoft Excel ® (Dashboard Model) Input Design Parameters 10 Phx ModelCenter ® Plugin ArtisanStudio ® Plugin Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Streamlined Analysis for Solution Exploration Spreadsheet “Dashboard” Simplifies Analysis Workflow bdd [Package] Beam-on-Target Analysis Context «constraint» Beam-on-Target Model constraints {function of beam-on-target model} 1 bot «constraint» EA Beam Pattern Model constraints {function of beam pattern model parameters} parameters Avg_SLL : Real Az : Angle BW_Ratio : Real El : Angle f_max : Frequency f_per : Frequency HPBW_Az : Angle HPBW_El : Angle nx : Real ny : Real Peak_SLL : Real Scan_Loss : Real 1 1 1 «block» Beam-on-Target Analysis constraints bot : Beam-on-Target Model bpm : EA Beam Pattern Model cue : ES Update Quality Model rgpo : RGPO Model sarm : Seeker Flyout Model «block» System Context 1 parts systems ship : Ownship skr : Seeker 1 1 bpm 1 1 1 1 cue sarm «constraint» Seeker Flyout Model Update Spreadsheet with Latest Design Parameters Captured in Artisan Studio® System Model… «constraint» ES Update Quality Model constraints {function of seeker angle rate parameters} parameters Pos : Real PosGoal : Real PosSave : Real constraints {function of angular accuracy parameters} parameters EA_cue : EA Cue Data 1 … Then Run Your Analysis Using ModelCenter to Execute the Collection of Integrated Matlab, MathCAD, and Excel Analysis Models rgpo «constraint» RGPO Model constraints {function of RGPO technique parameters} parameters MP : Real PosGoal : Real 11 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Mockup of TPM Comparison Spreadsheet “Analysis Model” Linked to Phx ModelCenter ® TPM Comparison Design Parameter (from Artisan) Range Units Constraint Requirement (from DOORS) Nominal: 8.5 Range Units Pass / Fail Lbs. Pass m^3 Fail … … Minimum: 5.95 Minimum: 6.0 Weight_lbs Objective: 9.0 Lbs. Wt_lbs Maximum: 10.2 Maximum: 10.0 Dist: Uniform Dist: Uniform Nominal: 2100.0 Minimum: 1350.0 Minimum: 1400.0 Size Objective: 1695.0 m^3 Size Maximum: 2000.0 Maximum: 1950.0 Dist: Gaussian Dist: Gaussian … 12 … … … … Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Integrated Tool Suite for Robust Analysis Descriptive Model Feeds Analytical Thru Bridge par [block] Beam-on-Target Analysis: EA Beam Pattern Model sewip : ES-EA Suite ea : EA Az : Real Az : Real El : Real Avg_SLL : Real El : Real Peak_SLL : Real f_max : Real f_max : Real Scan_Loss : Real f_per : Real bpm : EA Beam Pattern Model f_per : Real BW_Ratio : Real HPBW_Az : Real nx : Real HPBW_El : Real ny : Real BW_Ratio : Real nx : Real ny : Real bdd [Package] Beam-on-Target Analysis Context «constraint» Beam-on-Target Model constraints {function of beam-on-target model} 1 bot «constraint» EA Beam Pattern Model constraints {function of beam pattern model parameters} parameters Avg_SLL : Real Az : Angle BW_Ratio : Real El : Angle f_max : Frequency f_per : Frequency HPBW_Az : Angle HPBW_El : Angle nx : Real ny : Real Peak_SLL : Real Scan_Loss : Real 1 1 1 «block» Beam-on-Target Analysis constraints bot : Beam-on-Target Model bpm : EA Beam Pattern Model cue : ES Update Quality Model rgpo : RGPO Model sarm : Seeker Flyout Model «block» System Context 1 parts systems ship : Ownship skr : Seeker 1 1 bpm 1 1 1 1 cue sarm «constraint» ES Update Quality Model «constraint» Seeker Flyout Model constraints {function of seeker angle rate parameters} parameters Pos : Real PosGoal : Real PosSave : Real constraints {function of angular accuracy parameters} parameters EA_cue : EA Cue Data 1 rgpo «constraint» RGPO Model constraints {function of RGPO technique parameters} parameters MP : Real PosGoal : Real Heterogeneous set of analytical models driven by a single design database 13 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Visualization of the Solution Space ModelCenter Provides Unparalleled View of Output Data Quickly see how much impact various system design parameters have on key technical measures MOE 14 KPPs Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Path Forward: Incorporate RF, Signal Processing, Firmware, and Software Analytical Models 15 Brings System Simulation into Alignment with MBE/MDE Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Capabilities are now available with Modelcenter® Phoenix Integration MBSE Pak and MBSE Analyzer 16 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Abstract The automated integration of descriptive models and analytical models has been sought by the systems engineering community to manage the effects of requirements change to reduce defects, cost, and schedule. The Executable Model Integration Bridge (EMIB) demonstrates that objective through the integration of requirements in IBM Rational DOORS® traced to architectures in Atego Artisan Studio® with design parameters that affect cost values in cost tools built in Microsoft Excel®, and performance analyses in MATLAB® models via a central automation and integration tool (Phoenix Integration ModelCenter®). Through EMIB, we can analyze the impact of changes to constraint requirements by developing descriptive models in Systems Modeling Language (SysML) with links to the system performance requirements. The descriptive model contains behavioral views that capture the flow of system and actor behavior required to satisfy the requirements, structural views to depict the elements of the system as blocks with current design and constraint values, and parametric views to capture the relationships between system element blocks and constraint elements. This integrated approach allows us to ingest the design values from an analytical tool (MATLAB®) into a descriptive tool (Atego Artisan Studio®) to provide early validation that the input (given) and output (target) design values are within the allowed ranges imposed by the constraint values (i.e. requirements). To demonstrate this functionality, a simple, top-level radar model was contrived and implemented whereby cost and performance results associated with material choices can be evaluated against customer specifications for emplacement/displacement times, weight, power, cost, and antenna performance in the descriptive model. An interactive model to achieve the aforementioned objectives was realized. 18 Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission. Backup 19 Acronym List 20 • EMIB: Executable Model Integration Bridge • KPP: Key Performance Parameter • KSA: Key System Attribute • MOE: Measure of Effectiveness • MOP: Measure of Performance • SWAP: Size, Weight, and Power • SysML: Systems Modeling Language • TPM: Technical Performance Measure • UML: Unified Modeling Language Copyright © 2014 Tamara Valinoto, Published and used by SEDC and affiliated societies with permission.