Data Exchange and MBSE Developments for Space Hans

Data Exchange and MBSE Developments for Space Hans Peter de Koning (ESA/ESTEC, Noordwijk, The Netherlands) Harald Eisenmann (EADS/Astrium Satellites, Friedrichshafen, Germany) 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 1

European Space Agency • Intergovernmental organisation – 18 member states – Annual budget around 3 billion € – Approximately 2000 staff • http: //www. esa. int 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 2

Topics • Moving towards Model Based Systems Engineering – Standardization • INCOSE & OMG: Sys. ML • ECSS: "Engineering Database". . . – R&D • Studies to validate standardization approach • Virtual Spacecraft Design • Space Thermal Analysis Model Exchange – Status STEP-TAS implementations • Geometrical models and space mission apects • Thermal test results – New development • Conversion between ESATAN, SINDA/FLUINT, SINDA/G 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 3

What is Model Based Systems Engineering? (INCOSE definition) • Model-based systems engineering (MBSE) is the formalized application of modeling to support system requirements, design, analysis, verification and validation activities beginning in the conceptual design phase and continuing throughout development and later life cycle phases "INCOSE SE Vision 2020" INCOSE-TP-2004 -02 Sep 2007 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 4

European space industry Current approach and status • Space system development process – Overall system engineering process well established – applied on all ESA projects • Standard: ECSS-E-ST-10 -1 C "System engineering general requirements" – All domain (discipline-specific) processes are well established – applied on all ESA projects • Avionics, power, structures, mechanisms, thermal control, propulsion, optics, software (onboard, groundstation), communication, control (orbit & attitude, robotics), • Project management, cost, risk, logistics, product assurance, … • Each with own ECSS standard • Tools – – Domain disciplines are well supported in terms of engineering and analysis tools For some disciplines there is excellent integration of design and analysis tools Lack of support for system engineering: System level design is typically defined with office tools System level design is “scattered” in different tools – with little traceability 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 5

Industrial Practice Limited Model Based Capabilities • Current industrial practice: – Many elaborated tools (analysis) and databases in use – Pragmatic bottom-up integration of particular tool chains – Many ad hoc databases • System level design status – by system modeling – – Very good representation in phase 0/A (parametric models – standardized set of properties) Requirements specifications partly in databases (RE Tools), partly in documents, some traceability Very good representation for some disciplines (e. g. MCAD) Phase B representation often kept in MS-Office products • System modeling is often driven bottom-up (e. g. flight software or simulation) 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 6

Trend: System Engineering "V" with Model-based Validation & Verification Top-Down Design Phase B System / Preliminary Design Validation Deployment V V& rly Conceptual Design Phase E Verification Ea Phase A Ea Mission Need Feasibility Study Bottom-Up Production, V&V Integration Executable Model-based Validation & Verification Assembly Phase D Manufacturing Detailed Design Phase C (Development & Qualification) 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 7

Model Based (System) Engineering Expected benefits • Systematic application of models for specification, design, verification and validation – Both for structure and behavior of the system – Managed traceability between all aspects and integrated version/configuration control • Expected benefits of consistent application of models – As much as possible single source data – generate/transform where possible – Allows computer aided completeness, consistency and quality checking – Improved collaboration between disciplines and organisations • Multidisciplinary teams • Efficient and reliable communication through supply chain – Eases development iterations and change impact analysis – Enables and promotes knowledge capture and (re-)use – Enables multiple, concurrent views (subsets, presentations) on same data – Reduced “manual” work and thus improve cost, schedule and quality / reduce risk 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 8

Ultimate Goal – MBSE • Comprehensive system level design representation • Seamless data traceability from system level design into domain engineering design • Consistent “transformation” from system design model into analysis (executable) model – For domain analysis models – System simulators • Consistent representation of the decision making process (analysis design verification) • Continuous verification / validation based on consistent application of virtual models 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 9

Enabling Technologies • Overall Integration Framework required – For efficient tool integration – For data exchange and sharing – To ensure common and mappable semantics • Specification of the required data – In a formal way on a conceptual level – ref. OMG/MDA PIM (Platform Independent Model) – Enabling subsequent “generation” of required implementation technology • System level design editors – Definition, modification and presentation of engineering data – Capturing properties and values • Supported by open standards – Cannot and want not standardize on tools 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 10

Supporting standardization and R&D • ECSS-E-TM-10 -23 "Engineering Database" – – – Technical Memorandum evolving to a future standard Conceptual data model defining precise common semantics MBSE reference architecture – using OMG/MDA principles Re-use from STEP-NRF: formalized representation of quantities, units, dimensions, … As much as possible aligned with Sys. ML • ESA funded R&D – Space System Reference Model • Prototyping of centralized services as integration platform – validation of E-TM-10 -23 – Virtual Spacecraft Design • Demonstration of a space system engineering process relying on a virtual representation – Design Resource Center • Design data library to ease access 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 11

ECSS = European Cooperation for Space Standardization • Initiative started in 1995 to produce a coherent, single set of user-friendly standards for use in all European space activities • Previously there were different ESA and national space standards • Partners are space industry (via the Eurospace association), ESA and national space agencies • Published standards freely available from http: //www. ecss. nl in PDF (after a simple registration procedure) • Complete update per March 2009 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 12

ECSS E-TM-10 -23 "Engineering database" • Conceptual data model standard • Purpose: to provide common semantic reference for all engineering data repositories needed to develop and operate a space system through the whole life cycle • Defined using UML 2 (class diagram) – Looking at simultaneous specification in ontology language (OWL, ORM) and EXPRESS • First release imminent – Expected May/June 2009 – Will be published in hyperlinked wiki form to allow efficient review and evolution (generated from Open. AMEOS UML tool) 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 13

ECSS E-TM-10 -23 "Engineering database" 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 14

ECSS E-TM-10 -23 Adoption of the ECSS global conceptual data model for a project 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 15

ECSS E-TM-10 -25 ESTEC CDF • Technical Memorandum to agree the exchange of conceptual design models (Phase 0 / A) between concurrent design facilities • Aligned with E-TM-10 -23 (subset) • Agreed list of reference data – List of design parameters with responsible discipline (~600 parameters) – List of recognised disciplines – Basic system decomposition • First release expected May/June 2009 • Supported by OCDS (Open Concurrent Design Server) replacing Excel based infrastructure 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 16

Space System Reference Model ATV Tools ATV System Engineering Control GUI SSRM Tools SSRM Tool Connectors Control WS Basic WS Business Logic Engineering WS SSRDB Connector Integration Platform SSRM Database 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 17

Space System Reference Model • Objective – Prototype physical and semantic integration platform – Validation of draft E-TM-10 -23 data model w. r. t. STEP, MDA compatibility – Demonstration based on real industrial scenarios • Result – – Centralized webserver hosts shared applications like: adaptors, core application, version control, … Persistent storage of data in RDBMS Control applications MDA based development framework • Status – Study successfully finished in May 2008 – Results will be evolved in Virtual Spacecraft Design 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 18

Design Resource Center • Objective – Development of design data library – Definition of equipment properties (used for design) • Result – Web-server application for the core application – HTML front-end integrated in ESA portal – Underlying relational database – Linkage to ESA DMS 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 19

Virtual Spacecraft Design • Objectives – Demonstration of a model-based system engineering process using a full virtual representation of the space system • Status - End phase 1 Requirements analysis and conceptual design – Overall process analysis performed – Use case analysis and user requirements specification performed – Overall architecture defined lower level elements identified for prototyping • Next step – Phase 2 realization and demonstration (2009 – 2010) – – Space System Design Editor Space System Reference Database Space System Visualization Tool Builds further on E-TM-10 -23, SSRM centralized integration platform, MDA for design editors, STEP-NRF 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 20

Virtual Spacecraft Design 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 21

Overview Data Bases Supplier Handling of Data via Eng. Tool Input by domain Administrators ASTRIUM S/C Eng. Reference SDB (TM/TC) SEDB MCAD DB (CAD Model) CC DB = SCC DB Space System Engineering Repository EPPD Equip. Physical Props. Sim. DB MCAD ECAD Tool: CATIA Tool: Eng. Base Tool: Sim. DB Unit/ Equipment. SDB SRDB AP DB SCOE TM/TC DB Customer Tool: MOIS FDDB AIT SDB OBSW SDE AOCS OSE Tool: Matlab/ Simulink Parts/ Struct. / Harness Manufact. El. Test Tool IDAS DB Tool: IDAS Control Console MDVE Simulator 11 th EGSE DB DB Tool: Sim. Ops/ NASA-ESA Workshop Open. Center on Product Data Exchange 29 April - 1 May 2009 Development & Verification Tools 22

Example validation of E-TM-10 -23 in prototype space system design editor Eclipse (Ganymede) based tool Using GMF and EMF Code generated from E-TM-10 -23 data model using MDA approach Developed by Scope. SET (Germany) 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 23

Example validation of E-TM-10 -23 in prototype space system design editor 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 24

Support to Sys. ML • Support Sys. ML RTF 1. 2 and 2. 0 work through INCOSE/OMG – Upgrade of Quantities, Units and Dimensions – foundation for Sys. ML "value properties" • Now based on International Vocabulary of Metrology by BIPM • Combined concepts and lessons learned from ISO 10303 -41, STEP-NRF, MARTE, ontologies • Fully supports multiple quantities and units systems • Using new ISO/IEC 80000 standard (harmonised replacement of SI) as main reference – Integration of Sys. ML Parametrics and Modelica • Modelica is a neutral language to define object oriented non-causal simulation models see http: //www. modelica. org • Powerful definition of executable (high level to detailed) system models – Mathematical equations relating properties at ports of blocks – With or without predefined causality (order of variable assignment or expression evaluation) • Fits well with Sys. ML contraint blocks 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 25

STEP-TAS progress 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 Sheet 26

STEP-TAS Activities in 2008 • IITAS – Industrial Implementation of STEP-TAS – in progress – ESA completed full test suite with automation tools • TASverter by ESA TEC-MTV (Thermal analysis and verification section) – Now more than 150 different users (2~5 downloads per week) – Routine use in many projects – Under maintenance – but very few bugs reported • • Evolution of Expressik to support code generators STEP EXPRESS to C++ and Python Evolution mapping STEP data into HDF 5 format Completed ESATAP v 1 thermal analysis post-processor using STEP-TAS in HDF 5 format First validation of STEP-TAS Kinematics and Mission Aspects (CC 2, CC 4, CC 5, CC 6) – Implementation in TASverter for ESARAD • Proof of concept implementation in Dyna. Works® for import of STEP-TAS analysis predictions 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 27

In progress 2009 • Started IITAS-TMG activity with Maya (Canada) – Implementation of STEP-TAS import/export into TMG • Completion of IITAS and IITAS-TMG – Emphasis on testing and obtaining robustness of imports/exports • Full validation of STEP-TAS Kinematics and Mission Aspects • Implementation of full ESATAN / SINDA conversion in STEP-TAS and TASverter • Formalisation of STEP-NRF/TAS under ISO TC 184/SC 4 – Was planned for 2008 but put on-hold due to lack of resources – shifted to 2009 • R&D "Innovative methods for improved thermal testing" – o. a. using STEP-TAS to interface between analysis prediction and test results • Consolidate support software and test suites as true open source software – Depending on ESA open source software policy that is currently being finalised 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 28

ESATAN / SINDA conversion using STEP-TAS • ESATAN, SINDA/FLUINT, SINDA/G are major space thermal analysis tools • Old style FORTRAN finite difference / lumped parameter solvers – Common heritage to CINDA which was developed in 1960 s • Used on many space projects worldwide • Model defined in two parts: DATA and OPERATION blocks – DATA blocks define model structure: nodes, conductors, boundary and initial conditions, . . . – OPERATION blocks define behaviour simulation: user defined logic and solver calls • Quite sophisticated modelling capabilities for active thermal control: thermostats, PID controllers, heatpipes, fluid loops, evaporators/condensors, Peltier elements, mission (power) mode switching, model parameterization, non-linear material properties, . . . • Exchange of DATA blocks is supported by existing tool-to-tool converters – But incomplete and not reliable • Exchange of OPERATION blocks is manual: very complicated and time consuming 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 29

Data exchange implementation ESATAN / SINDA OPERATION blocks • Use ANTLR to parse ESATAN or SINDA MORTRAN language • Store resulting abstract syntax tree in prefix notation using JSON ANTLR (http: //www. antlr. org) is a very powerful OSS lexer/parser package. JSON (http: //www. json. org) is Java. Script Object Notation, a very powerful but simple serialization format supported by OSS in almost all programming languages. – For expressions we use the Content Math. ML approach • Only not in XML but JSON encoding • e. g. "X = 4. 0 * (2. 0 + 3. 5)" → ["assign", "X", ["times", 4. 0, ["plus", 2. 0, 3. 5]]] – For algorithms (sequence of statements, if-else blocks, loops, etc. ) we use simple keywords • e. g. ["loop_for", loop_var, start_expression, end_expression, step_expression, statement_block] • Integrated into STEP-TAS through EXPRESS ENTITY instances – nrf_algorithmic_language, nrf_algorithmic_expression, nrf_algorithmic_statement – Both in neutral JSON language and original ESATAN or SINDA language • • Prototypes work very well Perhaps interesting light approach also for other algorithm conversion need Upgrade of TASverter expected later this year Development contracted to DOREA (France) 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 30

References • European Space Agency http: //www. esa. int • Astrium Satellites http: //www. astrium. eads. net • European Cooperation for Space Standardization http: //www. ecss. nl • Sys. ML – Systems Modeling Language http: //www. omg. sysml. org • STEP-NRF and STEP-TAS http: //www. esa. int/thermalcontrol Look for "Standards” • TASverter https: //exchange. esa. int Look for "TASverter“ • ISO TC 184 / SC 4 standardization committee (a. o. STEP standards) http: //www. tc 184 -sc 4. org 11 th NASA-ESA Workshop on Product Data Exchange 29 April - 1 May 2009 31