Capella (once again) in space, meeting nanosatellites

Introduction
Please note your
microphone is muted
Click the Q&A icon and ask
your questions 10+ min
at the end of the presentation
Samuel ROCHET
samuel.rochet@obeosoft.com
www.obeosoft.com
Copyright
Obeo
@2022

Our Speaker
Jonathan LASALLE
Artal / Magellium
-
jonathan.lasalle@artal.fr
https://www.linkedin.com/in/johnlasalle/
-
https://capella.artal-group.com/
Copyright
Obeo
@2022

1 rue Ariane – 31520 Ramonville-Saint-Agne, France – 05 61 00 39 30 – artal@artal.fr
Jonathan Lasalle - ARTAL Technologies
Capella (once again) in space, meeting nanosatellites

The actors: CNES / Kinéis / Artal
 Capella expertise: training / coaching
 Capella customization
 Designs / manages complex space systems
 First satisfying Capella experience some months ago
 MBSE spreading in progress
 Strongly linked to the CNES
 Currently designs a new complex space system

 CNES/Artal first MBSE collaboration: discovering Capella
 Context
 SVOM: Space system dedicated to gamma ray detection
 Satellite with embedded sensors and ground segment (antenna and agencies)
 To be launched in 2023
 2018 to 2021 MBSE experimentation:
 Comparison: historical processes vs MBSE principles
 Capture of the system architecture using Capella
 Operational capture of system test
 Development of Capella extensions dedicated to V&V
History: introducing Capella inside CNES SVOM project

 Promising result
 MBSE/Capella is a real guide by structuring the work
 Non-ambiguous specification
 Better communication and process coverage
 Propagate these principles in another context
 Operationally
 Capella model being considered as single source of truth
 Generating less textual documents as possible
 On a similar… but maybe more complex... case
 Set-up of the Kinéis/Artal collaboration
History: introducing Capella inside CNES SVOM project
Click me !

 Kinéis
 Created in 2018
 Where NewSpace meets the IoT (Internet of Things)
 ~ 50 employees (+150% in 18 months)
 Initiated by:
 the CNES (French Space Agency)
 CLS (Collecte, Localisation, Satellites)
 Goal
 Democratization of Argos
technology
 Extend it to the entire
IoT market
Kinéis company

 Initiated in 1978
 Dedicated to studying, monitoring and
protecting our planet’s environment
 Structure
 Space segment: GPS and Argos satellites
 Ground segment: transmitters and process centers
 Data collection from various devices
around the world
 7 Argos Satellites
 22 000 active transmitters (8 000 dedicated to animal tracking)
 Over 100 countries
 3 Argos generation coexist: ARGOS 2, ARGOS 3 and ARGOS 4
 Revisit time: 1H30 / 2H (satellite connection period)
ARGOS system

 Kinéis main purpose:
 Extension of the ARGOS system to handle IoT principles
 Development, production and launch into orbit: 25 new nanosatellites
 Installation of 20 ground stations around the globe
 Revisit time => 5 min / 15 min (instead of 1H30 / 2H)
 Upgrade of the IT infrastructures
 Challenges:
 Design the new system
 Validate it
 Launch it
 MBSE process
 Use the Capella model as the (single) source of truth for system tests
Support the “regular” system tests need using Capella models
Kinéis main goal

Why using models instead of regular documents to describe the system ?
Reminder: principle benefits of MBSE/Capella approach
Communicate: use of a rigorous and reader-friendly language to
reduce ambiguities
Secure : validation of the specification using traceability and
coverage mechanism to ensure consistency, completeness …
Generate: take advantage of the formal description to
generate assets (and automate refinement steps)

 3 days training session:
 Go through all the concepts
of the Arcadia methodology
 Explore the main features of
Capella
 Provide guidelines and best
practices for the edition and
organization of the model
 Provide tips and tricks to
unleash the power of the tool
Step 1: learning Capella

 Coaching session to assist Kinéis
 Based on Logical Architecture layer
 Some “low-semantic” documents as inputs
combined with engineers knowledge
Step 2: model initialization (overall view)

 Two models were built in parallel:
 The main integration model (by Kineis)
 The model of one of the sub-component
(by Thales Alenia Space)
 Goals:
 The main model must integrate sub-model
“public interface” updates
 The main model must partially integrates
sub-items (to have a complete overall view)
 Challenges:
 How to link the two models ?
 How to “automatically” inject modifications
of the sub-model into the main one ?
(Iteratively ?)
Step 3: model reconciliation

 Solutions:
 Updates the models to reconnect “REC/RPL” mechanism
 Set traceability link “manually” in order to reconnect “system -> sub-system” mechanism
 Implement a specific “home-made” synchronization algorithm
 Implement a M2Doc template to easily compare “manual” synchronization
 Winner:
Step 3: model reconciliation
Integration model Sub-component model
Visual
comparison

 Capture functional chains:
 In order to highlight composite behaviors
 In order to identify validation objectives
 Entry point of the proposed process
defined during the CNES/Artal collaboration
 Limitation: hard to read/understand for
non-initiated people
 Create dedicated LAB
(Focusing on a given functional chain)
 Laborious to create manually
Step 4: focus on Functional Chain

 Implementation of a new tool:
 Import of functional chain content inside a LAB
 Synchronization of displayed item (in case of functional chain update)
Step 4 bis: assisted focus on Functional Chain

 Steps:
 Create a new LAB
 Select which components to display

 Steps:
 Select which functional chain(s) to “import”
 Read warning (if missing components)
 The user must explicitly/manually choose
which components to display

 Steps:
 Select which functional chain(s) to “import”
 Read warning (if missing components)
 The user must explicitly/manually choose
which components to display
 Then:
 Functions are automatically imported
 Unused component ports are automatically
hidden

 Application of the “system test” process
 Main steps:
 Capture the system architecture
 Capture validation needs
 Identification of the functional chains to validate
 Specification of corresponding tests (using exchange scenarios)
 Scenarios can be generated by a dedicated “home made” tool
Step 5: the system test process

 The CNES/SVOM process/tools not fully “compliant” with the Kinéis case
Step 5: generation of test scenarios

 New “sub-lifeline” generation feature
 Allows functions “superposition”
Clear/determinist reading
Step 5: generation of “readable” test scenarios

 Sometimes, test scenarios are “too long”
Step 5: generation of “simplified” test scenarios

 Sometimes, test scenarios are “too long”
 Injection of “sub-functional chain” support

 New mode
 3 functional chains
 5 functions
 Original mode
 17 functions

 In some complex cases,
it would be interesting to make scenarios even more compact
 Injection of a “compress” feature
Step 5: generation of “compact” scenarios

 Application of the “system test” process
 Main steps:
 Capture the system architecture
 Capture validation needs
 Identification of the functional chains to validate
 Specification of corresponding tests (using exchange scenarios)
 Scenarios can be generated by a dedicated “home made” tool
 Identification of user interactions and success criteria
Step 5: the system test process

 Inject test scenarios execution instructions
Step 6: user interaction and success criteria

 The edition of complex Capella scenarios can become laborious
 Several “graphical dependencies”
 Lot of impacts of moving an item
Step bonus: make scenarios more editable

 The edition of complex Capella scenarios can become laborious
 Several “graphical dependencies”
 Lot of impacts of moving an item
 Deletion of graphical constraints in generated scenarios
Step bonus: make scenarios more editable

 Review:
 1 year Kinéis / Artal collaboration
 Remaining 1 year before launch
 Kinéis have gained a new skill:
MBSE/Capella
 System tests capture in progress
 Based on Capella models (growing)
• 33 components / 112 component ports
• 200 functions / 500 function ports
• 33 functional chains
 Using a dedicated Capella extension
Conclusion

 Kinéis:
 Real facilitation of exchanges
 Functional chains: used as communication mean and system tests definition
• Huge gain for “under development” sub-systems
• Less useful for already developed sub-systems
 Scenarios:
• Necessary communication mean (for non initiated people)
• Satisfying system tests definition process
 Capella not easy to pick-up, assistance required
 Model reconciliations are laborious
 Essential to respect scheduling:
main view prior to sub-components in this case
Conclusion

 Artal:
 Support of Kineis activities
 Capella training and coaching sessions
 Development of Capella extensions to prevent limitations
 Next steps:
 Capture of missing tests scenarios
 Implementation of a bridge between
Capella and Jira RTM (Requirements & Test Management)
Conclusion

More information on
http://capella.artal-group.com

Q & A
Thanks for listening!
Do you have any questions?
Copyright
Obeo
@2022

Stay tuned!
▪ Capella Annual Message
by Juan Navas (Thales)
▪ System of Systems Modeling with Capella
by Tony Komar (Siemens)
March 15th, 2022
Don’t miss an opportunity to learn from Capella experts :
▪ March 21-28, 2022 | 2:00 pm - 5:30 pm UTC+1
An online training program in English, through 6 courses of 3.5 hours
each, from Monday to the following Monday.
February 17th, 2022
Contact sales@obeosoft.ca

Capella (once again) in space, meeting nanosatellites

More Related Content

What's hot (20)

Similar to Capella (once again) in space, meeting nanosatellites (20)

More from Obeo (20)

Recently uploaded (20)

Capella (once again) in space, meeting nanosatellites