The 13th ESA Workshop on Avionics, Data, Control and Software Systems (ADCSS) covers topics related to avionics for space applications, in the form of a set of round tables. The workshop acts as a forum for the presentation of position papers followed by discussion and interaction between ESA and Industry and between participants. Each theme part of ADCSS workshops will be first introduced and then expanded by presentations on related developments from technical and programmatic points of view. A round table discussion may follow, concluded by a synthesis outlining further actions and roadmaps for potential inclusion into ESA’s technology R&D plans.
Attendance to the workshop is free of charge. Registration is required via this website not later than November 08th, 2019.
All material presented at the workshop must, before submission, be cleared of any restrictions preventing it from being published on the ADCSS website.
The presentation will introduce the Space Avionics Open Interface (SAVOIR) initiative, the main outputs, the status on the work plan, and the trends, as an introduction to the following working group specific presentations.
Fault, Detection, isolation, Recovery has been subject to many issues in ESA reviews. There is no common vocabulary, no identified specific process, no agreed set of output. In order to establish FDIR as a discipline, and to agree on a common customers-suppliers approach, the FDIR working group aimed at producing a handbook. The document under publishing and the main results will be presented.
The SAVOIR-MASAIS Working Group aims at defining the functional, performance, operational and interface requirements of the Avionics System Reference Architecture Data Storage function and the definition of standard services and interfaces to manage data on-board. The status of this Working Group will be presented.
The SAVOIR-UNION Working Group aims at defining the functional, performance and interface requirements of the Avionics System Reference Architecture functional links. The status of the Working Group will be presented.
The OSRA Network Communication Specification aims to provide a set of high-level, generic requirements for modern and forthcoming avionic systems in a technology-agnostic way. Most of the currently available network technologies (e.g., SpaceWire, Ethernet, etc...) require complex upper layer protocols to fulfil the OSRA-NET requirements. The goal of this work is to assess the compatibility of link solutions with OSRA-NET, with a special focus on the Quality of Service and the Fault Detection and Isolation and Recovery (FDIR) requirements.
The SAVOIR on-board software architecture (OSRA) is now quite established, with a defined meta model and a supporting editor. Several technical documents are also describing the architecture and specifying the functions of the Execution Platform, this part of the on-board software that supports the application software for communication and real-time. The technical documents are gathered into a set of SAVOIR documents, including a generic specification of the execution platform. The presentation will show the status of the work.
Following the interest expressed by the SAVOIR Advisory Group, it has been decided to establish a dedicated working group on the automatic code generation process focused on AOCS Flight SW in order to prepare a handbook with the process guidelines. The presentation will introduce the objective of the working group, as well as the Table Of Content of the current SAVOIR Handbook.
Supported by a coming R&D activity on “Implementation of Attitude and Orbit Control Interface Commonalities”, this working group will aim at producing a handbook, in view of harmonisation of interface on board (Attitude and Orbit Control), with potential benefit on ground (Flight Dynamics) (e.g. list of common algorithms).
The presentation will introduce the coming Delay/Disruption Tolerant Network set of protocols and standards that allow, together with the CFDP, to achieve reliable communications for programs such as ISS, Epoxy, NASA lunar OISL, and the Gateway.
Mr HULT will give an overview and highlights of his avionics carreer at RUAG, as well as his substantial contribution to the SAVOIR working group
The CCSDS is developing a standard set of communication services to support the on-board applications, which are in turn supported by communication protocols. The idea of “plug and play” avionics units is a major goal while enabling use of standard building blocks and reuse. A major step in this direction is through the definition of Electronic Data Sheets (EDS). This concept is now taken up at system level, where it could become the corner stone of system data exchange
The presentation will address the status of the international standardisation, the activities in ESA and a view of future steps.
The presentation will recall the need for a generic OIRD as a way to reduce the variability of our avionics systems concerning operability. It will present the first published document, as well as the plan for its applicability.
This session, organized by the CNES, will highlight the impacts of the advanced Highly Integrated Avionics architecture physical implementation and technology on the SAVOIR specification.
In the frame of CNES R&D program, a new generic processing board based on the NanoXplore NG-Large SoC is being developed.
This board, evolution of the itar-free CPUGEN board commercialized by EREMS, will see the “Gaisler GR712 processor + ATMEL AT280 FPGA” couple replaced by the NG-Large SoC for more power efficiency and more versatility.
The form factor will be adapted to match the space VPX format while trying to still cope with the ITAR free requirement. The board will benefit from high speed link to connect with external high throughput equipment or accelerator board.
Artificial Intelligence (AI) is another one tool that humans need (among the others) for their lust for survival, comfort, improvement of quality of life, and reduction of tedious mental labour. But lots of human activities and efforts for making AI deviate a lot from this idea. This is due to the difficulty of understanding, grasping, and combining fundamental concepts from different fields of science and philosophy. This presentation clears this confusion and sets properly the pieces on the chessboard. It will give answers to fundamental questions and provide a different perspective of this beautiful scientific endeavour.
CloudScout is a project financed by the European Space Agency and run by a European consortium led by cosine Remote Sensing, participated by University of Pisa, Sinergise and Ubotica technologies. The project targets the development, and for the first time ever the in-orbit demonstration, of an Artificial Intelligence based algorithm for cloud detection.
The inference is applied on images acquired by HyperScout-2, a very compact and powerful hyperspectral/thermal imager which embarks an unprecedented processing capabilities.
The specification of HyperScout-2, with its very wide FOV (31deg) and small GSD (75 m in the VNIR and 390 m in TIR), as well as the large number of spectral bands (45 in VNIR and 4 in TIR), provides very powerful capabilities for a large number of applications. However, its management at system/mission level turns to be a challenging task, especially for what concerns the data management and download. For an effective use of the instrument potentiality the information processed from the Tera Bytes of data generated needs to reach the final user in a reasonable time. This pushes the mission to the edge of feasibility, being a very challenging task even for bigger classes of satellites.
In order to maximize the quality of the information included in the downloaded data, HyperScout-2 uses its on board processing capability without relying on platform subsystems. HyperScout-2 is in fact equipped with an hybrid processing platform composed of FPGA, CPU, GPU and VPU. The VPU is a state of art Vision Processing Unit developed by Intel and will be flown in space for the first time as part of HyperScout-2. Thus, HyperScout-2 will enable experimental programs to investigate the use of Artificial Intelligence (AI) for a variety of applications in the field of object detection and data inference.
This paper presents the first application that will fly on the ESA mission PhiSat-1 / FSSCAT, part of the ESA EOP initiative to leverage small satellites to foster technology breakthrough developments.
Machine Learning (ML) has been proven to be a promising tool for researchers in many
fields and disciplines. However, the barrier to entry has been too high as the time it takes to
become proficient enough on ML frameworks can be quite long. Yet domain experts, rather
than machine learning engineers, always have the best insight into solving problems. Fast.ai
abstracts away much of the difficulty in building quality ML models, making it easier for
domain experts to apply it to their own work. Here we demonstrate how Fast.ai works. We talk about its accuracy/reliability and footprint in both memory and FLOPS when we repeated some research papers, getting superior results with only a few lines of code
In the scope of the GSTP program, Fraunhofer INT will perform radiation tests and functional verification tests on COTS components under contract of the European Space Agency. It's key points are twofold: First it's to identify and test available COTS components whose functionality or properties would greatly benefit the space community. This will make use of the so-called spin-in approach. The project is currently in the first phase of test candidate identification. The basic suitability is then checked in a preliminary screening campaign and components that pass this will be subjected to in-depth testing. Suitable devices will be recommended for further qualification tests beyond radiation to enable use in space. The second key point is to assess test methods and methodology for COTS components in general, such that the data and experience can be used beyond the qualification of the specific parts but can rather lead to guidelines for future COTS testing.
Spacecraft Platforms avionics and Payload systems currently rely on three main datalink standards for covering the on-board communication: 1553, Can Bus and SpaceWire. However, none of these technologies provide an efficient and reliable solution for future applications requiring more that 1Mbps of data transfer rate together with stronger real-time and quality of service requirements. For Ariane6, Orion and other future manned flight and in-orbit space vehicles, an Ethernet based solution with the Time-Triggered protocol (TTE) has been selected and is currently being developed. The current development of Time Sensitive Network (TSN) standard protocol on Ethernet is an opportunity for providing future on-board communication solutions fitting needs of future space applications while allowing synergy with other markets with the expectation of high performance and more affordable space avionics products based on COTS building blocks and standard.
This presentation will provide the rationale for Airbus and other space industrial partners to evaluate and prepare a roadmap on Ethernet TSN technology for being use on future space applications. An overview of in-going relevant preparatory studies and developments with agencies and other partners will be provided.
This presentation will focus on the system aspects of TSN adoption in future satellites and which specific challenges will have to be met to enable such adoption. It will also present which specific advantages are brought by TSN with respects to existing space communication technologies
Since some years, worldwide, an effort has been made to find a new candidate for time critical application busses and the ‘retiring’ of the MIL-BUS-1553 has been in some way triggered. While several candidates have showed up, a ‘universal heir’ has not yet being found.
In order to add a piece to this complex puzzle, GMV and Seven Solutions would like to present their work about Deterministic GEthernet network.
An IPCORE based on a TSN light implementation, together with RTEMS drivers, has been designed and developed by Seven Solutions upon system requirements and validation methods provided by GMV.
A first use case has been identified: MIURA 1 sounding rocket, whose mission is to provide microgravity environment to payload experiments but also to provide a flying test bed for technologies that will fly with MIURA 5 micro launcher.
One of the innovations in the new generation of European launcher is the use of the Time
Triggered Ethernet for the communication system. This presentation will provide a feedback
from Ariane 6 experience with a focus on the communication controller configuration. Especially,
constraints regarding the schedule and the synchronization topics will be discussed.
RISC-V is an Instruction Set Architecture (ISA) that is rapidly growing in popularity in terrestrial applications. Its modularity and openness allow to implement the optimal microarchitecture for a wide range of applications in satellite data systems, ranging from microcontrollers to processors for Artificial Intelligence. Several open-source RISC-V processors are already available for terrestrial applications. However, there is still work to be done to employ them in satellite data systems, especially in terms of Fault Tolerance and Technology Readiness Level.
The presentation will provide an overview of the work performed in the ITI activity Introduction of Fault-Tolerant Concepts for RISC-V in Space Applications. ARIES
Aerospace Research and Innovation in Electronic
Systems) part of Nebrija University (ES), Cobham Gaisler (SE), and QinetiQ Space nv (BE) have evaluated the state of the RISC-V ecosystem, with focus on processor core implementations, and how RISC-V can be applied for use in European space. A processor implementation (Rocket from UC Berkley) was selected and integrated in a system representative of contemporary European space-grade system-on-chip designs, creating a FPGA demonstrator design implemented on Kintex Ultrascale. An end user
evaluation has been performed of the system in order to compare the specific RISC-V system to current LEON implementations from a functional and performance perspective and different solutions to mitigate radiation effects have been evaluated using error injection.
We introduce the first RISC-V orbital laboratory for microarchitecture fault-tolerance evaluation. The project builds on the space-qualified PULPino-compatible processor cores, belonging to the Klessydra core family, preliminarily introduced at the June 2019 RISC-V workshop. In small satellites, the use of Commercial Off-the-Shelf (COTS) components is increasingly interesting, along with the adoption of an open and extendable instruction set architecture like RISC-V. Fault-tolerant architecture techniques are essential, since COTS components are not intrinsically protected from the harsh space environment at physical level. Furthermore, the possibility of soft-cores implemented in FPGAs opens the way to dynamically adapting the system to harsh environment effects. The presentation addresses an extensive set of fault tolerance techniques, explicitly tailored to RISC-V and leveraging the inherent multi-threaded execution of the Klessydra microarchitecture. In addition to simulation-based and in-lab hardware-based fault-injection, the RISC-V Klessydra Orbital Lab (RV-KOL) is being set up to dynamically analyse the behaviour of the different techniques in a real space environment. The orbital lab will be the payload board launched on a space satellite, designed to be configured “Over-the-Air” (OTA). It can communicate through several standard interfaces and can run processing routines on data coming from on-board sensors. The launch of the PocketQube satellite with RV-KOL payload is expected in mid 2020, through the new European launch system VEGA-C.
The contribution will describe a new floating-point unit, and new integer instructions that have been implemented in LEON2-FT. The new integer instructions implement SIMD-within-a-register operations that are beneficial for speeding-up GNSS/SDR as well as cryptographic or image processing workloads. The FPU can be configured to act as a drop-in replacement of the former Meiko FPU that is no longer available, or to support just selected operations in user-defined precision (e.g. FP16 or bfloat16), with other operations and/or precisions computed in software. The main goal of these extensions is to increase the potential of LEON2-FT for current and future on-board data processing.
There has been a directed move over the last decade towards Interoperability, Portability, and the heavy use of Open and Published Standards and Open Architectures that, in turn, allow for those terms listed above. This is a major shift in paradigms that outlines a new Modular world with open and well defined logical and physical interfaces to facilitate a new method by which we truly architect systems for reuse and efficient tech refresh. The end result are logical and physical building blocks that allow for that reuse and reconfiguration.
Introduction to the model-based software engineering discipline, and its relationship to systems engineering. Presentation of the various past, present and future ESA activities related to that domain and its challenges from ESA point of view.
The first part of the talk will cover the principles of equation-based object-oriented modelling and simulation, the Modelica Language, and the generation of executable simulation code out of Modelica code (and other uses thereof). It will also address standards related to Modelica: FMI, SSP, DCP and show its application on the Modelling and Control of Satellite Attitude Control Systems. The second part of the talk will address the community and governance aspects, including the Modelica Association, its projects, in particular OpenModelica and the Open Source Modelica Consortium.
Francesco Casella got his master's and PhD degrees from Politecnico di Milano, where he is currently senior assistant professor at the Department of Electronics, Information and Bioengineering. He teaches graduate-level courses in Control Systems, Advanced Process Control, and Object-Oriented Modelling and Simulation at Politecnico di Milano; he is also guest co-teacher of a graduate course on Modelling, Simulation and Application of Propulsion and Power Systems at TU Delft. His main research interests are modelling and control of power generation systems, and equation-based, object-oriented modelling of engineering systems in general; he has published over 100 papers in peer-reviewed conferences and journals on this topic. He is member of the board of the Modelica Association and vice-director of the Open Source Modelica Consortium and has been General Chair and Program Chair of the International Modelica Conference in 2009 and 2017.
In this presentation, you will get feedback on ArianeGroup Capella tool deployment on Ariane 6 Program and will learn how ArianeGroup is meeting the challenges of multi-disciplinary system engineering. You will also discover the ArianeGroup next steps for MBSE and how the implementation of this process based on Capella is reinforced.
Clément Grise is a MBSE Project Leader at ArianeGroup. After graduation from the ESTACA Engineering School in 2006, he joined PSA Peugeot Citroën, the French automotive car manufacturer, in powertrain after-treatment Innovation department to develop the SCR (Diesel DeNOx system) algorithms. In 2012, he took the lead of the powertrain control innovation team and became leader of the Powertrain After-treatment Control development team in 2016. He joined ArianeGroup in 2017 to develop and deploy MBSE in the company.
Capella is an open-source and field-proven Model-Based Systems Engineering (MBSE) solution to successfully design systems architecture. It provides systems architects with rich methodological guidance relying on Arcadia, a comprehensive model-based engineering method based on both industrial experimentations and system engineers' feedback.
One of Capella's key attributes is its open source nature. This presentation will highlight the specificities of this model and will make out the state of the “community”. It will give an overview of the growing community supporting Capella. From Thales' initial decision to open-source the solution to its adoption by many other companies as an MBSE backbone. Operating rules and governance will be described as well as the many ways for an end user as the ESA to get involved and to take advantage of it.
Samuel Rochet is an Arcadia/Capella evangelist and an Open Source advocate at Obeo. He is also in charge of the Capella business development. Samuel has more than a dozen years of experience in the field of systems engineering and MBSE. Since a PhD dedicated to system engineering processes and modeling technologies, he has been continuously engaged in this field. He notably worked 5 years on behalf of Airbus as Method & Tools consultant supporting the deployment of MBSE projects before joining Obeo, an open source and model-driven software vendor. Since 2013, he’s been strongly involved in the Capella Community being one of the leading contributors in Capella adoption and its ecosystem development.
"After OMG issued the Request for Proposal (RFP) for SysML v2 on 24 Jan 2018 (https://www.omg.org/news/releases/pr2018/01-24-18.htm) a team of more than 100 experts from around 55 organisations -- the so-called SysML v2 Submission Team (SST) -- has started to develop the second version of the Systems Modeling Language standard (SysML).
In order to develop the RFP itself a working group analysed the strenghts and weaknesses of SysML v1 after use in industry for aproximately 10 years since its initial release in 2006. There are actually two RFPs and two specifications under development: (1) The SysML v2 language specification, (2) The SysML v2 Application Programming Interface (API). Version 2 is a complete redesign with the following aims:
1. Clean up the SysML meta-model and founding on a solid semantic basis;
2. Reduce the learning curve for systems engineers (who in general do not have a software background);
3. Fully harmonise the way decomposition of structure and behaviour is done;
4. In addition to XMI file based exchange, create a robust open standardised application programming interface, allowing for dynamic fine grained access to SysML model repositories;
5. Ensure a smooth migration from SysML v1.x, include an additional SysML v2 profile on top of the existing UML v2.5 standard.
6. Continuously validate the new specifications with a full implementation prototype.
In this presentation progress to date will be reported and explained, on the basis of the first intermediate public release that was made in October 2019. The currently scheduled release date of SysML v2 is June 2020. The author was a member of the RFP working group and is an active member of the SST.
This presentation will discuss the main challenges related to Autonomy and FDIR for Space Missions in a changing environment: constellations, debris management, complex integrated equipment, off-the-shelf equipment, increased availability requirements, human safety. AOCS/GNC functions are required to answer at the same time very high performance requirements and high resilience to ensure Spacecraft and Mission safety: robustness to radiation-induced effects (e.g. SEU), robustness to uncertain environment in critical phases (e.g. EDL, RDV), and robustness to degraded configurations (e.g. after multiple failures). Trade-offs and potential solutions currently developed by Airbus Defence and Space will then be presented:
- AOCS/FDIR Strategies: the different ways to implement Fail-Operational concept after critical failure (e.g. hot redundancy, warm restart, autonomous mission resume after restart, …), the different Safe Mode Strategies to improve Mission recovery (e.g. alternatives to Sun pointed attitude in Safe Mode, Star-Tracker based Safe Mode, autonomous Normal Mode entry and mission continuation, …)
- AOCS/FDIR Failure Detection improvements: technological vs functional monitoring, use of Artificial Intelligence techniques, independent safety monitoring, …
- AOCS/FDIR Engineering process: Generic AOCS/GNC Techniques & Design Framework for FDIR, Model-Based Engineering, handbooks, key principles and best practices,
The goal of this presentation is to present the state-of-the-art and perspectives of autonomy and GNC/FDIR coupling on current and near-future missions. We will clarify the need for increased autonomy and the main development axes identified to optimise the future missions.
This presentation will focus on:
· The novel FDIR handbook and its links toward GNC/FDIR coupling validation
· The needs for autonomy on current and near-future missions including robotic exploration
· A focus on ExoMars autonomous FDIR and specific recoveries to ensure the mission continuity in critical phases
· The potential benefits in mission operations and performance through highly autonomous systems
In our latest ESA activity, a flight test campaign was carried out in a representative environment of a Mars Landing for the verification of the Hybrid Camera-Lidar Hazard Detection and Avoidance System (H2DAS). For that purpose, a large multicopter platform was developed to carry aloft the Avionics Test Bench (ATB) experiment.
The multicopter-based flight test platform (FTP) can execute 15min autonomous flights while carrying an avionics payload of 12kg. The FTP functions independently from the ATB, with its own instrumentation suite that enables autonomous flights and provides a source of accurate navigation data for using as a reference when analysing and discussing the results of the avionics experiment.
The ATB integrates a Mars landing-representative sensor suite: an IMU, a laser altimeter, a visual camera and a LiDAR. The sensors are connected to both a space-representative processing unit (a soft-core LEON2 and an RTAX-2000 FPGA, for the H2DAS software) and a COTS-based Zynq-7020 board with dual-core ARM A9 CPUs and an FPGA (to accelerate the most computationally intensive tasks of Vision-Based Navigation).
The hardware-accelerated real-time software implementations of the Visual Based Navigation and Hazard Detection and Avoidance (VN&HDA) were first tested and verified in a processor-in-the-loop (PiL) facility.
Both the flight test platform and VN&HDA package were tested on the ground, and then validated individually in flight. Finally, the activity proceeded to carry out an extensive flight test campaign in one of the largest purpose-built Mars analogue terrains in Europe (>15000m2) at a quarry in Portugal.
Flight tests were also in a sufficiently high number (>650 tests) as to demonstrate a statistically significant result, indicating a safe site selection probability of >99%.
Current and future trends for launch vehicles include increasingly stringent requirements for mission responsiveness and adaptability. In a previous FLPP activity we developed a tool to analyze state-of-the-art approaches for on-board optimized guidance. FORCES Pro was used in the backend to generate fast embeddable optimization solvers. This paper describes the implementation of the simulation tools, the guidance and control algorithms, and the initial steps taken to qualify the flight software for demonstrator vehicles currently under development by FLPP and INCAS. Our software has been used to simulate thousands of end-to-end scenarios involving nominal and off-nominal conditions, different levels of dispersion, mid-flight target changes and partial engine failures. All missions in the test plan completed successfully with adequate landing states and constraint satisfaction throughout the entire flight
GNC for In Orbit Assembly” (IOA-GNC) is an ESA TRP project by a consortium led by GMV and targeting the study of autonomous assembly of large structures in space, using a modular design approach, to obtain a final structure that could not be achieved by means of a single element. The main applied principles are: use of advanced control techniques to perform rendezvous/mating operations in systems growing in complexity and changing their physical properties throughout the assembly process; and the increase of mission safety in case of contingencies, and the use of autonomous planning techniques to correct mission level failures by making on-board decisions and generating new mission plans.
The presentation will focus on autonomy aspects, internally within the GNC system and externally through on-board replanning system/capabilities.
The use of electric propulsion for orbit raising is a key factor for reducing the cost of access to space, for telecommunication satellites going to geostationary orbit (E172B, SES14, SES12) as well as for LEO satellites climbing to their operational altitude. The increased time-to-orbit is however a downside, with additional operation costs from ground station usage and personnel. Moreover, when envisaging large LEO constellations, the number of satellites that must be simultaneously operated might result in operation bottlenecks.
Increasing autonomy is therefore an important objective for reducing EOR costs, and the anticipated generalization of on-board GNSS receivers is a key element towards that goal. We present the results of a 2-year R&D study (ARTES funding) which looked into guidance strategies that would make autonomous EOR possible while being compatible with on-board CPU limitations.
The study investigated two major classes of solutions:
- semi-autonomous solutions are an extension of what Airbus is currently implementing for telecom missions with EOR: the use of on-board navigation data and smart compression allows to extend the validity of the ground-optimized guidance profile from a few weeks to a few months, thus reducing ground intervention by 80% with only minor changes to the current architecture.
- fully autonomous solutions represent a significant step further, where the on-board system can autonomously compute an optimum guidance scenario (thrust profile and complete 3-axis attitude trajectory), based on the latest GNSS data, thus limiting ground intervention to health-checks and collision-risk monitoring. It is this latter class of solutions which we present here.
After comparing several alternatives for the fully autonomous guidance approach, a best candidate solution was selected and implemented in detail. Functional simulations over a complete transfer showed that despite the disturbances and uncertainties, the performance in terms of optimality was remarkable (propellant expenditure was within 1% of the optimum, as computed by Airbus's OptElec tool).
The guidance software prototype then went through the same formal code-generation process as for real on-board software developments, and tested on a LEON3 processor. Preliminary results from the Processor-In-the-Loop test campaign show that the peak combined CPU load for the 2 asynchronous tasks plus the cyclic task amounts to less than 1%: this demonstrates that the fully autonomous guidance algorithms can indeed be implemented on board.
GENEVIS presents a full-software, Vision-Based Navigation solution dedicated to precise lunar landing. The outputs of two image processing algorithms, providing information on relative and absolute positioning respectively, are hybridized with inertial measurements. GENEVIS validates the performance and robustness of the solution with simulations, the CPU load performance on space-graded processor in real-time, and the in-flight accuracy with real hardware and representative dynamic conditions.