SONATE-2

Project Overview

The SONATE-2 mission will verify novel artificial intelligence (AI) hardware and software technologies in miniaturized format in a low Earth orbit. By using such AI technologies, the satellite will be able to analyze the environment autonomously with its multiple sensors in visible and near infrared spectrum. Deep learning plays a central role as a versatile image processing tool. Thus, in addition to the classification of targets already known at the start of the mission, the payload will also have the capability of anomaly detection by on-board training for the detection of previously unknown objects or phenomena. As one of the secondary objectives, SONATE-2 will also be able to detect transient light phenomena such as lightnings by using image processing algorithms. In the future, these technologies could be used, for example, in small satellite missions to explore the solar system.

Mission Statement:

Development and operation of a technology testing satellite for highly autonomous payloads and artificial intelligence.

Start of the project: 01.03.2021
Duration: 4 years
Funding Code: 50RU2100

Contact:

Oleksii Balagurin (Project Manager)
Prof. Dr.-Ing. Hakan Kayal

The SONATE-2 mission will verify novel artificial intelligence (AI) hardware and software technologies in miniaturized format in a low Earth orbit. By using such AI technologies, the satellite will be able to analyze the environment autonomously with its multiple sensors in visible and near infrared spectrum. Deep learning plays a central role as a versatile image processing tool. Thus, in addition to the classification of targets already known at the start of the mission, the payload will also have the capability of anomaly detection by on-board training for the detection of previously unknown objects or phenomena. As one of the secondary objectives, SONATE-2 will also be able to detect transient light phenomena such as lightnings by using image processing algorithms. In the future, these technologies could be used, for example, in small satellite missions to explore the solar system.

The successful launch of SONATE-2 took place on March 4, 2024 from California on board a Falcon 9 launcher as part of the Transporter 10 mission.

SONATE-2 project team after completion of the flight model (Image: H. Kayal)

Project Manager:

O. Balagurin

Team Memebers:

Information for Radio Amateurs

2024[ to top ]

6U+ CubeSat SONATE2: Operation of an Optical AI Payload in Low Earth Orbit. Herbst, Tobias; Balagurin, Oleksii; Greiner, Tobias; Kaiser, Tobias; Kayal, Hakan; Maurer, Andreas; Schwarz, Tobias. In 75th International Astronautical Congress (IAC). 2024.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
Optical payloads and other instruments can generate large amounts of data. Data acquisition is often limited by downlink capabilities, especially in the context of space exploration, as the NASA deep space network is said to be at capacity, for example. Intelligent onboard data processing can decrease the amount of data that has to be transferred to Earth. The Nanosatellite SONATE-2, developed and operated at the University of Würzburg, tackles this problem by demonstrating artificial intelligence technologies in a low Earth orbit. SONATE-2's AI payload has specialized hardware, featuring an NVIDIA Jetson NX System on Module (SoM) with a 6-Core ARM CPU and an NVIDIA Volta architecture GPU with Tensor cores that can infer and train neural networks. This SoM enables data processing using pre-trained models on the ground and training neural networks onboard. The Linux operating system of the AI Payload allows for the use of standard libraries, like CUDA, CudNN, or TensorFlow. The 6U+ CubeSat SONATE-2 was launched with the Transporter 10 Mission by Space X into a sun-synchronous orbit in March 2024. This work presents the results of the first months of operation of the SONATE-2 Nanosatellite. The primary mission objective is to demonstrate AI technologies in space, including anomaly detection, image segmentation, and object detection using four cameras, two in the visible spectrum and two in the near-infrared. This work focuses on the technical challenges and presents the methods used to operate the AI Payload in space. In addition, we give an overview of our experiences with managing the payload, including the thermal and power considerations. A brief outlook on the planned experiments within the mission duration of one year is presented.

@inproceedings{herbst2024cubesat, abstract = {Optical payloads and other instruments can generate large amounts of data. Data acquisition is often limited by downlink capabilities, especially in the context of space exploration, as the NASA deep space network is said to be at capacity, for example. Intelligent onboard data processing can decrease the amount of data that has to be transferred to Earth. The Nanosatellite SONATE-2, developed and operated at the University of Würzburg, tackles this problem by demonstrating artificial intelligence technologies in a low Earth orbit. SONATE-2's AI payload has specialized hardware, featuring an NVIDIA Jetson NX System on Module (SoM) with a 6-Core ARM CPU and an NVIDIA Volta architecture GPU with Tensor cores that can infer and train neural networks. This SoM enables data processing using pre-trained models on the ground and training neural networks onboard. The Linux operating system of the AI Payload allows for the use of standard libraries, like CUDA, CudNN, or TensorFlow. The 6U+ CubeSat SONATE-2 was launched with the Transporter 10 Mission by Space X into a sun-synchronous orbit in March 2024. This work presents the results of the first months of operation of the SONATE-2 Nanosatellite. The primary mission objective is to demonstrate AI technologies in space, including anomaly detection, image segmentation, and object detection using four cameras, two in the visible spectrum and two in the near-infrared. This work focuses on the technical challenges and presents the methods used to operate the AI Payload in space. In addition, we give an overview of our experiences with managing the payload, including the thermal and power considerations. A brief outlook on the planned experiments within the mission duration of one year is presented.}, author = {Herbst, Tobias and Balagurin, Oleksii and Greiner, Tobias and Kaiser, Tobias and Kayal, Hakan and Maurer, Andreas and Schwarz, Tobias}, booktitle = {75th International Astronautical Congress (IAC)}, keywords = {ai}, title = {6U+ CubeSat SONATE2: Operation of an Optical AI Payload in Low Earth Orbit}, year = 2024 }
In-Orbit Validierung des MultiView-Sternsensors auf dem SONATE-2 CubeSat. Neumann, Tobias; Balagurin, Oleksii; Kaiser, Tobias; Kayal, Hakan; Maurer, Andreas; Riegler, Clemens; Schwarz, Tobias. In Deutscher Luft- und Raumfahrtkongress. Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V., Hamburg, 2024.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
Es wurde ein mehrköpfiger Sternsensor für CubeSats für den Einsatz im sog. TunaCanSlot konzipiert. Der Sensor zeichnet sich durch die gegebene Redundanz auf Grund der Verwendung einer Vielzahl an Sensorköpfen trotz des geringen Bauvolumens aus. Im Falle eines Ausfalls oder bei fehlenden Messdaten werden die Sensoren autonom umgeschaltet, die so die Verfügbarkeit der Lagedaten erhöht. Der MultiViewSternsensor wurde an Bord des SONATE2 Nanosatelliten Anfang März 2024 in die Erdumlaufbahn gebracht. Neben anderen Subsystemen als Technologiedemonstration, wurde der Sensor während der Missionslaufzeit in Betrieb genommen und bereits mehreren Tests unterzogen. Damit wird die Raumfahrttauglichkeit sowie der Einsatz im LowEarth Orbit nachgewiesen. In diesem Paper werden die für die Validierung benötigten Aspekte dargestellt. Wichtig sind die Überprüfungen sowohl der Grundfunktionen als Subsystem gegenüber dem Satelliten, als auch der Messprozesse der Sensorköpfe selbst. Die bisherigen Ergebnisse der InOrbitTests werden in Verbindung mit den dargestellten Validierungsaspekten präsentiert. Sie bestätigen die Funktion des Konzeptes und bilden bereits eine gute Grundlage für die in Aussicht gestellte Validierung auf TRL 9.

@inproceedings{neumann2024inorbit, abstract = {Es wurde ein mehrköpfiger Sternsensor für CubeSats für den Einsatz im sog. TunaCanSlot konzipiert. Der Sensor zeichnet sich durch die gegebene Redundanz auf Grund der Verwendung einer Vielzahl an Sensorköpfen trotz des geringen Bauvolumens aus. Im Falle eines Ausfalls oder bei fehlenden Messdaten werden die Sensoren autonom umgeschaltet, die so die Verfügbarkeit der Lagedaten erhöht. Der MultiViewSternsensor wurde an Bord des SONATE2 Nanosatelliten Anfang März 2024 in die Erdumlaufbahn gebracht. Neben anderen Subsystemen als Technologiedemonstration, wurde der Sensor während der Missionslaufzeit in Betrieb genommen und bereits mehreren Tests unterzogen. Damit wird die Raumfahrttauglichkeit sowie der Einsatz im LowEarth Orbit nachgewiesen. In diesem Paper werden die für die Validierung benötigten Aspekte dargestellt. Wichtig sind die Überprüfungen sowohl der Grundfunktionen als Subsystem gegenüber dem Satelliten, als auch der Messprozesse der Sensorköpfe selbst. Die bisherigen Ergebnisse der InOrbitTests werden in Verbindung mit den dargestellten Validierungsaspekten präsentiert. Sie bestätigen die Funktion des Konzeptes und bilden bereits eine gute Grundlage für die in Aussicht gestellte Validierung auf TRL 9.}, address = {Hamburg}, author = {Neumann, Tobias and Balagurin, Oleksii and Kaiser, Tobias and Kayal, Hakan and Maurer, Andreas and Riegler, Clemens and Schwarz, Tobias}, booktitle = {Deutscher Luft- und Raumfahrtkongress}, keywords = {cubesat}, publisher = {Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V.}, title = {In-Orbit Validierung des MultiView-Sternsensors auf dem SONATE-2 CubeSat}, year = 2024 }
Novel On-Board Data Processing Strategies on Nanosatellite SONATE-2. Maurer, Andreas; Balagurin, Oleksii; Greiner, Tobias; Herbst, Tobias; Kaiser, Tobias; Kayal, Hakan; Schwarz, Tobias. In Small Satellites, System & Services Symposium (4S). Palma de Mallorca, Spain, 2024.
- [ Abstract ]
- [ BibTeX ]
Reliable on-board data handling is the backbone of every successful satellite mission. New scientific experiments can only be validated, and results obtained, if the recorded scientific data reaches the ground completely and correctly. For this, several special techniques were applied during the implementation of the SONATE-2 mission. Generic communication protocols were defined between all payloads, subsystems, and the On-board data handling (OBDH) of SONATE-2 to reduce the overall complexity of the data management system. With the help of these protocols, the development effort could be reduced, and the data transfers simplified. In addition, this allowed the communication interfaces to be simulated and quickly tested during development with a desktop computer using an OBDH-Simulator.

@conference{maurerandreas2024novel, abstract = {Reliable on-board data handling is the backbone of every successful satellite mission. New scientific experiments can only be validated, and results obtained, if the recorded scientific data reaches the ground completely and correctly. For this, several special techniques were applied during the implementation of the SONATE-2 mission. Generic communication protocols were defined between all payloads, subsystems, and the On-board data handling (OBDH) of SONATE-2 to reduce the overall complexity of the data management system. With the help of these protocols, the development effort could be reduced, and the data transfers simplified. In addition, this allowed the communication interfaces to be simulated and quickly tested during development with a desktop computer using an OBDH-Simulator.}, address = {Palma de Mallorca, Spain}, author = {Maurer, Andreas and Balagurin, Oleksii and Greiner, Tobias and Herbst, Tobias and Kaiser, Tobias and Kayal, Hakan and Schwarz, Tobias}, booktitle = {Small Satellites, System & Services Symposium (4S)}, keywords = {obdh}, title = {Novel On-Board Data Processing Strategies on Nanosatellite SONATE-2}, year = 2024 }
Early Results and In-Flight Experience of the 6U-Mission SONATE-2. Schwarz, Tobias; Balagurin, Oleksii; Greiner, Tobias; Herbst, Tobias; Kaiser, Tobias; Kayal, Hakan; Maurer, Andreas. In Small Satellites, System & Services Symposium (4S). Palma de Mallorca, Spain, 2024.
- [ Abstract ]
- [ BibTeX ]
SONATE-2 is a 6U+ CubeSat for technology demonstration developed at the University of Würzburg. It was launched on SpaceX' Transporter-10 mission into a 527x505 km Sun synchronous orbit on March 4, 2024.

@inproceedings{schwarzEARLYRESULTSINFLIGHT2024, abstract = {SONATE-2 is a 6U+ CubeSat for technology demonstration developed at the University of Würzburg. It was launched on SpaceX' Transporter-10 mission into a 527x505 km Sun synchronous orbit on March 4, 2024.}, address = {Palma de Mallorca, Spain}, author = {Schwarz, Tobias and Balagurin, Oleksii and Greiner, Tobias and Herbst, Tobias and Kaiser, Tobias and Kayal, Hakan and Maurer, Andreas}, booktitle = {Small Satellites, System & Services Symposium (4S)}, keywords = {results}, title = {Early Results and In-Flight Experience of the 6U-Mission SONATE-2}, year = 2024 }
Erste Flugergebnisse der Satellitenmission SONATE-2. Balagurin, O.; Greiner, T.; Herbst, T.; Kaiser, T.; Kayal, H.; Männel, J.; Maurer, A.; Neumann, T.; Riegler, C.; Schwarz, T. In Deutscher Luft- und Raumfahrtkongress. Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V., Hamburg, 2024.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
SONATE-2 ist eine Technologiedemonstrationsmission im 6U -CubeSat-Format, deren Hauptziel es ist, neuartige Methoden der künstlichen Intelligenz für den Einsatz auf Nanosatelliten im Orbit zu verifizieren. Nach etwa drei Jahren Entwicklungszeit wurde der SONATE-2-Satellit Anfang März 2024 durch eine Falcon-9-Rakete in den Low Earth Orbit befördert und wird seitdem vom eigenen Missionskontrollzentrum der Uni-versität Würzburg betrieben. Dieses Paper fasst die Ergebnisse der ersten fünf Monate des Satellitenbe-triebs zusammen.

@inproceedings{https://doi.org/10.25967/630218, abstract = {SONATE-2 ist eine Technologiedemonstrationsmission im 6U -CubeSat-Format, deren Hauptziel es ist, neuartige Methoden der künstlichen Intelligenz für den Einsatz auf Nanosatelliten im Orbit zu verifizieren. Nach etwa drei Jahren Entwicklungszeit wurde der SONATE-2-Satellit Anfang März 2024 durch eine Falcon-9-Rakete in den Low Earth Orbit befördert und wird seitdem vom eigenen Missionskontrollzentrum der Uni-versität Würzburg betrieben. Dieses Paper fasst die Ergebnisse der ersten fünf Monate des Satellitenbe-triebs zusammen.}, address = {Hamburg}, author = {Balagurin, O. and Greiner, T. and Herbst, T. and Kaiser, T. and Kayal, H. and Männel, J. and Maurer, A. and Neumann, T. and Riegler, C. and Schwarz, T.}, booktitle = {Deutscher Luft- und Raumfahrtkongress}, keywords = {cubesat}, publisher = {Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V.}, title = {Erste Flugergebnisse der Satellitenmission SONATE-2}, year = 2024 }

2023[ to top ]

SONATE-2 – A Technology Demonstration Mission of Artificial Intelligence for Image Processing and Anomaly Detection. Greiner, Tobias; Balagurin, Oleksii; Herbst, Tobias; Kaiser, Tobias; Kayal, Hakan; Maurer, Andreas; Schwarz, Tobias. In 16th Pico- and Nano-Satellite Workshop. Würzburg, Germany, 2023.
- [ BibTeX ]
@conference{greiner2023sonate2, address = {Würzburg, Germany}, author = {Greiner, Tobias and Balagurin, Oleksii and Herbst, Tobias and Kaiser, Tobias and Kayal, Hakan and Maurer, Andreas and Schwarz, Tobias}, booktitle = {16th Pico- and Nano-Satellite Workshop}, keywords = {ai}, title = {SONATE-2 – A Technology Demonstration Mission of Artificial Intelligence for Image Processing and Anomaly Detection}, year = 2023 }
Attitude-independent calibration of magnetometers in orbit using Particle Swarm Optimization. Kaiser, Tobias; Thois, Dominik; Balagurin, Oleksii; Greiner, Tobias; Kayal, Hakan; Schwarz, Tobias; Maurer, Andreas; Herbst, Tobias; Riegler, Clemens. In 14th IAA Symposium on Small Satellites for Earth Observation. 2023.
- [ Abstract ]
- [ BibTeX ]
SONATE-2 is an upcoming CubeSat for AI applications in space. The payload includes widefield and nearfield cameras to take images of the Earth and perform tasks like image segmentation and anomaly detection. The Attitude Determination and Control System (ADCS) is designed to provide accurate pointing for these tasks. In addition, a propulsion system and a star sensor will be validated in orbit. In normal operations, the magnetic field vector and the sun vector are measured to determine the CubeSat attitude via the TRIAD or QUEST method. The harsh space environment subjects the sensors to disturbances, and it is expected that the calibration parameters will change over time. The magnetometers will be recalibrated in orbit using Particle Swarm Optimization (PSO). The algorithm records the predicted and measured magnetic field vectors from hundreds of points in time and minimises the error between their magnitude by changing the calibration parameters. This calibration was tested at the JMU Würzburg on demonstrator hardware for the ADCS and a set of Helmholtz coils simulating the Earth Magnetic Field. The resulting calibration parameters were only slightly less accurate than those found by calibrating the magnetometer in the laboratory. The calibration algorithm will be optimized further and then implemented on SONATE-2.

@inproceedings{noauthororeditorattitudeindependent, abstract = {SONATE-2 is an upcoming CubeSat for AI applications in space. The payload includes widefield and nearfield cameras to take images of the Earth and perform tasks like image segmentation and anomaly detection. The Attitude Determination and Control System (ADCS) is designed to provide accurate pointing for these tasks. In addition, a propulsion system and a star sensor will be validated in orbit. In normal operations, the magnetic field vector and the sun vector are measured to determine the CubeSat attitude via the TRIAD or QUEST method. The harsh space environment subjects the sensors to disturbances, and it is expected that the calibration parameters will change over time. The magnetometers will be recalibrated in orbit using Particle Swarm Optimization (PSO). The algorithm records the predicted and measured magnetic field vectors from hundreds of points in time and minimises the error between their magnitude by changing the calibration parameters. This calibration was tested at the JMU Würzburg on demonstrator hardware for the ADCS and a set of Helmholtz coils simulating the Earth Magnetic Field. The resulting calibration parameters were only slightly less accurate than those found by calibrating the magnetometer in the laboratory. The calibration algorithm will be optimized further and then implemented on SONATE-2.}, author = {Kaiser, Tobias and Thois, Dominik and Balagurin, Oleksii and Greiner, Tobias and Kayal, Hakan and Schwarz, Tobias and Maurer, Andreas and Herbst, Tobias and Riegler, Clemens}, booktitle = {14th IAA Symposium on Small Satellites for Earth Observation}, keywords = {adcs}, title = {Attitude-independent calibration of magnetometers in orbit using Particle Swarm Optimization}, year = 2023 }
SONATE-2 – Using Artificial Intelligence for In-Orbit Image Processing and Anomaly Detection. Greiner, Tobias; Balagurin, Oleksii; Herbst, Tobias; Kaiser, Tobias; Kayal, Hakan; Maurer, Andreas; Schwarz, Tobias. In 14th IAA Symposium on Small Satellites for Earth Observation. Berlin, 2023.
- [ BibTeX ]
@inproceedings{greiner2023sonate2, address = {Berlin}, author = {Greiner, Tobias and Balagurin, Oleksii and Herbst, Tobias and Kaiser, Tobias and Kayal, Hakan and Maurer, Andreas and Schwarz, Tobias}, booktitle = {14th IAA Symposium on Small Satellites for Earth Observation}, keywords = {ai}, title = {SONATE-2 – Using Artificial Intelligence for In-Orbit Image Processing and Anomaly Detection}, year = 2023 }

2022[ to top ]

Hardware and Software Redundancy Concepts on-board of SONATE-2. Maurer, Andreas; Balagurin, Oleksii; Greiner, Tobias; Herbst, Tobias; Kaiser, Tobias; Kayal, Hakan; Riegler, Clemens; Schwarz, Tobias. Vol. IAC 2022 congress proceedings, p. 11. International Astronautical Federation (IAF), 100 Avenue de Suffren 75015 Paris, France, 2022.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
SONATE-2 is a 6U-CubeSat, which main mission goal is the in-orbit demonstration of a newly developed AI processing platform. Experiments running on this platform include lightening detection, image segmentation and anomaly detection in images taken by the satellite. Thereby, on-board training of neural networks for AI based image processing applications will be demonstrated. Secondary mission objectives include the verification of AOCS components, tests of the ADCS’s target-pointing capabilities and the operation of an amateur radio payload. The SONATE-2 satellite is planned to be launched into a SSO in Q1 2024. par In satellites, redundancy plays a key role due to the lack of hardware-based maintenance availability, but also because of the harsh space environment. Therefore, SONATE-2 consists of two fully redundant 5V and 12V busses which supply the satellite’s subsystems and payloads with the required amount of power. Redundant subsystems are connected to one of the two power busses. In case of a power failure on one of the two busses, the redundant component supplied by the other bus is then still operational. par A key element of the design of SONATE-2 is hardware-based redundancy of the communication subsystem, the ADCS, the OBDH and the AI payload. Each of these subsystems is included multiple times on-board. SONATE-2’s communication subsystem consists of two VHF-, two UHF- and two S-Band transceivers and antennas. The ADCS includes two processing computers and a total of four three-axis magnetometers, four three-axis gyroscopes, six reaction wheel, six magnetorquer and 12 sun sensors. On-board data handling tasks are executed by one of four available OBDH computers. The decision which OBDH is executing the telecommand processing and housekeeping tasks is determined by a software-based role arbitration. Furthermore, the AI-payload which consists of a Nvidia Jetson Xavier NX and four cameras is integrated in the satellite twice. par Each of these subsystems have further built-in software-based redundancy concepts, including fault detection, isolation, and recovery algorithms as well as redundant saved configurations. This paper gives a brief overview of the SONATE-2 mission with the focus on the built-in hardware and software redundancy concepts of the satellite bus and the payloads on SONATE-2.

@conference{maurer2022hardware, abstract = {SONATE-2 is a 6U-CubeSat, which main mission goal is the in-orbit demonstration of a newly developed AI processing platform. Experiments running on this platform include lightening detection, image segmentation and anomaly detection in images taken by the satellite. Thereby, on-board training of neural networks for AI based image processing applications will be demonstrated. Secondary mission objectives include the verification of AOCS components, tests of the ADCS’s target-pointing capabilities and the operation of an amateur radio payload. The SONATE-2 satellite is planned to be launched into a SSO in Q1 2024. \par In satellites, redundancy plays a key role due to the lack of hardware-based maintenance availability, but also because of the harsh space environment. Therefore, SONATE-2 consists of two fully redundant 5V and 12V busses which supply the satellite’s subsystems and payloads with the required amount of power. Redundant subsystems are connected to one of the two power busses. In case of a power failure on one of the two busses, the redundant component supplied by the other bus is then still operational. \par A key element of the design of SONATE-2 is hardware-based redundancy of the communication subsystem, the ADCS, the OBDH and the AI payload. Each of these subsystems is included multiple times on-board. SONATE-2’s communication subsystem consists of two VHF-, two UHF- and two S-Band transceivers and antennas. The ADCS includes two processing computers and a total of four three-axis magnetometers, four three-axis gyroscopes, six reaction wheel, six magnetorquer and 12 sun sensors. On-board data handling tasks are executed by one of four available OBDH computers. The decision which OBDH is executing the telecommand processing and housekeeping tasks is determined by a software-based role arbitration. Furthermore, the AI-payload which consists of a Nvidia Jetson Xavier NX and four cameras is integrated in the satellite twice. \par Each of these subsystems have further built-in software-based redundancy concepts, including fault detection, isolation, and recovery algorithms as well as redundant saved configurations. This paper gives a brief overview of the SONATE-2 mission with the focus on the built-in hardware and software redundancy concepts of the satellite bus and the payloads on SONATE-2.}, address = {100 Avenue de Suffren 75015 Paris, France}, author = {Maurer, Andreas and Balagurin, Oleksii and Greiner, Tobias and Herbst, Tobias and Kaiser, Tobias and Kayal, Hakan and Riegler, Clemens and Schwarz, Tobias}, keywords = {redundancy}, month = {09}, pages = 11, publisher = {International Astronautical Federation (IAF)}, title = {Hardware and Software Redundancy Concepts on-board of SONATE-2}, volume = {IAC 2022 congress proceedings}, year = 2022 }
Testing the NVIDIA Jetson Xavier NX Module for the SONATE-2 Nano Satellite Mission. Herbst, Tobias; Balagurin, Oleksii; Greiner, Tobias; Kaiser, Tobias; Kayal, Hakan; Maurer, Andreas; Riegler, Clemens; Schwarz, Tobias. Vol. IAC 2022 congress proceedings, p. 7. International Astronautical Federation (IAF), 100 Avenue de Suffren 75015 Paris, France, 2022.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
SONATE-2 is a 6U+, LEO CubeSat mission, which will demonstrate the feasibility of in-orbit image processing using neural networks. These networks shall also be trained on-board. Therefore, a powerful computational unit with a small form factor and low power consumption is needed. One possible candidate, the NVIDIA Jetson Xavier NX is analysed and tested for a possible use in this mission. Thus, thermal vacuum chamber tests, total ionizing radiation dose tests, single event effect tests and vibration tests were performed to expose the module to conditions equivalent to space environment present in LEO. The module has been Total Ionizing Dose (TID) tested with a Co-60 source at a dose of 10,krad. Single Event Effect (SEE) testing has been done with protons at energy levels of 30, 50 and 68 MeV. In addition to that, vibration tests have been conducted on acceptance level for a launch with a Soyuz-2 rocket for a containerized CubeSat. The thermal-vacuum tests included an outgassing test and a thermal cycling test with 5 cycles from -30$$^\circ$$C to 90$$^\circ$$C. After the TID tests, the vibration tests and the thermal tests, no damage to the module could be detected. During the outgassing test, a total mass loss of at least 0.08,% was measured. During the SEE tests, two single event effects occurred, of which one caused the device to reboot and another one caused a memory error in the modules eMMC memory. Thus, it is recommended to implement a redundancy strategy for the content of the eMMC memory and operate the device, such that spontaneous reboots do not lead to critical failures.

@conference{noauthororeditor, abstract = {SONATE-2 is a 6U+, LEO CubeSat mission, which will demonstrate the feasibility of in-orbit image processing using neural networks. These networks shall also be trained on-board. Therefore, a powerful computational unit with a small form factor and low power consumption is needed. One possible candidate, the NVIDIA Jetson Xavier NX is analysed and tested for a possible use in this mission. Thus, thermal vacuum chamber tests, total ionizing radiation dose tests, single event effect tests and vibration tests were performed to expose the module to conditions equivalent to space environment present in LEO. The module has been Total Ionizing Dose (TID) tested with a Co-60 source at a dose of 10\,krad. Single Event Effect (SEE) testing has been done with protons at energy levels of 30, 50 and 68 MeV. In addition to that, vibration tests have been conducted on acceptance level for a launch with a Soyuz-2 rocket for a containerized CubeSat. The thermal-vacuum tests included an outgassing test and a thermal cycling test with 5 cycles from -30$^\circ$C to 90$^\circ$C. After the TID tests, the vibration tests and the thermal tests, no damage to the module could be detected. During the outgassing test, a total mass loss of at least 0.08\,\% was measured. During the SEE tests, two single event effects occurred, of which one caused the device to reboot and another one caused a memory error in the modules eMMC memory. Thus, it is recommended to implement a redundancy strategy for the content of the eMMC memory and operate the device, such that spontaneous reboots do not lead to critical failures.}, address = {100 Avenue de Suffren 75015 Paris, France}, author = {Herbst, Tobias and Balagurin, Oleksii and Greiner, Tobias and Kaiser, Tobias and Kayal, Hakan and Maurer, Andreas and Riegler, Clemens and Schwarz, Tobias}, keywords = {ai}, pages = 7, publisher = {International Astronautical Federation (IAF)}, title = {Testing the NVIDIA Jetson Xavier NX Module for the SONATE-2 Nano Satellite Mission}, volume = {IAC 2022 congress proceedings}, year = 2022 }
SONATE-2 - A Technology Demonstration Mission for Artificial Intelligence. Greiner, Tobias; Balagurin, Oleksii; Herbst, Tobias; Kaiser, Tobias; Kayal, Hakan; Maurer, Andreas; Neumann, Tobias; Riegler, Clemens; Schwarz, Tobias. In Small Satellites, System & Services Symposium (4S), G. Tobias, B. Oleksii, H. Tobias, K. Tobias, K. Hakan, M. Andreas, N. Tobias, R. Clemens, S. Tobias (eds.). Vilamoura, Portugal, 2022.
- [ Abstract ]
- [ BibTeX ]
The increasing amount of data produced by sensors, combined with low download rates, are a challenge for nanosatellites. On-board data processing and classification can be used to reduce the amount of data that needs to be downloaded. This edge computing approach has become more and more present in various terrestrial applications and embedded devices. Recent missions have used artificial intelligence (AI) models, pre-trained on the ground, for on-board image analysis. The SONATE-2 mission expands on this approach by also introducing on-board training capabilities, enabling new applications including on-board model generation and improvement.

@inproceedings{greiner2022sonate2, abstract = {The increasing amount of data produced by sensors, combined with low download rates, are a challenge for nanosatellites. On-board data processing and classification can be used to reduce the amount of data that needs to be downloaded. This edge computing approach has become more and more present in various terrestrial applications and embedded devices. Recent missions have used artificial intelligence (AI) models, pre-trained on the ground, for on-board image analysis. The SONATE-2 mission expands on this approach by also introducing on-board training capabilities, enabling new applications including on-board model generation and improvement.}, address = {Vilamoura, Portugal}, author = {Greiner, Tobias and Balagurin, Oleksii and Herbst, Tobias and Kaiser, Tobias and Kayal, Hakan and Maurer, Andreas and Neumann, Tobias and Riegler, Clemens and Schwarz, Tobias}, booktitle = {Small Satellites, System & Services Symposium (4S)}, editor = {Tobias, Greiner and Oleksii, Balagurin and Tobias, Herbst and Tobias, Kaiser and Hakan, Kayal and Andreas, Maurer and Tobias, Neumann and Clemens, Riegler and Tobias, Schwarz}, keywords = {ai}, title = {SONATE-2 - A Technology Demonstration Mission for Artificial Intelligence}, year = 2022 }

2021[ to top ]

Development and operation of a technology testing satellite for highly autonomous payloads and artificial intelligence. Balagurin, Oleksii; Greiner, Tobias; Herbst, Tobias; Kaiser, Tobias; Kayal, Hakan; Maurer, Andreas; Neumann, Tobias; Riegler, Clemens; Schwarz, Tobias. In 14th Pico- and Nano-Satellite Workshop. Würzburg, Germany, 2021.
- [ BibTeX ]
@conference{noauthororeditor2021development, address = {Würzburg, Germany}, author = {Balagurin, Oleksii and Greiner, Tobias and Herbst, Tobias and Kaiser, Tobias and Kayal, Hakan and Maurer, Andreas and Neumann, Tobias and Riegler, Clemens and Schwarz, Tobias}, booktitle = {14th Pico- and Nano-Satellite Workshop}, keywords = {ai}, month = {09}, title = {Development and operation of a technology testing satellite for highly autonomous payloads and artificial intelligence}, year = 2021 }

Artificial Intelligence for Space

04/14/2021

Building a satellite with artificial intelligence on board that is trained in space: For this project, Professor Hakan Kayal from Würzburg is receiving 2.6 million euros from the German Federal Ministry for Economic Affairs and Energy.

more

Page 4 of 4.

previous
1
2
3
4

BR Video Report on SONATE-2 (DE)

A video report (DE) with an interview with Prof. Kayal was published for the BR Frankenschau. In it, he explains the background to the satellite launch and gives an insight into the profressorship's mission control center, which was described by the Bavarian Minister of Science, M. Blume as 'Houston am Main'.

(Bild: R.Emmerich)

Tagesschau Artikel zu SONATE-2 Start

The SONATE-2 launch to test AI technologies in space attracted a great deal of media attention, including an article in the Tagesschau (DE), which shed light on the background ot the mission and the work at the professorship.

Additional press releases on the project:

06.03.2024 - BR (Videobeitrag) Nanosatellit erfolgreich ins All gestartet

06.03.2024 - BR - Meilenstein in Würzburg: Unisatellit in den Weltraum gestartet

05.03.2024 - MainPost - Erfolgreicher Start in den Weltraum

05.03.2024 - Tagesschau - Deutsche KI im All

05.03.2024 - JMU press article on the successful launch of SONATE-2

06.11.2023 - JMU press article on the development and objectives of SONATE-2

02.07.2021 - Vorbereitet auf die Zukunft, Presseartikel Süddeutsche Zeitung

14.04.2021 - JMU press article on SONATE-2

Impressions from Orbit

1
2
3
4
»

+++ Liveticker +++

Link to livestream (finished): https://www.youtube.com/watch?v=4qOBXHmDCUU

13.05.2024 First high-resolution image of SONATE-2

We have taken and downloaded the first high-resolution images of the SONATE-2 nano-satellite. Here you can see an image from our wide-angle colour camera. It shows a part of North Africa and the Strait of Gibraltar.

16.04.2024 We are celebrating 42 days of uninterrupted operation of the SONATE-2 on-board computer.

02.04.2024 First time training of the anomaly detection networks in orbit completed!

20.03.2024 Official NORAD ID assigned

In the past few days SONATE-2 has been added to the official NORAD catalog aufgenommen.

International Designator: 2024-043Q
NORAD-ID: 59112

12.03.2024 First image of SONATE-2 downloaded successfully!

First wide field of view image from SONATE-2 downloaded successfully!
This is a reduced and compressed version of the raw image.

Can you guess which coastline you can see here?

08.03.2024 First SSTV message received from SONATE-2

Today, for the first time, an image previously stored on board was successfully transmitted via SSTV (Slow Scan TV).
This is a low-resolution image transmitted via an analogue transmission method in an amateur radio band.

Dates for further transmissions can be found here.

06.03.2024 Press conference

Together with the President of the University, we celebrated the successful launch of SONATE-2 at a press conference. (Image: R.Emmerich)

06.03.2024 First ADCS data downloaded

As part of the LEOP operations, the redundant ADCS system was operated for the first time today and various status data was collected. All downloaded data of the satellite are within the expected range. The plot shows the rotation rates of SONATE-2 over time. (Figure T.Kaiser).

06.03.2024 Onboard computer is running very reliably

The SONATE onboard computer, which forms the "brain" of the satellite and an in-house development of the Professorship of Space Technology, is proving to be very reliable and has so far run without rebooting. An up-time of > 1 day was already achieved yesterday, as the housekeeping data shows (parameters OBDH1 UPTIME, OBDH3 UPTIME; image: H.Kayal).

06.03.2024 Picture of the SONATE-2 deployment

Image of SONATE-2 during the deployment as one of SpaceX's Transporter-10 missions payloads on March 4, 2024 (Image source: SpaceX, edited by J. Männel)

05.03.2024 - 15.30 New TLE data published

Updated TLE data has been published:

SONATE-2
1 70309C 24043J   24064.95799882 .00043189 00000+0 22599-2 0    01
2 70309 97.4633 190.2549 0015255 299.5135 40.1876 15.15805287    14

05.03.2024 - 15.15 SONATE-2 in stable condition, first day passes completed

After the first two day passes completed today (4 passes in total since deployment), over 1000 telemetry frames have already been downloaded thanks to the successful and robust communication with the satellite. The analyzed housekeeping data show a safe and stable state of the satellite.

05.03.2024 - 01.10 End of Stream

The livestream is now offline. Many thanks to all viewers who accompanied us during this exciting mission phase! Further updates on SONATE-2 will follow in the coming days.

05.03.2024 - 01.00 SONATE-2 in a healthy condition

The analyses of the first telemetry data show that SONATE-2 is in a healthy state. We are now starting the LEOP procedures.

05.03.2024 - 00.27 SONATE-2 received by SatNOGS ground stations

Ground stations of the SatNOGS network have received SONATE-2! Thanks to all operators of SATNOGS stations for their support! Link to SatNOGS DB

05.03.2024 - 00.25 Uhr Telemetry is being analyzed!

Screenshot from livestream during first contact with SONATE-2.

Communication with SONATE-2 is stable, initial telemetry data has been received and is being analyzed in the next minutes. The next contact with SONATE-2 is expected at 01:47 (CET)

05.03.2024 - 00.15 Uhr First Signal Received

It's official! We have made first contact with our SONATE-2 satellite! We are now starting to analyze the telemetry data received, which shows the status of the satellite.

05.03.2024 - 00.03 Uhr Groundstation is awaiting contact

The ground station is ready for operation and pointed in a south-easterly direction, where SONATE-2 will be visible for the first time. Everything is ready for the first contact with SONATE-2 in orbit. First contact is expected around 00:14 (CET) on 05.04.2024.

05.03.2024 - 00.01 Uhr Deployment confirmed

The deployment of SONATE-2 into its Sun Synchronous Orbit (SSO) was successful!

04.03.2024 - 23.35 Operational plan after deployment:

While we wait for the arrival at the deployment point, some information about the planned actions by SONATE-2 until the first contact attempt:

T+0s: Deployment from Exolaunch Container
T+2s: Boot Onboard Computer
T+3min: Deploy Antennas
T+4min: Deploy Solar Array
T+5min: Start CW Transmission
T+15min: First Contact

04.03.2024 - 23.07 Scheduled engine shutdown of the 2nd stage

After the scheduled engine shutdown, the upper stage will fly to the deployment locations. For SONATE-2, this is reached at around 23:59 (CET).

04.03.2024 - 23.07 Uhr Stage Seperation

The separation of the Falcon 9 first stage has just been confirmed by SpaceX. The rocket will continue its flight with the second stage up to the S-2 point (separation of the 2nd stage). This is expected to be reached at around 23:59 (CET). So far, everything is going according to plan.

04.03.2024 - 23.05 Uhr Lift-Off

LIFT-OFF! The Falcon-9 blasts off from Space Launch Complex 4E in Vandenberg SFB, CA! Have a safe journey SONATE-2!

04.03.2024 - 22.55 Uhr T-10 Minutes Countdown Started

Launch preparations for the Transporter-10 mission with SONATE-2 on board are almost complete, the Countdown is ticking! Scheduled launch at (23:05:00 MEZ)

04.03.2024 - 22.35 Final preparations for the start are underway

Less than 30 minutes until launch. According to the plan, the rocket is now being fueled and final checks are being carried out.

04.03.2024 - 22:28 Uhr Livestream - problems solved

The technical problems in the livestream have been resolved. Thanks to the team in the background.

04.03.2024 - 22:23 Uhr Livestream - smaller technical problems

We are currently experiencing smaller technical issues with our livestream. Our media team is working on a solution at full speed.

04.03.2024 - 22:08 Good weather conditions at the launch site

SpaceX has officially confirmed that the current weather conditions at the launch site at Vandenberg Space Force Base are very good.

04.03.2024 - 21.30 Uhr Livestream gestartet

The livestream has now officially started! Follow the launch event live on YouTube:https://www.youtube.com/watch?v=4qOBXHmDCUU

Timeline for the stream (Image: T.Neumann)

04.03.2024 - 21.05 Uhr The tension is rising!

Less than an two hours left until launch! The image shows SONATE-2 in the Exolaunch deploy container (yellow marker) on the Transporter-10 mission. (Image source: SpaceX)

04.03.2024 - 15.15 Uhr Planned First Contact Opportunities

Based on the TLE data, the first orbits of SONATE-2 could be simulated including the first possible contact times to our ground station in Würzburg. Further information on the reception of SONATE-2 is summarized on the Information for Radio Amateurs page.(Image: T.Neumann)

04.03.2024 - 12.00 Uhr TLE published

SONATE-2
1 70309C 24043J   24064.95832326 .00068411 00000+0 35030-2 0    01
2 70309 97.4548 190.2543 0011790 296.9003 44.7014 15.16600872    16

04.03.2024 - 10.30 a.m. Livestream URL published

Link to the livestream: https://www.youtube.com/watch?v=4qOBXHmDCUU

Key events (CET):

Stream Start 21:30
Rocket Launch 23:05
Expected first contact: (approx.) 00:15

04.03.2024 - 07.30 Uhr a.m. Launch details published by SpaceX - Launch on 04.03.2024 at 23:05 (CET)
SpaceX has published the launch details of the Transporter-10 mission this morning, which will transport SONATE-2 into orbit. The launch will take place today, March 4, 2024 at 23:05 (CET). A first contact attempt with SONATE-2 is expected at approximately 70 minutes after launch.

Link to the SpaceX announcement

29.02.2024 SONATE-2: Launch event - livestream announcement

Preparations are underway for the live broadcast of the SONATE-2 launch (Transporter-10 mission). A livestream will cover the entire launch event and provide exclusive insights from the mission control center. Further information on the exact launch date and the link to the livestream will be published here as soon as the official launch date has been announced.

26.02.2024 - SONATE-2: Information for radio amateurs

In preparation for the launch, details of SONATE-2 reception for radio amateurs have been published, see here: Information for Radio Amateurs

21.02.2024 - SONATE-2: Test Operations started

The test operations for SONATE-2 have now started. As of today, the SONATE-2 team is testing the procedures and commands for the safe commissioning of SONATE-2 directly after the launch in a few weeks at the Mission Control Center of the professorship for space technology.

20.02.2024 - Liveticker announcement

This live ticker will accompany the SONATE-2 launch. The launch is currently planned for the beginning of March 2024 with transporter mission 10. Exact details on the launch date will be published at the end of February.

The live ticker will accompany the entire launch event. There will also a livestream. Further information will also be published here at the end of February. Check also the german site for additional updates.

To top

Social Media

Contact

E-Mail: sonate-2@uni-wuerzburg.de

Satellite Information

Name: SONATE-2
Callsign: DP0SNX
Int.-Desig.: 2024-043Q
Norad-ID: 59112

Mission Patch

Research Funding

Hubland Nord, Geb. 32