VIREO Spacecraft

Collaborative project between C3S and AIMotive a leading provider of camera-first, AI powered, Level 5 autonomous driving technology.

The mission is a collaborative project between C3S, a high reliable nanosat developer company and AIMotive a leading provider of camera-first, AI powered, Level 5 autonomous driving technology. The results of this collaboration are expected to accelerate the commercialization of a wide range of services for both specialized and mass-market applications, such as IoT, telecommunications, Earth and space observation, autonomous satellite operation, docking support, asteroid mining etc.

VIREO’s on-board demonstration experiments have been running for months with excellent results, from the very beginning of the space experiments, the on-board AI algorithm successfully pre-selected the images that the system sent back to Earth.

A number of space experiments have been carried out on board VIREO, including a joint development with the aiMotive team. This involved processing image data captured by the satellite’s cameras in space using AI-based image processing algorithms. This demonstration was also an innovative technology transfer, as we integrated an automotive hardware accelerator, aiWare, into the satellite. Another significant experiment was the on-board run of an AI-enabled software algorithm suite from KP Labs in Poland, which learned on-board cloud detection and cloud mask generation using VIREO on-board images after reconfiguring the neural network.

A new feature on the satellite is that, in addition to the experiments already presented, we have also tested a payload control subsystem of our own design in the 10 × 10 × 10 cm cargo bay. This will support future CubeSat projects by simplifying and speeding up integration and providing outstanding on-board computing capacity for the various sensors and experiments. We are also prepared for future Earth observation missions.

We have also demonstrated several imaging techniques with four different types of cameras in space. For example, we have demonstrated that the satellite can capture images by focusing on a point on the ground and tracking it as it moves in orbit. We have also taken ground target tracking and burst shooting images and developed automated image acquisition processes for Earth-observation purposes.

During its 16-month mission, CubeSat captured breathtaking images. These stunning images of the Earth from space provides a unique perspective of our planet that also shows its fragile beauty.

Although the mission is over, the success of VIREO has proven that artificial intelligence technology can make earth observation faster and more efficient, which can mean advances in sectors such as agriculture, disaster management, and energy extraction.

Images from space can be used to quickly and accurately track changes in the earth’s surface and unexpected events. In addition, its further development could pave the way for areas such as better coordination of satellite fleets, bringing the internet to remote areas, space mining, reducing space debris or the success of very distant space missions.

Today, on 7th August 2024 morning, the satellite will deorbit over New Zealand at 11:39am, reach our atmosphere and will be destroyed forever.

New platform capabilities demonstrated in VIREO mission:
Within 1U payload bay:

• 2 general-purpose, AI-capable Single-Board Computers designed from scratch
• 2 × SpaceWire + 1 × Gigabit Ethernet links (harness-based)
• 2 × automotive GMSL + 2 × commercial JPEG camera systems in 2 different lens configurations
• AI-related IODs: cloud-detection capabilities demonstrated using:

1. NextChip’s Apache5 SoC with aimotive’s aiWare CNN accelerator
2. KP Labs’ „The Herd” AI-based algorithm collection 

Technological advancements in C3S’ portfolio:

• Hardware: System-on-module, System-on-Chip, high-density storage, high-speed serial links, and FPGA integration
• Software: combined FreeRTOS & Linux-based OBDH approach, adaptive payload-handling mechanism
• FPGA: first mission of an in-house developed soft-processor system incl. ≈30 proprietary IP cores
• Mechanical engineering: 3D-printed camera support structure
• Operation: fully automated image acquisition processes for Earth-observation purposes

1. random thumbling
2. ground target tracking
3. burst shooting (in nadir direction or with an automatically calculated roll angle)

Unsure about the deployment time. Using Jonathan McDowell's.

[1]

• BRO-9
• Brokkr-1
• Connecta T2.1
• FACSAT-2
• INSPIRE-Sat 7
• It’s About Time
• KILICSAT (KILIÇSAT)
• Lemur-2 164
• Lemur-2 165 (Adler-2)
• Lemur-2 163 (KAUST)
• LS2f
• Platform-3
• Pleiades - Sapling Giganteum
• Pleiades Yearling 2.0
• RoseyCubesat-1
• Space-D DEWA 2
• SSS-2B
• TAIFA-1
• ELO-3
• EPICHyper-1
• Kepler-20
• Kepler-21
• VCUB1
• CIRBE
• DISCO
• IRIS-C
• REVELA
• VIREO
• LLITED A
• LLITED B