Do you need clouds in space

, , . , , « », - . Cloud4Y , .

A joint project between the space agencies of the United States and India involves the launch of a dual-frequency synthetic aperture radar from the Sriharikot space center in Andhra Pradesh in 2022. The NISAR satellite will be the first radar to use dual frequencies (L and S bands). The use of advanced radar technology should provide an unprecedented, highly detailed view of the Earth. The NISAR satellite allows observing and measuring ecosystem disturbances, ice cover destruction, as well as predicting and tracking natural disasters - earthquakes, tsunamis, volcanic eruptions and landslides.

What does a satellite look like

NISAR


NISAR

NISAR is expected to generate 100 terabytes per day. This is a lot of data that needs to be collected and processed for several space agencies involved in the project. Own data centers of space agencies cannot cope with such a load. Therefore, NASA will use the cloud. Thanks to this technology, we have a good place to store, analyze and analyze data that will be used for the benefit of mankind.

Interestingly, scientists intend to actively use GPU computing technology. Previously, they were not familiar with it, but when studying the applicability of the technology, it was possible to establish that GPUs are sometimes 100 times better than traditional methods of calculation, and in general give results four times faster. In addition, scientists will also be able to switch between the CPU and GPU, depending on what is cheaper and more appropriate to use at the moment.

The data collected during the NISAR mission will help to reveal information about the development and current state of the earth's crust, help scientists better understand the processes of climate change on our planet, and theoretically provide an opportunity to manage available resources and predict disasters. The cloud will help researchers process data differently. They will be able to use machine learning to identify trends, both in scientific data and in telemetry data.

Spacecraft tracking and identification

The American Space Development Agency ( SDA ) believes that cloud technology can provide significant assistance in developing a project to track and quickly identify rockets and other spacecraft. Tracking sensors generate a huge amount of information, which must be quickly processed and transmitted further. According to the military, this is a matter of national security. The later the information is received, the greater the risk for the United States in case of emergency.

Cloud platforms are designed to eliminate a bottleneck, eliminating unnecessary delays. Satellites equipped with spaceborne sensors will continue to perform partial data processing on board. But a significant part of the processes will be given to the cloud provider.

According to representatives of the Agency, on-site data processing is attractive due to the variety of commercial cloud providers and services, which can also offer greater flexibility than hardware and software on satellites. Cloud computing is one of those architectures that gives companies the flexibility to use the data you collect, process and transmit with greater efficiency.

At the same time, the agency notes the risks associated with cybersecurity. However, they notice that there have always been security problems. But they should not interfere with the development of technology and new discoveries. So soon monitoring of space objects will be carried out using cloud technology.

Cosmic clouds

There are also ideologists of the exact opposite approach to the use of clouds. Aerospace Corp. Team together with Intel, it created the Space Cloud artificial intelligence system, which uses modern cloud computing to allow satellites to transmit meaningful data and discard the rest.

Given that there are about 1800 satellites in orbit and their number is expected to increase tenfold in the coming years, the researchers plan to demonstrate how ground-based cloud computing and artificial intelligence can be moved into space for processing on board.

Engineers admit that satellites are good for observation, but weak in data processing. Modern spacecraft can collect and transmit a large amount of data, but analysts on Earth must sort through all this material in order to find something valuable. But if you teach a satellite how to process the collected information and transmit only the data that analysts need, then this will solve a number of problems.

Aerospace Corp. Team worked on a device that could fit in cubsat- an extremely small Earth satellite weighing several kg. The team created their technology using Intel Movidius and Kubernetes. The software is borrowed from Google's open source cloud software, which moves data between server farms depending on traffic and other factors.

This is what AI looks like.

Space Cloud is an artificial intelligence system that uses modern cloud computing to allow satellites to detect and transmit only relevant data. The satellite collects data, analyzes it, sends the necessary information to the ground, and then is ready to perform a new task. Saving time and resources is evident.

If satellites do not have to spend so much energy transmitting unnecessary data, it will be possible to reduce communication systems, mission power systems and solar panels, and therefore the size of the satellite itself.

They want to demonstrate the work of technology this fall, when Cubsat with AI support will be launched. Space sensors in low Earth orbit should always be in motion, so they spend part of their time above the water, and the other part above the ground. The event is codenamed "goats and boats."

Space Cloud will teach AI, when above water, to remove only boats. If he finds this “significant data”, he discards unnecessary data and transmits only the image or video of the boat to the ground. Being above the ground, he changes his schedule to search for goats. And that’s all he will be looking for. You can also be instructed to search only one region: for example, goats in Australia.

Of course, goats are a euphemism. Analysts will replace them with more interesting objects: perhaps tanks or missile bunkers.

If satellites do not have to spend so much energy transmitting unnecessary data, it will be possible to reduce communication lines, mission power systems and solar panels, and therefore the size of the satellite itself.

Summary

I must say that cloud systems can be used for other purposes. For example, to track and predict the trajectory of space debris and other objects that could threaten both ground and space objects. Perhaps you can offer other ways to apply the technology. However, one thing can definitely be said: there can be clouds in space!

What else can you read on the Cloud4Y blog

→ What is the geometry of the universe?
→ Easter eggs on topographic maps of Switzerland
→ A simplified and very short history of the development of “clouds”
→ How the bank “broke”
→ Computer brands of the 90s, part 3, final

Subscribe to our Telegram-channel, so as not to miss another article. We write no more than twice a week and only on business.