Satellite Edge Computing (SEC): Towards AI-Enabled Satellites

Satellite Edge Computing (SEC): Towards AI-Enabled Satellites
Namla Team

Published on May 08, 2023

Satellite Edge Computing (SEC): Towards AI-Enabled Satellites
Ready for a demo of
Namla
?

Satellites have proven to be very beneficial for large segments of the world population in various applications such as earth observation and space exploration. However, traditional satellites have been limited in their scalability due to their exorbitant costs.

Fortunately, with the recent miniaturization of hardware, new generations of satellites are being launched into space every year at a fraction of the cost. On the other hand, the large volumes of data that are captured make it impractical to perform manual analysis on. Thus, the need for AI-enabled satellites has emerged.

To address the issue of congested communication capabilities of Low Earth Orbit (LEO) satellites, researchers and industry experts are exploring various solutions. One promising approach is to incorporate AI & ML algorithms locally to automate the processing and analysis of vast amounts of generated data before being transmitted to the ground, thus easing the burden on communication capabilities.

This approach is known as edge computing, a form of distributed computing. With satellite and cloud service providers strengthening their relationship, edge solutions have become a powerful enabler for satellite networks, from ground terminals to spacecraft in orbit.

The value of applying computational power at the edge of the network is evident to civil, commercial, and military space leaders.

"We need computational and storage power in space," says Dr. Lisa Costa, the US Space Force CTIO.

Industry leaders have begun to view cloud technologies like edge computing and virtualization as necessary preconditions to unlock the potential of multi-orbit constellations, proliferated LEOs, and high-throughput satellites.

What is Edge Computing ?

Edge computing is based on the idea of moving data processing and AI inference near its source, where it is generated - so-called “Edge” - to provide near real-time insights and reduce bandwidth.

Edge computing is a decentralized or distributed architecture that brings a range of networks and devices at or near the user for better decision-making. It seeks to address the problem of data in motion. Rather than backhauling large amounts of raw data to be processed in the cloud or data center, edge computing occurs on-site.

Edge computing moves part of the storage and compute resources out of the central data center, which means relying less on cloud computing. Edge devices can vary depending on their role within the edge network, from sensors, routers, and IoT gateways to AI Boxes with built-in processors (CPU/GPU). These devices serve as entry-points to the edge network, which is mostly a LAN (Local Area Network) - a network within the same building - interfacing with the internet for cloud access for updating and monitoring purposes.

Why Satellites at the Edge ?

Satellite Edge networks have a number of advantages that make them well-suited for handling the large amounts of data generated by space-based systems. Some of the key benefits of Satellite Edge Computing (SEC) include:

  • Reduced latency

    Edge computing can help reduce latency in the satellite network by processing the data near close to its source. This is especially important for applications that require real-time processing, such as remote sensing or disaster response.

  • Improved reliability

    By distributing computing power across the network, edge computing can help improve the reliability and resiliency of satellite systems. This is particularly important for critical applications where downtime or data loss could have serious consequences.

  • Better security

    Edge computing can also help improve the security of satellite networks by reducing the amount of data that needs to be transmitted over long distances. By processing sensitive data locally, edge computing can help minimize the risk of interception or tampering.

  • Increased efficiency

    Processing data at the edge within satellite networks can reduce the volume of data that requires transmission back to ground stations. This can help improve the overall efficiency of the network and reduce costs associated with data transmission and storage.

Orbital Edge Computing is still not fully explored. However, some emerging technologies, both in software and hardware, are playing a crucial role in enabling the intersection of satellites and edge computing.

Nanosatellites, CubeSats, and even smaller Chip-Sats are examples of low-cost and lightweight satellites that can be deployed in constellations to improve connectivity and enable edge computing. These small satellites are equipped with powerful processors and have the capability to perform edge computing tasks in space, such as pre-processing data before transmitting it to the ground. They can also be launched in swarms for wide coverage.

Different Satellite Sizes

Different Satellite Sizes

On the other hand, software-defined networking (SDN) and network function virtualization (NFV) are two software-related technologies that can enable the efficient management of satellite networks and the deployment of edge computing resources.

As the demand for high-speed, low-latency data processing in space continues to grow, these emerging technologies will play an increasingly important role in the success of satellite networks at the edge.

Satellites at the Edge: Applications

The integration of edge computing technology with satellites can enable the development of a wide range of innovative applications. Below, we mention some of them:

  • Weather Observation & Forecasting Climate change

    Including the analysis of real-time data to monitor weather and climate status like monitoring global warming for example.

  • Precision Agriculture

    Satellite imagery and edge computing are used to provide real-time insights on crop health and yield predictions, allowing for more informed decision-making.

  • Risk Monitoring & Disaster Response

    Satellite data can be combined with edge computing to quickly analyze and respond to natural disasters, such as wildfires or hurricanes.

  • Territory Mapping & Urban Growth

    Detailed, real-time mapping of urban areas and land use can help city planners make informed decisions about urban growth and development. This can improve infrastructure planning and help mitigate the impact of urbanization on the environment.

  • Remote Monitoring of Critical Infrastructure

    Satellites and edge computing can play a vital role in providing real-time data and insights to help with infrastructure planning and maintenance.

  • Security & Surveillance

    This can include monitoring borders and coastlines for illegal activities, tracking the movements of ships and aircraft, and providing situational awareness for military and law enforcement applications.

As the capabilities of satellites and edge computing continue to enhance, the potential applications are everywhere, with the possibility of enabling new industries and driving innovation in existing ones.

Perspectives & constraints

Many prospects of SEC lie ahead, including:

  • Faster Response to Natural Disasters and Climate Crisis

    Data processing at the edge enables satellite networks can provide real-time insights to help with disaster response and climate monitoring.

  • Efficient Global Monitoring of Natural Resources on Earth

    Satellite imagery and edge computing can be used to monitor natural resources and provide real-time insights for more efficient resource management.

Although SEC is a promising path to explore and despite having the base technology already developed, the actual implementation and launch of this high-tech industry remain subject to many constraints, including:

  • Cost
  • Complexity
  • Regulation

Beyond that, there are some interesting ongoing projects to feature here:

  • Ф-sat-1 (or: Phi-Sat-1) - launched by ESA (European Space Agency) in 2020 as the 1st AI-deployed satellite.
  • CogniSat-6 - Another example of an AI-enabled satellite.
Takeaway

The integration of AI and edge computing with satellites has the potential to deliver significant value to various space applications. Technological advancements in hardware capabilities and growing community support for AI are driving this potential.

Rapid improvements in camera technologies and the emergence of quantum computing are expected to facilitate a new era of on-orbit AI. However, there are still challenges that need to be overcome. Environmental constraints, data security issues, and legal considerations have hindered the widespread adoption of AI for space applications.

Deploying and managing edge computing solutions within a satellite's ecosystem presents also a significant challenge. To address this issue, Namla has developed a cloud-edge management and orchestration platform to offer an efficient solution that enables seamless deployment of edge computing hardware and software within the satellite's ecosystem while automating the deployment at the edge.

By addressing these challenges, including through the use of edge computing, businesses can potentially leverage AI-enabled satellites to benefit society. The future looks promising for the intersection of satellites and edge computing.