Home / Technology / AI & IT / The Age of Onboard AI: Revolutionizing Space and Satellite Missions

The Age of Onboard AI: Revolutionizing Space and Satellite Missions


The exploration of outer space has long been a testament to human curiosity, innovation, and technological advancement. In an era of unprecedented advancements in space exploration, the role of artificial intelligence (AI) has become increasingly prominent. AI is now an integral part of satellite and spacecraft missions, empowering them with autonomous decision-making, data processing capabilities, and onboard diagnostic expertise. This article explores how AI is transforming the way we manage data, process information, and enhance the efficiency of space and satellite missions.


The Data Deluge

One of the key benefits of using AI on board spacecraft is that it can help to reduce the amount of data that needs to be downlinked to Earth. This is important because downlinking data can be expensive and time-consuming. AI can be used to filter and compress data on board the spacecraft, sending only the most important information back to Earth.


Data Analysis and Interpretation

Space is filled with data, and lots of it. Satellites and spacecraft gather colossal volumes of information, from celestial observations to Earth monitoring. Without AI, processing and deciphering this data would be an overwhelming task. AI algorithms can swiftly identify patterns, anomalies, and valuable insights within these datasets, aiding scientists and researchers in their quest for knowledge.

Space missions, particularly those involving satellites, are producing an immense volume of data. These missions capture data from Earth and beyond, creating a treasure trove of information that demands effective management. AI is the key to sifting through this data and transforming it into meaningful insights. Whether it’s weather observations, remote sensing, or planetary exploration, AI plays a crucial role in speeding up data delivery and improving the efficiency of space missions. AI makes sense of the data deluge, helping us understand our universe and our planet better.


Remote Sensing and Earth Monitoring

One of the most remarkable applications of AI in space missions is its ability to enhance remote sensing. AI-equipped satellites can track changes on Earth’s surface, monitor environmental shifts, and detect natural disasters. The precision and speed of AI-driven remote sensing applications have revolutionized disaster response, environmental research, and even agricultural practices. AI in space helps us keep a watchful eye on our home planet.


Autonomous Decision-Making

Imagine sending a spacecraft millions of miles from Earth, far beyond our direct control. The communication lag could be detrimental in critical situations. This is where AI steps in. Onboard AI allows spacecraft to make decisions autonomously by analyzing data, adjusting to changing conditions, and even reacting to unforeseen events without human intervention. It’s like having a virtual astronaut onboard, making real-time choices to ensure mission success.


Interplanetary Exploration

As we set our sights on other celestial bodies, AI becomes an invaluable partner. Spacecraft traveling to distant worlds must navigate unfamiliar terrain and make critical decisions. For example, NASA’s Perseverance rover used AI for terrain analysis and obstacle avoidance during its dramatic landing on Mars. AI’s prowess in autonomous navigation opens up a world of possibilities for future planetary exploration.

Resource Optimization and Efficiency

Onboard AI contributes to resource optimization on spacecraft. It manages power usage, propulsion, and data storage, extending mission lifespans and improving efficiency. These resource-conscious AI systems ensure that every watt of energy, every drop of propellant, and every byte of data are utilized to their full potential.

Enhanced Communication

Efficient communication is vital for space missions, especially when faced with vast distances. AI systems improve satellite communication by optimizing signal processing, reducing interference, and maximizing bandwidth efficiency. This translates into clearer and more reliable connections with Earth, even when traversing the far reaches of our solar system.

Redundancy and Resilience

AI also plays a key role in ensuring mission resilience. In the event of hardware failures, AI systems can take over critical functions, providing redundancy that is essential for the success of long-duration missions. This autonomy enhances a mission’s ability to adapt to unexpected challenges.

The NASA-developed Research in Artificial Intelligence for Spacecraft Resilience (RAISR) software represents a groundbreaking leap in space technology. In essence, RAISR acts as a diagnostic tool for spacecraft, similar to a car’s check engine light. Conventional fault tree diagnosis relies on simple physics and known parameters, where the spacecraft responds to predefined scenarios. However, it falls short when dealing with unexpected or complex faults, such as a spacecraft entering Earth’s shadow or sustaining damage from micrometeoroids.

RAISR bridges this gap by employing AI to decipher the underlying causes of these issues, emulating human reasoning to draw connections between seemingly unrelated events. For instance, it can link a decrease in the spacecraft’s temperature to a malfunction in its internal heat regulation system, even in cases of severe faults.

Crucially, RAISR’s real-time fault diagnosis has the potential to expedite the recovery process, reducing downtime during missions. While human analysis takes time and consumes valuable resources, particularly in terms of communication networks and bandwidth, RAISR’s AI capabilities facilitate prompt responses to emerging issues.

Moreover, RAISR’s adaptability and utilization of both machine learning and classical AI techniques make it invaluable for diagnosing anomalies. Machine learning excels at identifying known faults based on historical data, but its effectiveness is limited by the quantity and diversity of data available. In situations involving novel faults, classical AI steps in to make informed decisions, compensating for the absence of prior data.


Onboard AI in Action

The On-Board Computer (OBC) of a satellite serves as the brain of the spacecraft. It’s responsible for controlling the payload, power generation, attitude control, and communication. Designing OBCs is a complex task, especially for small satellites that require miniaturization and energy efficiency. These computers have a finite lifespan in space, heavily influenced by the radiation level of the environment. For long-duration missions, radiation-hardened components are essential. AI technologies offer solutions to these challenges, making satellite missions more efficient and productive.

In April 2022, Titan Space Technologies successfully deployed a suite of machine learning models on the HPE Spaceborne Computer-2 aboard the International Space Station (ISS) in collaboration with Axiom Space and HPE. This achievement highlights the practical application of artificial intelligence (AI) in space experiments and addresses the evolving demands of modern space stations and spacecraft. The success underscores the crucial role of AI capabilities in supporting mission-critical functions in space, both in low Earth orbit (LEO) and beyond. The industry’s commitment to AI for space infrastructure development is essential as space commercialization and human activity in LEO continue to grow, signifying a significant milestone in space technology advancement.


Intel Empowers Satellites with AI: A Space Revolution

In October 2020, the European Space Agency (ESA) unveiled PhiSat-1, a groundbreaking satellite that marked a significant leap in space technology. This innovative satellite was equipped with a new hyperspectral-thermal camera and onboard AI processing, courtesy of Intel’s Movidius Myriad 2 Vision Processing Unit (VPU), a chip known for its use in smart cameras and even consumer drones. PhiSat-1 was part of a pair of satellites on a mission to monitor polar ice and soil moisture, all while testing intersatellite communication systems, laying the foundation for a future network of federated satellites.

The primary challenge PhiSat-1 aimed to address was the enormous data generated by high-fidelity cameras. The growth in sensor capabilities far outpaced our ability to download and process the data on Earth. A factor of 100 increase in data generation for every sensor generation compared to a factor of three, four, or five for data download posed a significant bottleneck. To compound the issue, roughly two-thirds of the Earth’s surface is typically obscured by clouds, leading to the capture, storage, and transmission of vast quantities of redundant cloud images that were eventually deleted.

This is where onboard artificial intelligence, located at the edge of space, played a pivotal role. The idea was simple yet transformative – leverage onboard processing to identify and discard cloudy images in real-time, conserving approximately 30% of precious downlink bandwidth. Space, as they say, is the ultimate edge, and PhiSat-1’s AI technology, developed in collaboration with Irish startup Ubotica, cosine, the camera’s manufacturer, and the University of Pisa and Sinergise, has harnessed the power of AI in this extreme environment. By pioneering hardware-accelerated AI inference for Earth observation images on an in-orbit satellite, PhiSat-1 ushered in a new era of efficient satellite data utilization, significantly reducing downlink costs and saving scientists’ valuable time on the ground.

PhiSat-1 also symbolizes the promising future of low-cost, AI-enhanced miniaturized satellites that can run multiple applications, offering versatility and adaptability in space missions. Rather than being limited to a single purpose, these satellites can switch between various tasks and applications, introducing the concept of “satellite-as-a-service.”

Furthermore, AI in space is not limited to scientific research and Earth observation. It also plays a vital role in defense systems, particularly for military satellites. These next-generation defense systems are increasingly autonomous, necessitating advanced data processing and decision-making capabilities at the edge of space. The integration of artificial intelligence and machine learning (AI/ML) in space systems is poised to provide higher levels of autonomous command and control, ensuring faster responses to emerging threats.

To harness the full potential of AI/ML onboard space systems, a substantial increase in computing capabilities is required. This shift will likely see the emergence of a diverse range of heterogeneous systems, including combinations of CPUs, GPUs, FPGAs, and purpose-built ASICs, marking a new frontier in space technology.

Recent Developments

NASA’s Jet Propulsion Laboratory (JPL) is developing a new type of AI-powered spacecraft called the Autonomous Mission Operations (AMO) spacecraft. The AMO spacecraft will be able to plan and execute its own missions, without the need for human intervention. This is a significant step forward in the development of autonomous spaceflight, and it has the potential to revolutionize the way that space missions are conducted.

The AMO spacecraft will use a combination of AI technologies, including machine learning, planning, and scheduling, to make decisions about its own operations. This will allow the spacecraft to respond to unexpected events and to optimize its performance. The AMO spacecraft is also expected to be more resilient to failures than traditional spacecraft, as it will be able to reconfigure its operations around any problems that it encounters.

The AMO spacecraft is currently in the early stages of development, but it has the potential to be used for a wide variety of missions, including exploration missions, scientific missions, and Earth observation missions. The AMO spacecraft could also be used to support commercial space operations.

Here are some of the potential benefits of using AMO spacecraft:

  • Reduced costs: AMO spacecraft could reduce the cost of space missions by automating tasks that are currently performed by ground control teams.
  • Increased efficiency: AMO spacecraft could be more efficient than traditional spacecraft, as they would be able to make decisions and take actions more quickly.
  • Improved safety: AMO spacecraft could be safer than traditional spacecraft, as they would be more resilient to failures.
  • New mission possibilities: AMO spacecraft could enable new types of space missions, such as missions to distant planets or missions that require a high degree of autonomy.

The European Space Agency (ESA) is also developing a new type of AI-powered satellite called the EarthCARE satellite. The EarthCARE satellite will be used to study clouds and aerosols. The AI on the EarthCARE satellite will be used to process and analyze data from the satellite’s instruments in real time. This will allow the EarthCARE satellite to provide more timely and accurate information about clouds and aerosols than traditional satellites.

The EarthCARE satellite is expected to be launched in 2024. It will be placed in a geostationary orbit, which means that it will remain stationary above a fixed point on Earth’s surface. This will allow the EarthCARE satellite to provide continuous monitoring of clouds and aerosols.

The EarthCARE satellite is expected to have a significant impact on our understanding of clouds and aerosols. This information will be used to improve weather forecasting, to study climate change, and to develop new air quality policies.

Overall, the development of AI-powered spacecraft and satellites is a major step forward in the space industry. AI has the potential to revolutionize the way that space missions are conducted and the way that we collect and use data from space.

Ubotica & IBM usher in a new era with autonomous AI-integrated satellite tech, revolutionising space missions and industry efficiency

Ubotica Technologies and IBM have formed a transformative partnership to revolutionize the space technology sector by integrating autonomous AI into satellite technology. This collaboration combines Ubotica’s space AI expertise with IBM’s advanced cloud infrastructure and watsonx.ai components, enabling more efficient application deployment on satellites. The adoption of Commercial Off-The-Shelf (COTS) technology has significantly reduced space computing costs, allowing real-time applications like forest fire detection and space debris removal.

Ubotica’s CogniSAT space AI compute platform, hosted directly on satellites, handles large data volumes from Earth Observation missions, streamlining data analysis and reducing reliance on ground systems. Their recent initiative, CogniSAT-6, represents a pivotal step toward enhancing satellite functionalities with an emphasis on optimizing technology for smaller satellites. Ubotica leverages IBM’s Cloud services to create a robust space AI platform on Low Earth Orbit (LEO) satellites, providing flexibility and growth opportunities through the collaboration. This partnership is set to reshape the satellite industry, making space technology more accessible and powerful.


AI for Satellite swarms

NASA engineers are actively developing the concept of satellite swarms with advanced communication capabilities, allowing these small spacecraft to work in unison and capture diverse datasets while extending communication networks. Artificial intelligence (AI) will be integral in the future use of satellite constellations, enabling functions such as autonomous healing and repositioning based on real-time onboard health and performance analysis.

Small satellites, or SmallSats, could communicate and coordinate with one another, employing machine learning algorithms to observe vital weather patterns from multiple angles and at different times. This collaborative approach has the potential to revolutionize weather and climate research, such as studying phenomena like Saharan dust affecting cloud formation.

Engineer Sabrina Thompson is developing software to enable SmallSats to identify high-value observation targets, adjust their positions, and optimize viewing strategies over time. These satellite swarms may utilize different configurations, including various orbits, to provide comprehensive observations of specific phenomena, greatly enhancing our understanding of weather system dynamics.



The integration of AI into space and satellite missions is ushering in a new era of exploration, discovery, and efficiency. As AI continues to evolve and advance, it opens the door to more ambitious and complex missions within our solar system and beyond. The collaboration between humans and AI in the final frontier is a testament to our unending quest for knowledge and understanding, and it promises to unlock the secrets of the universe in ways we could only dream of a generation ago.




References and Resources also include:






About Rajesh Uppal

Check Also

China’s Tiangong Space Station: A Beacon of Innovation and Global Collaboration

China’s Tiangong space station, also known as the Heavenly Palace, is a testament to the …

error: Content is protected !!