Edge Computing in Orbit: Why the Next Generation of AI is Moving into Space
Oct. 9 2026, Updated 12:33 p.m. ET
For decades, satellites have primarily served as the eyes of the global economy, collecting vast amounts of information about Earth and sending it back to ground stations for processing. But as the volume of satellite data grows, the economics of simply collecting more information are changing.
The next opportunity is to make satellites more intelligent.
By placing artificial intelligence and high-performance computing directly on spacecraft, satellites can increasingly analyze information while still in orbit, identifying what matters before transmitting it to Earth. The result is more than faster satellite imagery. It is a shift from selling raw data toward selling usable intelligence, creating new revenue opportunities while reducing costs for industries on Earth.
Turning Data Into a Commercial Product
Earth observation is already a significant commercial market. The OECD estimates that global space-related revenues reached roughly $550 billion to $600 billion in 2025, with satellite data and services supporting industries including transport, agriculture, energy and finance.
The economic opportunity in onboard AI is therefore not simply about putting more computing power into space. It is about increasing the value of every observation.
A conventional Earth observation system may capture an image, transmit it to Earth, process it and then deliver an analysis to a customer. An AI-enabled satellite can begin that process in orbit, identifying changes, classifying objects or detecting patterns and transmitting a smaller, more useful dataset.
That can reduce communications costs, lower processing requirements on the ground and allow customers to receive information closer to when an event occurs.
Orbitworks and the Intelligent Satellite
The UAE's Orbitworks provides an emerging commercial example. Its Altair constellation consists of 10 AI-enabled Earth observation satellites being assembled and integrated in Abu Dhabi. The satellites combine multiple sensors with onboard computing designed to process data in orbit and deliver analysis-ready intelligence with lower latency.
The commercial significance extends beyond the satellites themselves.
Altair is designed to support applications including agriculture, energy, infrastructure, disaster response and maritime activity. A shipping company could use satellite intelligence to improve route visibility, while agricultural businesses could identify changes in crops and insurers could assess climate-related risks more quickly.
Orbitworks is also positioning satellite manufacturing as an economic activity. Its Abu Dhabi facility is designed for production of up to 50 satellites annually, creating demand for engineering, electronics, software, advanced manufacturing and testing.
India Takes Computing Into Orbit
India offers another example of the emerging commercial model. Startup TakeMe2Space is preparing to launch MOI-1A, an orbital-computing satellite equipped with Nvidia processors that will analyze data in space rather than sending raw information back to Earth.
The company says it already has 23 customers across agriculture, mining, supply chains, and insurance. Its model is particularly revealing from a financial perspective because customers can upload AI models for targeted analysis, potentially reducing the amount of data that needs to be transmitted and processed on the ground.
If successful, this could turn computing capacity itself into a commercial space service, rather than making customers buy and operate their own satellites.
Europe Has Already Demonstrated the Technology
Europe provides an earlier demonstration of how onboard AI can improve the economics of Earth observation.
The European Space Agency's Φ-sat-1, launched in 2020, was designed to use AI onboard to filter satellite imagery before transmission. One application involved identifying images affected by cloud cover, allowing unsuitable data to be discarded before it consumed valuable communications capacity. ESA describes this as a way to improve the efficiency of sending Earth observation data back to Earth.
ESA is now developing the next stage through EOCognitiveLab, which will explore multiple satellites working together through onboard processing and inter-satellite communications.
The Value Is Down on Earth
The biggest economic effects of orbital AI may ultimately occur outside the space industry.
In agriculture, faster satellite analysis can help businesses make decisions about irrigation and crop conditions. In maritime trade, near-real-time information could help logistics companies and ports monitor shipping activity. For infrastructure and insurance, faster detection of changes can potentially reduce the time and cost involved in assessing risk.
This creates a different economic model for space. Instead of value being concentrated in building and launching spacecraft, more value can be generated by the information those spacecraft produce and the decisions that information enables.
Countries investing in the sector can capture value at multiple stages, from manufacturing spacecraft and developing AI systems to operating constellations and selling data services.
Orbitworks' Altair, India's emerging orbital-computing market, and Europe's continuing investment in onboard AI all point toward the same shift.
The significance of intelligent satellites is not simply that computers are moving into orbit. It is that economic value is moving with them.
Never Miss an
Daily updates from the heart of Hollywood, right to your inbox
By entering your email and clicking Sign Up, you’re agreeing to let us send you customized marketing messages about us and our advertising partners. You are also agreeing to our Terms of Service and Privacy Policy.
