Space technology startup Satlyt has raised $8 million in new funding to develop artificial intelligence computing systems designed to operate directly on satellites. The company is working on technology that could allow spacecraft to process large amounts of information in orbit instead of sending all raw data back to Earth.
The new funding comes as the space industry increasingly looks toward artificial intelligence for processing satellite imagery, communications data and other information generated by spacecraft.
AI Computing Moves Closer to Orbit
Satellites generate enormous amounts of data from cameras, sensors and other instruments. Traditionally, much of that information has to be transmitted to ground stations for processing.
This approach can create delays and requires significant communications bandwidth.
Satlyt is developing systems that can perform artificial intelligence calculations directly on spacecraft. This approach is commonly described as edge computing in space.
Instead of transmitting every piece of raw information to Earth, a satellite could analyze the data in orbit and send only the most important results.
Why Processing Data in Space Matters
Satellite operators increasingly collect high-resolution images and other complex data. As the number of spacecraft increases, the amount of information generated in orbit is also growing.
Sending all of that data to Earth can be expensive and time-consuming.
AI systems operating on satellites could potentially identify objects, detect changes, classify images or analyze other sensor information before the data reaches a ground station.
This could reduce the amount of information that needs to be transmitted and allow operators to respond more quickly to events detected by spacecraft.
Satlyt's New Funding
The company has raised $8 million in new funding to expand development of its space-based AI computing technology.
The investment will support engineering, hardware development and the deployment of technology designed specifically for the demanding conditions of space.
Satlyt was founded by former technology and space-industry professionals with experience in developing products for large-scale computing and spacecraft systems.
Space Hardware Creates Unique AI Challenges
Running artificial intelligence systems in space is considerably different from operating AI infrastructure on Earth.
Satellites have limited power, limited physical space and strict weight requirements.
Electronic components must also operate in an environment exposed to radiation and extreme temperature changes.
For these reasons, space-based AI systems need to be smaller, more power-efficient and more resistant to environmental conditions than many conventional data-center systems.
AI Could Reduce Satellite Data Bottlenecks
One of the biggest potential advantages of orbital AI is reducing the amount of data that must travel between space and Earth.
A high-resolution satellite camera can produce large quantities of information. Transmitting every image in full resolution can consume valuable communication capacity.
An AI processor could analyze images onboard and determine which information is most relevant.
For example, an Earth-observation satellite could potentially identify changes in a particular area and prioritize those observations for transmission.
The technology could also be useful for scientific missions, weather monitoring and other applications where rapid data analysis is important.
Growing Interest in Space-Based Computing
Satlyt's investment comes during a broader increase in interest in space-based computing.
Technology companies and space startups are exploring whether some computing workloads can be moved closer to the point where data is generated.
Google's Project Suncatcher is investigating another version of the idea, involving AI processors operating in orbit and eventually networking satellites together.
However, Satlyt's approach focuses on bringing AI processing capabilities directly to satellites and spacecraft rather than building traditional data centers in orbit.
Space AI Could Support Earth Observation
Earth-observation satellites are among the most obvious potential users of onboard AI.
Modern spacecraft can capture detailed information about agriculture, forests, oceans, infrastructure and changes occurring on the Earth's surface.
AI algorithms can help analyze this information and identify patterns that may otherwise require significant computing resources on the ground.
Processing some information directly in space could allow satellite operators to make faster decisions about which data should be transmitted.
Power Efficiency Is Critical
AI computing normally requires substantial electrical power, making energy efficiency particularly important for satellites.
Unlike terrestrial data centers, spacecraft cannot simply connect to a large electricity grid.
Satellites typically depend on solar panels and batteries, which limits the amount of computing power available at any given time.
Satlyt's technology therefore needs to balance computing performance with the limited power available aboard spacecraft.
The Space Computing Market Is Expanding
The number of commercial and government satellites in orbit has increased rapidly in recent years.
Large satellite constellations are being developed for communications, Earth observation and other applications.
As more spacecraft are deployed, demand for efficient onboard computing is expected to increase.
Artificial intelligence could become an important part of this next generation of satellite technology.
From Satellites to Autonomous Spacecraft
Onboard AI could eventually enable spacecraft to make more decisions without waiting for instructions from Earth.
A satellite equipped with advanced computing could analyze sensor information, identify important events and prioritize communications.
More advanced systems could potentially support autonomous navigation, scientific experiments and other spacecraft operations.
However, autonomous space systems also require careful testing and strong safeguards because mistakes in orbit can be difficult and expensive to correct.
A New Direction for Artificial Intelligence
The development of space-based AI demonstrates how artificial intelligence is moving beyond traditional computers, smartphones and data centers.
AI processors are increasingly being integrated into vehicles, industrial machines, robots and other devices that operate at the edge of networks.
Satellites represent another frontier where local processing could provide significant benefits.
Satlyt's $8 million funding round gives the company additional resources to develop this technology and move AI computing closer to the source of satellite-generated data.
As satellite constellations continue to expand, the ability to analyze information directly in orbit could become an increasingly important part of the global space technology industry.
Journalist: Vijay Singh