For over two decades, millions of personal computers around the world have joined forces to scan radio signals from the cosmos, hoping to identify a trace of technological life beyond our planet. Today, about one hundred signals deemed worthy of interest, resulting from this tremendous collective mobilization, are undergoing thorough examination by the planet's most imposing radio telescope.
Launched in 1999 and active until 2020, the SETI@home project offered anyone interested the chance to participate in the search for extraterrestrial intelligence by harnessing the unused computing power of their own computer. This program processed recordings from the Arecibo radio telescope, isolating no fewer than twelve billion candidate signals. This effort brought together a community of unprecedented scale, turning simple screensavers into genuine scientific instruments.
The five-hundred-meter (~1640-foot) aperture FAST radio telescope.
Image Wikimedia Supercomputers were later added to sort through this colossal mass of information. Dedicated algorithms gradually reduced the list to one million, then to one thousand signals. Each of these was manually inspected by scientists, allowing the selection of about a hundred particularly promising cases warranting complementary observation.
Since July 2025, it is the FAST radio telescope in China, with its antenna five hundred meters (~1640 feet) in diameter, that has been tasked with this meticulous verification. The original instrument, Arecibo, which provided the initial data, has been out of service since its collapse in 2020.
Public enthusiasm far exceeded the founders' expectations. While they hoped to gather a few tens of thousands of users, over two million people joined the project in the first year alone. This exceptional participation made it possible to explore billions of stars in our Galaxy with a sensitivity never before achieved for this type of study.
The 12 billion signals detected by SETI@home have been reduced to 100 candidates for follow-up observations.
Credit: Robert Sanders/UC Berkeley.
No confirmed extraterrestrial emission has been formally identified to date, but the initiative sets a new standard for future searches. The expertise thus developed opens the door to new projects, possibly equipped with more powerful technologies.
A question persists among some researchers: could the amassed data still conceal a clue that has so far gone unnoticed? With advances in artificial intelligence and distributed computing, a comprehensive reanalysis might one day be possible. This possibility keeps alive the idea that the work of millions of volunteers may not yet have yielded all its substance.
The search for narrowband radio signals
In the pursuit of traces of a technological civilization, scientists frequently target 'narrowband' radio signals, meaning those concentrated on a very precise frequency. Within the natural cosmic noise, emissions are typically broad and spread over an extended spectrum. An artificial emission, designed to communicate over long distances, would have a better chance of being perceived if it were sent on a narrow and stable band.
This method helps to more easily distinguish a potentially intelligent emission from the many natural astrophysical phenomena, such as pulsars or stellar flares. The SETI@home project was precisely designed to spot these tiny energy peaks on a specific frequency, coming from a given point in the sky. Algorithms continuously scanned the data to detect these anomalies.
However, the vast majority of signals captured actually come from radio interference generated by terrestrial human activity. Satellites, radars, and even some electronic devices can produce emissions that, to sensitive instruments, resemble extraterrestrial signals. The main challenge is therefore to filter out this terrestrial 'noise' to keep only the truly intriguing candidates.
This work establishes detection thresholds: if a civilization emitted a sufficiently powerful and targeted signal in the observed areas, projects like SETI@home should have intercepted it. The absence of a positive detection thus allows us to state that no emission of this type, beyond a certain power threshold, has been detected in the examined portion of the Galaxy, thereby refining the field of possibilities.
Distributed computing for science
Distributed computing allows the use of the power of many networked computers to process a problem too large for a single machine. Each participant installs a small software running in the background, examining batches of information when the computer is idle. This approach makes feasible simulations or processing that would otherwise require extremely expensive supercomputers.
SETI@home is among the best-known illustrations of this principle, having generated an immense virtual computing capacity from domestic resources. This model has been applied to other scientific fields, such as biology for the study of protein folding or climatology for modeling atmospheric changes. It thus democratizes research by allowing everyone to contribute directly to advancements.
Its main advantage lies in its ability to handle astronomical volumes of data at a relatively low cost. Projects can progress faster by mobilizing an international community of volunteers, without requiring an excessive central infrastructure. This collaborative method transforms personal computers into a valuable collective scientific resource.