The Galileo Project: Someone Finally Builds the Sensors, and We Finally See How Hard It Is
A Harvard professor did what we had been asking for eighty years: his own instruments, his own data, peer-reviewed publications. And also an expedition to the Pacific that may have been searching in the wrong place because of a truck.
The Galileo Project, founded in July 2021 by Harvard astronomy professor Avi Loeb with private funding, is the first systematic sky-observation program with its own instruments devoted to this subject. It works along three lines: interstellar objects, continuous multispectral observation and a search for artifacts. It has published the description of its observatory and its results, including false alarms. Its most striking conclusions (the spherules recovered in the Pacific in 2023 and the speculation about the third interstellar object in 2025) have been rebutted with solid arguments by outside specialists.
Executive summary of the case file. Everything essential before you dig deeper.
Fact sheet
OSINT - Traceability
Friends, we have been complaining about the same thing for eighty years: that science won’t deign to study this. That there’s the stigma, that there are careers at stake, that nobody wants to sign a paper about lights. Well then: in July 2021, a Harvard professor of astronomy set up a project to do exactly that, with instruments, with private funding and with peer-reviewed publications.
And from there, the story gets complicated in a way that says far more about us than about the sky.
Getting to the point: what exactly this project does
First, cut the noise. The Galileo Project is an observation program with equipment, with a protocol and with published papers, not a hobbyist association with a video channel. Its approach is so sensible that it almost offends that nobody thought of it before:
- Don’t argue over other people’s videos. No analyzing leaked military recordings or frames of dubious provenance. If you want good data, you buy the sensors and generate it yourself, with known calibration and the full record saved.
- Watch the whole sky, all the time. A wide-angle infrared camera running continuously, plus radio, audio, magnetic field and weather at the same site. When something shows up, you have its signature in several bands at once.
- Study interstellar objects. Bodies that enter the solar system from outside are, beyond any dispute, samples from another stellar system. Studying them is pure astronomy, with no need to invoke anything.
- And publish the noise too. This is what I like most. The observatory’s first papers are, to a large extent, catalogs of birds, planes, drones and balloons. In other words: the boring part, which is the part that gives the other one credibility.
A project that publishes how many birds it has mistaken for something strange deserves a hearing when it says something doesn’t fit. One that publishes only what doesn’t fit does not.
Oumuamua arrives
The first known interstellar object passing through the solar system is detected. Its elongated shape, its variable brightness and a non-gravitational acceleration with no visible cometary tail set off a legitimate scientific debate about its nature.
The book
Avi Loeb publishes a popular-science book arguing that the hypothesis of an artificial origin for that object deserves to be taken seriously. The planetary science community reacts harshly.
The Galileo Project is born
Loeb announces a systematic observation program with private funding and a base at Harvard, with three lines: interstellar objects, instrumental observation of the sky, and a search for artifacts.
The Space Command memo
A scientific official of U.S. Space Command signs a note confirming that the velocity estimate for a fireball recorded in 2014 over the Pacific is accurate enough to indicate an interstellar trajectory. It is the basis for the later expedition.
The Pacific expedition
A team sweeps the seabed off Papua New Guinea with a magnetic sled looking for remains of that fireball. It recovers metallic spherules of submillimeter size.
The controversy
Loeb maintains that the composition of some spherules is anomalous. Meteorite specialists reply that it is consistent with coal ash and industrial contamination, a material extremely abundant on the seafloor because of shipping routes.
The location problem
An independent study argues that the seismic signal used to narrow down the fireball's impact zone did not come from the impact, but from a truck driving on a nearby road. If so, the wrong site was searched.
Third interstellar object
The third known interstellar visitor is discovered. Loeb and collaborators publicly raise the possibility of an artificial origin. Later observations show cometary activity and the astronomical community treats it as what it appears to be: a comet.
The good, and there is a lot of it
Let’s be fair: here we hand out blows to everyone, but in order.
The Galileo Project has done three things nobody else had done. First: it has built and put into operation a dedicated observatory, with its own calibrated data, and has published the description of the system so anyone can replicate it. Second: it has brought the subject into scientific journals through the front door, that is, with methodology and peer review, not with a press conference. And third, the most important: it has shown that it can be done. That a professor can devote himself to this without having his chair taken away.
That last point is worth a fortune. For decades, the argument that no serious scientist could touch the subject without ruining his career was true, and it worked as a self-fulfilling prophecy: nobody studied it, so there was no data, so there was nothing to study. Loeb has broken the loop. You can disagree with him on almost everything (I do on quite a lot) and still give him that.
And the bad, which is also a lot
Now the other half, and here we have to be tough because the project brought it on itself.
The spherule problem
The 2023 expedition to the Pacific is a case study in how not to do this. The seabed was swept with a magnetic sled and hundreds of metallic beads smaller than a millimeter were recovered. Loeb maintained that the composition of some was anomalous and consistent with an origin outside the solar system.
The problem is this, and it is devastating for how prosaic it is: the ocean floor is full of metallic spherules. They have been there for more than a century, and nobody brought them from another star. Steamships brought them. The coal ash dumped and thrown overboard along shipping routes over a hundred and fifty years produces exactly this kind of iron-rich bead. Several meteorite specialists pointed it out immediately, with comparative analyses.
And on top of that there is the matter of location. The search zone was narrowed down using a seismic signal recorded at a nearby station, attributed to the fireball’s entry. A later analysis argued that the signal was consistent with a truck driving on a road close to the seismograph station. If that is correct (and the argument is solid), the expedition searched the wrong place.
A truck. In Astro language: we went down to the bottom of the Pacific to look for remains of another star by following the vibration of a truck on its way to deliver goods. No script tops that.
The announcement problem
And then there is the manner. Loeb has developed a habit of communicating very strong hypotheses in advance, on blogs and in interviews, before the analysis is closed. That is no minor sin: when the headline comes out before the data, the later correction never catches up with the headline. The third interstellar object, discovered in July 2025, is the most recent example: the possibility that it was an artifact was raised publicly, and later observations showed a comet doing comet things. The retraction existed. The headline had already gone around the world.
Oumuamua, or the legitimate debate that became a trench
To understand why this project exists we have to go back to October 2017, and do so honestly, because the story is told badly by both camps.
That month the first known object entering the solar system from outside and leaving again was detected. A small body, with an unambiguous hyperbolic trajectory, already moving away when we discovered it. It was observed for a few weeks and left forever. And it left a list of oddities that are nobody’s opinion: a light curve implying a very elongated or very flattened shape, uncommon among known minor bodies; no detected cometary tail; and an acceleration beyond gravity as it moved away from the Sun.
That last point is the knot. Comets accelerate like that because ice sublimates and pushes, and that produces a tail. Here there was a push and no tail was seen. Natural explanations were proposed: sublimation of molecular hydrogen, a porous, spongy body, radiation pressure on an extraordinarily thin object. All plausible, none settled. Loeb added another: that a very thin object pushed by light fits well with a solar sail, and a solar sail is an artifact.
This is where I get off both bandwagons. Raising the hypothesis was perfectly legitimate and that is how science is done: you propose the strange, try to rule it out, and see what holds. What no longer seems legitimate to me is the next step, the popular-science book with the artifact thesis turned into the central argument, aimed at a public that has no way to evaluate the light curve. Nor do I think the profession’s reaction was right, which in many cases was personal contempt instead of technical refutation.
Result: a debate that could have been excellent ended up as two trenches. And the project was born out of those trenches, which explains a good deal about its character: it was born angry.
One thing I do consider a finding, and it isn’t the one that made the papers
I’m going to defend the project on something almost nobody credits it with, because it isn’t showy.
When you set up a wide-angle infrared camera watching the sky twenty-four hours a day for months, and automatically classify everything that passes, you get a data point of enormous value that nobody had: the base rate. How many birds, how many planes, how many drones, how many balloons and how many unclassified objects per unit of time and of sky area.
That is exactly what this field has lacked for eighty years. Without a base rate, a sighting means nothing: you don’t know whether it is one in a thousand or one in a million. With a base rate, for the first time, you can say something like “out of so many detections, so many remain unclassified,” and that fraction can be compared across sites, across seasons and across instruments.
It doesn’t make the headlines because there is no possible headline in “we have counted birds with rigor.” But it is the only thing in all of this that will still be useful in twenty years. Announcements age terribly. Catalogs don’t.
And in that the project resembles, much more than it would like to admit, something that had been working in a Norwegian valley for forty years without anyone paying attention: set up the apparatus, endure the boredom and publish the statistics. We covered it in the Hessdalen case file. The difference is the budget and the press. The method is the same.
Why this case file matters even though there is no UFO in it
Because it is the field test of an idea we have defended for years here at Broken UFO: that the problem with this subject was never a lack of phenomenon, but a lack of our own data.
All of 20th-century ufology was built on other people’s material: other people’s photographs, other people’s testimonies, documents leaked by others, declassified military videos arriving two years late and without metadata. Arguing over material you don’t control is a dead end, and that is why we have been in the same place for eighty years. Galileo is the first serious attempt to leave by the other door: if you want to know what is up there, set up a sensor and look. The results will be disappointing most nights. Perfect. Disappointing data are data.
And it is also a warning about the other danger, the one from inside. When a project with good method lets its spokesman run ahead of its laboratory, the whole method gets contaminated in the public’s eyes. The project is not Loeb, but for the headlines it is. That is a structural risk, and we have already seen it with the opposite sign in the institutional chapter: there the bureaucracy braked too much, and here enthusiasm accelerates too much.
Eighty years asking for science to take this on. When it finally does, it turns out that science is slow, expensive, boring and wrong in public. Welcome. That is exactly how it works.
OSINT – Source traceability
An express distinction is drawn between the project's method, which this site values, and its most striking conclusions, which are not taken as sound. No confirmed finding is attributed to the project.