The trouble with listening for the faintest events in the universe is that your own instrument often drowns them out. However, physicists at Imperial College London have now shown that a prototype quantum sensor can still pull out a real signal. This happens even when noise appears to erase each measurement.

That result, reported in Nature, tackles one of the central technical problems facing a new class of detectors known as long-baseline atom interferometers. These devices are being developed to search for gravitational waves in a frequency range current observatories cannot reach. Moreover, they aim to look for signs of ultralight dark matter.

Instead of using mirrors and laser beams in the usual way, atom interferometers use lasers to split and recombine the wave-like motion of atoms. Tiny changes in that motion can reveal equally tiny disturbances in space, time, or the atoms themselves.

The promise is huge. So is the noise.

The Imperial team built a tabletop prototype using two clouds of ultracold strontium-87 atoms, held 1 millimeter apart and interrogated by the same ultrastable clock laser. This setup was meant to mimic, in miniature, the conditions expected in much larger detectors now being planned.

In those future systems, researchers hope to compare atom clouds separated by distances closer to a kilometer, or even more. A passing gravitational wave could slightly change the effective travel time of laser light between the two locations. A dark matter field, in theory, could slightly shift atomic energy levels. Either way, the signal would be tiny.

The problem is that the laser itself carries phase noise, fluctuations that can be far larger than the effect anyone is trying to detect. Additionally, on a single interferometer, that noise can swamp the measurement completely.

To get around that, physicists have long proposed a differential approach. If two interferometers are driven by the same laser, much of the laser noise should be shared between them. Consequently, compare the two, and the common noise should cancel, at least in principle. That idea sits at the heart of proposed next-generation detectors. Nonetheless, proving it under realistic experimental conditions had remained an open challenge.

Dr Charles Baynham, co-lead of the Ultracold Strontium Laboratory at Imperial College London, said, “We’ve known for a long time that quantum sensors can help us understand the universe, but it’s only recently that it’s become possible to build them with the resolution needed.

“We’re immensely proud of our team’s efforts to make these sensors a reality – I can’t wait for the day when signals from an atom are telling us about a black hole that merged millions of years ago.”

To push the test hard, the researchers deliberately added large amounts of artificial phase noise, far more than their clock laser naturally produces. The goal was to imitate the conditions expected in long-baseline detectors. There, noise control becomes much tougher.

The effect on the individual instruments was dramatic. Their normal interference fringes vanished into the noise. Looked at one by one, the interferometers had become useless.

But when the two were analyzed together, the buried signal reappeared.

Across 56,623 shots taken over 61.9 hours, the team found that the differential measurement continued to operate at the standard quantum limit, the fundamental noise floor expected from the finite number of atoms being measured. In other words, adding several radians of laser phase noise did not produce a statistically significant increase in the differential phase noise.

That matters because it suggests the cancellation worked as intended. The researchers also used Monte Carlo simulations matched to the experiment. They found that the measured behavior agreed with the statistical model, a useful check that the system was really performing at that limit.

The group then added a second test. They injected a controlled oscillating signal into one interferometer, designed to mimic the sort of time-dependent effect a gravitational wave or dark matter field might create. Even in the high-noise regime, where neither interferometer alone retained usable phase information, the differential analysis still recovered the signal.

That proof of principle is important because atom interferometers are being developed for a part of the gravitational-wave spectrum that remains largely unobserved. Ground-based laser interferometers such as LIGO, Virgo and KAGRA are most sensitive at higher frequencies. The space-based LISA mission will target much lower ones. Between them sits a middle band, roughly from 0.1 to 10 hertz, that could contain signals from intermediate-mass black hole mergers and the early inspiral of solar-mass systems.

A kilometer-scale atom interferometer could, in principle, reach into that band. Moreover, the same kind of detector could also look for bosonic dark matter fields with masses around 10^-15 electronvolts, according to the paper.

That is where collaborations such as AION, led by Imperial and involving researchers from Birmingham, Cambridge, Liverpool, King’s College London, Oxford and STFC Rutherford Appleton Laboratory, come in. This program is part of a broader international push. It also includes MAGIS at Fermilab and the proposed Atom Interferometry CERN Experiment, or AICE.

Dr Richard Hobson, co-lead of the Ultracold Strontium Laboratory at Imperial, said, “We have taken some of the most precise instruments ever built—atomic clocks and atom interferometers—and shown that they can be repurposed to open entirely new windows onto the invisible parts of our Universe.

“Our current experiment is just a prototype, but scaling it to a full-scale facility at laboratories such as CERN or Fermilab will allow us to tackle some of the deepest mysteries in physics, including the nature of dark matter.”

The work does not mean those giant detectors are around the corner. The authors are clear that major hurdles remain, including stronger cold-atom sources, longer baselines, tighter control of systematic shifts, larger momentum transfer, and new quantum-state techniques such as squeezing. Still, the result clears one of the more basic and stubborn questions in the field. Specifically, whether laser noise rejection really works when conditions get messy.

For physics, the immediate importance is confidence. The study gives experimental backing to a measurement strategy. This is what future long-baseline atom interferometers will depend on if they are to search for mid-frequency gravitational waves and ultralight dark matter.

For quantum technology more broadly, the work shows that tools developed for atomic clocks and precision sensing can be combined in ways that stay useful even when noise is severe.

Over time, techniques refined for these fundamental-physics detectors could also influence precision navigation, geodesy, and resource exploration. In these fields, extremely sensitive measurements matter.

Research findings are available online in the journal Nature.