Can Quantum Sensors Beat the Noise? Inside the Research Changing Precision Measurement

For decades, researchers have investigated whether carefully engineered quantum states can make measurements more precise than conventional techniques allow.

In 2025 and 2026, several research teams advanced this effort along different paths. One team experimentally generated strongly squeezed microwave states and used them to measure small displacements. Other researchers developed protocols based on quantum grid states and spin-dependent squeezing, seeking better performance under specified measurement constraints.

These studies address a common question: Can scientists engineer the quantum state of a sensor so that a weak signal becomes easier to distinguish from noise?

The answer is promising, but it depends on what kind of noise is present, which physical quantity is being measured, and how the sensor is operated.

This report compares three important research directions:

1. Experimentally generated squeezed states: A superconducting microwave cavity produced squeezed states reaching 14.6 decibels, which were subsequently used in a displacement-sensing experiment.
2. Stabilized GKP states: A theoretical protocol uses continuously stabilized quantum grid states to estimate two components of an unknown displacement.
3. Spin-dependent squeezed states: A separate theoretical and numerical study investigates optimal displacement sensing and faster preparation of specially engineered quantum states.

These approaches are related, but they are not interchangeable. One has demonstrated a physical squeezing-and-sensing experiment; the others establish proposed protocols and modeled performance that require further experimental validation.

Understanding the difference is essential for evaluating the actual progress of quantum technology.

Noise is not one single phenomenon.

In an ordinary electronic instrument, noise can arise from thermal fluctuations, electrical interference, imperfect components, and limitations in the measuring equipment.

Quantum systems can experience those same kinds of disturbances, but they also have quantum-mechanical uncertainty and state-dependent measurement statistics.

A quantum sensor may be affected by several distinct sources of uncertainty:

• Quantum uncertainty: The uncertainty inherent in the state being measured.
• Environmental noise: Unwanted interactions with the surrounding environment.
• Decoherence: The loss of useful quantum coherence through environmental interactions.
• Control errors: Imperfect laser pulses, microwave drives, or other operations.
• Readout errors: Incorrect or incomplete measurements of the final state.
• Technical drift: Slow changes in the instrument that can imitate or obscure the signal.

The solution depends on which limitation dominates.

If uncertainty in a particular quadrature is the main problem, squeezing may help. If the sensor is vulnerable to small displacement errors, a grid-state encoding may offer a way to detect or correct those errors. If state preparation is too slow, an improved control protocol could reduce the time during which certain disturbances accumulate.

The crucial point is that reducing one source of uncertainty does not automatically eliminate every other source of noise.

A sensor that performs exceptionally well in a controlled measurement can still struggle when moved into a different environment or asked to detect a different kind of signal.

The standard quantum limit, or SQL, is a widely used reference for measurement precision.

In many settings involving independent probes, the uncertainty of an estimate decreases approximately as:

where N represents the number of independent resources or measurements under the chosen convention.

Increasing the number of measurements can therefore improve precision, but with diminishing returns: reducing the uncertainty by a factor of two may require roughly four times as many independent measurements.

Quantum metrology investigates whether specially prepared quantum states and suitable measurements can do better than this scaling in a particular sensing task.

Under idealized conditions, some quantum strategies can achieve Heisenberg scaling, often written as:

These expressions are benchmarks, not universal formulas that apply identically to every experiment. The relevant resource count, noise model, measurement task, and assumptions matter.

In real systems, the advantage can be reduced or lost because of decoherence, finite state-preparation quality, imperfect control, or the time and energy required to prepare the quantum state.

This is why the most convincing research does more than report an impressive theoretical limit. It also explains the conditions under which that limit applies and tests how the proposed strategy behaves under realistic constraints.

A team led by Yanyan Cai and colleagues reported an experimental approach to generating and amplifying squeezed microwave states in a superconducting cavity.

The paper, titled Quantum squeezing amplification with a weak Kerr nonlinear oscillator, appeared in Nature Communications. It was published online in December 2025 and appeared as part of volume 17 in 2026.

The researchers addressed a practical problem: producing strongly squeezed states often requires substantial nonlinear interactions, which can introduce additional decoherence.

Instead of relying on a strong nonlinearity, the team engineered the evolution of a superconducting microwave cavity with a weak Kerr nonlinearity.

The results included three key achievements:

• Experimental generation and amplification of squeezed microwave states.
• A maximum reported squeezing level of 14.6 dB.
• An experimental demonstration that the generated states could improve small-displacement measurement precision relative to the standard quantum limit.

How the cavity produces squeezed states

The experiment used a superconducting microwave cavity coupled to an ancillary superconducting qubit.

The cavity acts as a quantum oscillator. The qubit helps characterize the cavity state, while the cavity's nonlinear response provides a way to manipulate the quantum fluctuations.

In an ideal linear harmonic oscillator, the energy levels are evenly spaced. A Kerr nonlinearity makes the energy levels slightly unequally spaced, allowing the state to evolve in ways that can produce squeezing.

The team carefully engineered an off-resonant microwave drive to control that evolution.

They also used a technique called Trotterization, which approximates a desired evolution by applying a sequence of simpler operations. In this case, alternating displacement operations helped build up squeezing over successive steps.

The researchers characterized the resulting states using measurements of their Wigner functions. A Wigner function is a phase-space representation of a quantum state that can reveal its structure, including the anisotropic uncertainty characteristic of squeezing.

As the sequence progressed, the state became compressed in one phase-space direction and expanded in the complementary direction.

The team reported a maximum squeezing level of 14.6 dB.

What does 14.6 dB mean?

Squeezing is often expressed in decibels relative to a reference noise level.

The decibel value is logarithmic, so it should not be interpreted as a percentage improvement or as a direct measure of how much better every possible measurement becomes.

The reported 14.6 dB describes the squeezing degree of the generated state under the convention used in the paper. It does not mean that every source of noise in the device was reduced by 14.6 dB.

The significance is that the experiment created a strongly nonclassical microwave state using a controlled sequence of operations, then used the state in a measurement task.

The sensing experiment

The team tested the squeezed states by measuring small displacements along one quadrature.

They used a parity measurement to map information about the displacement onto the state of an ancillary qubit. Repeated measurements were then used to estimate the displacement.

The researchers reported that displacement-estimation uncertainty decreased as they generated states with greater squeezing, and that their measurements surpassed the standard quantum limit in the experiment.

This is an important distinction from a paper that only predicts a theoretical advantage. The team physically generated the states and tested them in a displacement-measurement protocol.

However, the demonstration was specific to the system and sensing task used in the experiment. It did not establish that every force sensor, atomic clock, or dark-matter detector would receive the same advantage.

What limits the approach?

The researchers identified practical constraints on the achievable squeezing, including limitations related to the displacement-drive amplitude and the approximations used to describe the nonlinear cavity.

They also observed that nonlinearity could cause the state to twist and stretch in phase space as the squeezing increased.

This illustrates a recurring theme in quantum engineering: an operation can be useful up to a point, but the same physical effects that enable it can also introduce limitations.

The experiment is a meaningful advance because it demonstrates both state preparation and a sensing benefit — not because it eliminates the engineering challenges of quantum measurement.

A second approach comes from the research paper Quantum Sensing of Displacements with Stabilized Gottesman-Kitaev-Preskill States, published in PRX Quantum on April 1, 2026.

The authors — Lautaro Labarca, Sara Turcotte, Alexandre Blais, and Baptiste Royer — investigated how a quantum state originally developed for bosonic quantum error correction could be used for precision sensing.

Their proposed sensor estimates two quadratures of an unknown displacement and uses continuous stabilization to maintain a useful state while gathering information. I cover this paper in depth in my related article, Quantum Sensing With GKP States.

Why use a quantum grid state?

Gottesman-Kitaev-Preskill states, usually called GKP states, have a periodic grid-like structure in the phase space of an oscillator.

They were developed as a way to encode quantum information in a bosonic system while allowing certain small displacement errors to be identified and corrected.

The same property that makes these states interesting for error correction can also be useful for sensing: their structure provides a way to distinguish different displacements.

An ideal GKP state is not physically achievable because it requires infinitely sharp features and unbounded energy. Practical systems must use finite-energy approximations, which have imperfect grid structures and are more vulnerable to some errors.

The sensing problem is therefore not simply to create an ideal grid state. It is to design a protocol that continues to work when the state is imperfect and the system is noisy.

What the researchers proposed

The team adapted a stabilization technique known as the small-big-small protocol.

In broad terms, the method uses controlled interactions with an auxiliary qubit to steer the oscillator toward a desired grid state. The proposed sensing protocol incorporates information extraction into the stabilization process.

This gives the approach a feature the researchers call backaction evasion: the sensor is designed to continue operating without a complete reset after each measurement cycle.

That is potentially valuable for signals that change over time. A sensor that must repeatedly stop, reset, and prepare itself may miss information during the gaps.

What did the paper establish?

The authors showed theoretically that their proposed strategy can approach the multivariate quantum Cramér–Rao bound for two-quadrature displacement sensing.

They also provided numerical simulations indicating that the protocol can surpass the Gaussian limit for displacement sensing with prior information, even in the presence of modeled realistic noise.

The quantum Cramér–Rao bound is a precision benchmark derived from quantum estimation theory. The multivariate version matters when estimating several related parameters together, rather than treating each parameter as an independent single-variable problem.

The Gaussian limit, meanwhile, applies to a defined class of Gaussian-state strategies under the assumptions used in the paper. Beating that benchmark does not mean the protocol beats every possible sensor in every setting.

Most importantly, the paper does not report an experimental demonstration of the complete proposed sensor. Its conclusions are based on theoretical analysis and numerical modeling.

The result is a research blueprint with quantified advantages under specified assumptions. Experimental work is needed to determine whether those advantages survive the practical constraints of a real device.

A third direction is explored in Optimal Displacement Sensing with Spin-Dependent Squeezed States, published in PRX Quantum on June 23, 2026.

The paper includes researchers affiliated with the University of Amsterdam and collaborators in Australia and India. I cover this paper in depth in my related article, Optimal Quantum Sensing With Spin-Dependent Squeezing.

The researchers examine how to connect the quantum spin degrees of freedom in a system with the motion of an oscillator.

The goal is to construct a state in which the oscillator's squeezing depends on an auxiliary spin. This relationship can make the final measurement more informative about an unknown displacement.

How spin-dependent squeezing works

A spin is an intrinsic quantum property, not simply a miniature object spinning around an axis.

In quantum-control experiments, a spin can be used as an auxiliary degree of freedom. Researchers can prepare it in a chosen state, couple it to an oscillator, and measure it to learn about the oscillator.

Squeezing reduces uncertainty in one oscillator quadrature while increasing uncertainty in the complementary quadrature.

In spin-dependent squeezing, the oscillator's state depends on the spin configuration. This allows a carefully designed sequence of operations to map information about a displacement onto a measurable spin outcome.

The protocol does not physically enlarge the unknown displacement. It engineers the quantum system so that the displacement can produce a more distinguishable measurement result.

The reported numerical result

The authors numerically demonstrated a state-preparation protocol that achieved 8.7 dB of spin-dependent squeezing.

They reported that this preparation was 15 times faster than a standard comparison approach based on second-order sidebands in trapped ions.

The paper also analyzes how the proposed states can reach the Heisenberg limit within its theoretical framework.

These are meaningful results, but their evidence level matters. The 8.7 dB result is a numerical demonstration, not a reported experimental measurement of the complete sensing protocol.

The study does not demonstrate a finished quantum computer, a new dark-matter detector, or a universal improvement in measurement precision.

Instead, it provides a protocol that can be tested experimentally to determine whether the predicted advantages survive control errors, decoherence, and other practical constraints.

The studies share a broad objective — improving the extraction of information from quantum systems — but their methods and evidence differ.

The comparison shows why it is misleading to rank quantum sensors using a single number.

A squeezing value alone does not tell us the sensor's final precision. A theoretical bound does not establish real-world performance. A fast preparation method is valuable only if the complete protocol remains accurate after noise and control overhead are included.

A fair comparison must consider the same sensing task, the same definition of resources, the same measurement time, and comparable noise assumptions.

These terms often appear together in quantum-sensing articles, but they describe different things.

Squeezing describes a property of a quantum state: uncertainty is reduced in one quadrature at the expense of increased uncertainty in another.

The Heisenberg limit refers to a precision-scaling benchmark achievable under specified idealized conditions and resource assumptions. It is not a universal guarantee for real instruments.

The Gaussian limit is a benchmark for a particular family of Gaussian-state sensing strategies. It is meaningful only when the measurement task and assumptions are clearly defined.

A non-Gaussian state may outperform a Gaussian strategy for one sensing problem without outperforming all conceivable strategies.

Likewise, a state can have a high squeezing value but perform poorly if it is difficult to prepare, quickly loses coherence, or is measured with insufficient accuracy.

The real scientific question is not simply whether a state is more quantum. It is whether the entire sensing protocol delivers a measurable advantage under a fair comparison.

The most promising proposals still face a demanding engineering process.

State preparation

Researchers must create the required quantum state with sufficient fidelity. A theoretically ideal state may be impossible to prepare exactly, so practical implementations rely on approximations.

Stability and coherence

The sensor must maintain its useful quantum properties long enough to collect the signal. A more sensitive state may also be more fragile.

Readout accuracy

The measurement must extract the information that matters. If the readout process introduces large errors, a state-preparation advantage can be lost.

Measurement overhead

Preparation, stabilization, feedback, and readout all take time and consume resources. The complete protocol must be evaluated, not just its most impressive operation.

Application-specific noise

A sensor designed to measure a controlled displacement in a laboratory may face different disturbances when used to measure a force, a field, or a signal that changes continuously.

Independent replication

Other researchers must be able to reproduce the results and verify that the claimed advantage holds under the specified conditions.

These are not minor details. They determine whether a quantum-sensing technique can move from a compelling paper to a reliable instrument.

Dark matter remains one of the major open questions in physics. Its gravitational effects are well supported by observations, but its underlying particle nature has not been established.

Some theoretical models predict that dark matter could interact weakly with ordinary matter or fields. Researchers have proposed precision-measurement experiments to search for signals associated with such interactions.

Improved quantum sensors could contribute to these efforts if the proposed interaction produces a signal that the instrument can detect.

However, none of the three studies compared here reports a dark-matter discovery.

The squeezed-cavity experiment demonstrates improved displacement measurement in a particular setup. The GKP and spin-dependent squeezing papers propose sensing strategies with potential applications in precision measurement.

To use such techniques in a dark-matter search, researchers would need to specify the predicted signal, design an appropriate detector, characterize backgrounds, and show that the instrument can distinguish a potential dark-matter signal from ordinary disturbances.

A better sensor can expand the range of hypotheses that scientists can test. It cannot, by itself, prove that a particular hypothesis is correct.

The three studies reveal a broader trend: quantum error correction, state engineering, and precision measurement are increasingly interconnected.

Squeezing offers one way to shape the uncertainty of a quantum state. GKP grid states provide a structured way to represent and detect displacements. Spin-dependent squeezing links oscillator motion to an auxiliary quantum degree of freedom.

Each approach attempts to engineer a state that makes a specific measurement more informative.

The strongest near-term evidence in this comparison comes from the superconducting-cavity experiment because the team generated the squeezed states and tested them in a displacement-sensing task.

The GKP and spin-dependent squeezing papers contribute different kinds of progress: mathematical analysis, protocol design, and numerical predictions that can guide future experiments.

Those theoretical contributions matter. They help researchers decide which states to prepare, which operations to implement, and which measurements could offer an advantage.

But the path from theory to technology requires testing the complete system under realistic conditions.

The next major advance may not be the sensor with the largest headline squeezing value. It may be the sensor that combines adequate sensitivity with reliable preparation, stable operation, low overhead, and reproducible results.

Quantum sensing is advancing through a combination of experimental demonstrations and theoretical innovation.

The superconducting-cavity experiment reported 14.6 dB of squeezing and demonstrated a displacement-sensing advantage over the standard quantum limit in its specific measurement task. The GKP-state study developed a theoretical strategy for continuous two-quadrature sensing, while the spin-dependent squeezing paper proposed an approach that combines oscillator motion and auxiliary spin states.

These studies do not prove that all quantum sensors can defeat noise, nor do they establish a universal route to perfect measurement.

They show something more precise and more useful: researchers can engineer quantum states and measurement protocols to improve sensitivity under carefully defined conditions.

The next step is to test the newer protocols in physical systems, measure their performance against fair baselines, and establish which advantages survive real-world imperfections.

That is how a quantum physics result becomes an engineering capability — and how precision measurement may eventually help scientists investigate forces, fields, and physical phenomena that remain difficult to observe.

Read more in my related articles, Quantum Superposition Explained, Quantum Entanglement Explained, Long-Lived Quantum States in Ytterbium Ions, and The Optical Magnus Effect.