This blog summarizes research trends based on published paper abstracts. Specific numbers or findings may contain inaccuracies. For scholarly rigor, always consult the original papers cited in each post.
In 2024, a team led by Valahu and colleagues at the University of Sydney did something that sounds impossible. They took a single trapped ytterbium-171 ion and used it to measure two physical quantities---displacement and momentum---simultaneously, with precision below the standard quantum limit for both. The uncertainty principle, the bedrock constraint of quantum mechanics, says you cannot know both position and momentum of a particle with arbitrary precision at the same time. Yet here was an experiment that beat the classical precision limit on both measurements at once.
This was not a violation of quantum mechanics. It was something more subtle and, for the future of precision measurement, more consequential: the first experimental demonstration that quantum resources can be distributed across multiple parameters to achieve simultaneous sub-classical sensitivity. The result opens the door to a new generation of quantum sensors that measure not just one thing very well, but many things at once---better than any classical sensor can.
The Problem: Why One Parameter Is Not Enough
Most quantum sensing experiments optimize for a single measurable quantity. An atomic clock measures frequency. A gravitational wave detector measures strain. A magnetometer measures field strength. The theory of single-parameter quantum estimation is mature: prepare a quantum state, let it interact with the quantity of interest, measure the state, and extract the parameter. The precision is bounded by the quantum Cramer-Rao bound, which depends on the quantum Fisher information (QFI) of the probe state.
But real-world sensing problems are rarely single-parameter. A magnetic field has three spatial components. A gravitational field has a gradient tensor with multiple independent elements. Imaging requires estimating phase across many spatial points. Navigating with inertial sensors demands simultaneous acceleration and rotation measurements. Traditionally, each parameter gets its own dedicated sensor, or a single sensor cycles through parameters sequentially---sacrificing bandwidth, integration time, or both.
The question that drives multiparameter quantum sensing is: can a single quantum system estimate multiple parameters simultaneously, and can quantum resources provide an advantage over classical strategies for all parameters at once?
Pezze and Smerzi (2025) provide the most comprehensive review to date of the theoretical foundations underlying multiparameter quantum estimation. Their framework centers on the quantum Fisher information matrix (QFIM), a generalization of the scalar QFI to the multiparameter case.
For a single parameter, the quantum Cramer-Rao bound states that the variance of any unbiased estimator is lower-bounded by the inverse of the QFI. For multiple parameters, the bound becomes a matrix inequality: the covariance matrix of the estimators is lower-bounded by the inverse of the QFIM. This sounds like a straightforward generalization, but it conceals a fundamental difficulty.
The Non-Commutativity Problem
When the generators of the parameters of interest do not commute---as is the case for position and momentum, or for different components of a magnetic field---the optimal measurements for individual parameters are incompatible. You cannot simultaneously perform the measurement that is optimal for parameter A and the measurement that is optimal for parameter B. The QFIM bound may not be simultaneously achievable for all parameters.
Pezze and Smerzi introduce a quantitative measure of this incompatibility: the parameter R, which quantifies the gap between the attainable precision (the Holevo bound, the tightest known matrix bound for quantum multiparameter estimation) and the QFIM bound. When R = 0, the parameters are compatible and the QFIM bound is achievable. When R > 0, there is a fundamental trade-off, and the Holevo bound describes the actual limit.
This distinction matters enormously in practice. For compatible parameters, multiparameter quantum sensing is "easy" in the sense that no precision is sacrificed by measuring all parameters simultaneously. For incompatible parameters, the challenge is to find states and measurements that minimize the trade-off---coming as close as possible to the QFIM bound for each parameter while respecting the Holevo constraint.
Distributed Quantum Sensing: Four Strategies
One of the most practically important results in the Pezze and Smerzi review concerns distributed quantum sensing (DQS): using entangled networks of sensors to estimate spatially distributed parameters. They classify four main strategies, ranging from simplest to most powerful:
Multiple Single-Parameter Sensors (MSPS): Each sensor estimates one parameter independently. No entanglement between sensors. This is the baseline classical approach.Simultaneous Multiparameter Estimation with Local States (SMELS): A single probe estimates multiple parameters, but uses separable (unentangled) states. Modest improvements over MSPS from optimized measurement design.Simultaneous Multiparameter Estimation with Entangled States (SMEES): Entanglement within each probe enables better simultaneous estimation. This is the strategy demonstrated by Valahu et al.Multiparameter Estimation with Probe Entanglement (MEPE): Entanglement across the sensor network, enabling correlations between spatially separated probes. This achieves an N-times-d improvement in total precision (N probes, d parameters) over the MSPS baseline---a genuine quantum advantage that scales with both the number of probes and the number of parameters.The progression from MSPS to MEPE represents a roadmap for practical quantum sensor networks. Current experiments are at the SMEES stage; MEPE remains largely theoretical but represents the ultimate scaling advantage.
Bayesian Adaptive Optimization
Classical estimation theory assumes many repeated measurements from which frequentist statistics extract parameters. Real quantum sensing experiments often operate in regimes where each measurement is expensive (in time, in the destruction of a fragile quantum state, or in the consumption of entangled resources). Pezze and Smerzi emphasize Bayesian adaptive protocols: after each measurement, update the posterior distribution over parameters, and choose the next measurement setting to maximize expected information gain. This approach naturally handles non-Gaussian posteriors, finite-sample effects, and the sequential optimization of measurement bases---all critical for multiparameter sensing where the optimal measurement depends on the (unknown) parameter values.
The Experiment: A Single Ion Measuring Two Things
Valahu et al. (2024) provide the experimental proof of concept. Their platform is a single Yb-171 ion confined in a Paul trap. The motional state of the ion in the trap is a quantum harmonic oscillator---a system whose position and momentum are canonically conjugate variables, related by the Heisenberg uncertainty relation.
Grid States and Number-Phase States
The key innovation is the preparation of non-Gaussian quantum states tailored for multiparameter sensing. Valahu et al. use two types of engineered states:
Grid states (also called GKP states, after Gottesman, Kitaev, and Preskill) are quantum states whose Wigner function---the quantum analog of a phase-space probability distribution---has a periodic grid structure. This periodicity allows simultaneous sensitivity to both position and momentum displacements, because a shift in either direction produces a detectable change in the state's measurement statistics.
Number-phase states are superpositions of Fock states (states with definite numbers of motional quanta) that provide simultaneous sensitivity to the oscillator's amplitude and phase. These are conceptually analogous to the squeezed states used in gravitational wave detectors, but generalized to provide sub-classical sensitivity in two quadratures simultaneously rather than one.
Results: Sub-Standard-Quantum-Limit on Both Parameters
The central result is that both grid states and number-phase states enable estimation of two non-commuting observables with precision below the standard quantum limit (SQL) for both parameters simultaneously. Specifically:
- Grid states achieved 5.1 dB below the SQL for displacement sensing and 3.1 dB below the SQL for momentum sensing, simultaneously.
- The force sensitivity reached 14.3 yN per root-Hz (yoctonewtons, 10^-24 newtons)---an extraordinary sensitivity achieved in a single trapped ion.
These numbers deserve context. The SQL is the best precision achievable using coherent states (the quantum states closest to classical behavior). Beating it by 5 dB means the measurement variance is about one-third of the classical limit. Doing this for one parameter is routine with squeezed states. Doing it for two non-commuting parameters simultaneously is the breakthrough.
What Makes This Possible
The experiment does not violate the uncertainty principle. The total "uncertainty area" in phase space is constrained by quantum mechanics. What the engineered states do is redistribute the uncertainty: instead of concentrating sensitivity in one direction (as squeezed states do), they create a pattern of sensitivity that provides information about both parameters. The periodic structure of grid states, for instance, means that a small displacement in any direction produces a measurable shift in the interference pattern, even though the total phase-space area is Heisenberg-limited.
The Textbook Foundation: Quantum Sensing Platforms and Protocols
Degen, Reinhard, and Cappellaro (2017) provide the canonical reference for the field of quantum sensing broadly. Their review in Reviews of Modern Physics covers 14 experimental platforms---from nitrogen-vacancy centers in diamond to superconducting circuits to atomic ensembles---and the measurement protocols common to all of them.
Core Protocols
Three measurement protocols underlie most quantum sensing experiments:
Ramsey interferometry: Prepare a superposition state, allow free evolution under the Hamiltonian of interest, and measure the accumulated phase. Sensitivity scales as 1/T, where T is the free evolution time. Limited by decoherence.
Rabi spectroscopy: Drive transitions between quantum states at a frequency near resonance with the parameter of interest. The lineshape of the resonance encodes the parameter value. Less sensitive than Ramsey but more robust to certain types of noise.
Multipulse (dynamical decoupling) sequences: Apply periodic pulses during the sensing interval to selectively filter the noise spectrum. This enables sensing of AC signals while rejecting DC noise, extending coherence times by orders of magnitude.
Entanglement and Quantum Error Correction
Degen et al. also review the role of entanglement in enhancing sensitivity beyond the SQL (the Heisenberg limit, scaling as 1/N rather than 1/sqrt(N) for N particles) and the emerging application of quantum error correction to protect quantum sensors from decoherence. These ideas are now being combined with multiparameter strategies: entangled sensor networks with error-corrected nodes represent the long-term vision for quantum-enhanced sensing infrastructure.
Connecting the Three Perspectives
These three papers form a coherent narrative. Degen et al. (2017) established the single-parameter foundations---platforms, protocols, and the SQL as the benchmark. Valahu et al. (2024) demonstrated that multiparameter sensing is experimentally feasible, achieving sub-SQL precision on two incompatible observables in a single quantum system. Pezze and Smerzi (2025) provide the theoretical framework that explains why this works, what the ultimate limits are, and how to scale it from a single ion to networks of entangled sensors.
The trajectory is clear. Single-parameter quantum sensing is mature and deployed (atomic clocks, magnetometers, gravitational wave detectors). Multiparameter quantum sensing is at the proof-of-concept stage, with the first experiments confirming the theoretical predictions. The next frontier is distributed multiparameter sensing with entangled networks---the MEPE regime---where the quantum advantage scales with both the number of sensors and the number of parameters.
Open Questions
Scalability of non-Gaussian state preparation: Grid states and number-phase states are difficult to prepare. Can they be generated reliably in larger systems---multiple ions, photonic networks, solid-state platforms?Decoherence in multiparameter protocols: The Valahu experiment operates in a well-isolated single-ion system. How does multiparameter advantage degrade in noisier, larger-scale systems? Can quantum error correction preserve it?Optimal measurement design: For a given set of incompatible parameters, what is the best measurement strategy? Bayesian adaptive methods are promising but computationally expensive. Practical heuristics are needed.Applications at scale: The most impactful applications---multi-axis inertial navigation, full magnetic field tensor mapping, multi-point gravitational field measurement---require sensor networks. The gap between SMEES (demonstrated) and MEPE (theoretical) remains wide.What To Watch
The Valahu et al. result is the starting gun for experimental multiparameter quantum sensing. Watch for extensions to more than two parameters, demonstrations in photonic and solid-state systems, and the first proof-of-principle distributed quantum sensor networks. The QFIM/Holevo framework from Pezze and Smerzi provides the theoretical scoreboard against which all future experiments will be judged.
For researchers outside quantum physics, the practical message is this: quantum sensors are moving beyond single-purpose instruments. The same theoretical and experimental advances that enable multiparameter sensing also enable quantum sensor networks for environmental monitoring, navigation, medical imaging, and fundamental physics. The uncertainty principle constrains what a single measurement can reveal, but it does not prevent quantum resources from being distributed across parameters in ways that classical sensors cannot match.
In 2024, a team led by Valahu and colleagues at the University of Sydney did something that sounds impossible. They took a single trapped ytterbium-171 ion and used it to measure two physical quantities---displacement and momentum---simultaneously, with precision below the standard quantum limit for both. The uncertainty principle, the bedrock constraint of quantum mechanics, says you cannot know both position and momentum of a particle with arbitrary precision at the same time. Yet here was an experiment that beat the classical precision limit on both measurements at once.
This was not a violation of quantum mechanics. It was something more subtle and, for the future of precision measurement, more consequential: the first experimental demonstration that quantum resources can be distributed across multiple parameters to achieve simultaneous sub-classical sensitivity. The result opens the door to a new generation of quantum sensors that measure not just one thing very well, but many things at once---better than any classical sensor can.
The Problem: Why One Parameter Is Not Enough
Most quantum sensing experiments optimize for a single measurable quantity. An atomic clock measures frequency. A gravitational wave detector measures strain. A magnetometer measures field strength. The theory of single-parameter quantum estimation is mature: prepare a quantum state, let it interact with the quantity of interest, measure the state, and extract the parameter. The precision is bounded by the quantum Cramer-Rao bound, which depends on the quantum Fisher information (QFI) of the probe state.
But real-world sensing problems are rarely single-parameter. A magnetic field has three spatial components. A gravitational field has a gradient tensor with multiple independent elements. Imaging requires estimating phase across many spatial points. Navigating with inertial sensors demands simultaneous acceleration and rotation measurements. Traditionally, each parameter gets its own dedicated sensor, or a single sensor cycles through parameters sequentially---sacrificing bandwidth, integration time, or both.
The question that drives multiparameter quantum sensing is: can a single quantum system estimate multiple parameters simultaneously, and can quantum resources provide an advantage over classical strategies for all parameters at once?
The Theoretical Framework: Quantum Fisher Information Matrix
Pezze and Smerzi (2025) provide the most comprehensive review to date of the theoretical foundations underlying multiparameter quantum estimation. Their framework centers on the quantum Fisher information matrix (QFIM), a generalization of the scalar QFI to the multiparameter case.
For a single parameter, the quantum Cramer-Rao bound states that the variance of any unbiased estimator is lower-bounded by the inverse of the QFI. For multiple parameters, the bound becomes a matrix inequality: the covariance matrix of the estimators is lower-bounded by the inverse of the QFIM. This sounds like a straightforward generalization, but it conceals a fundamental difficulty.
The Non-Commutativity Problem
When the generators of the parameters of interest do not commute---as is the case for position and momentum, or for different components of a magnetic field---the optimal measurements for individual parameters are incompatible. You cannot simultaneously perform the measurement that is optimal for parameter A and the measurement that is optimal for parameter B. The QFIM bound may not be simultaneously achievable for all parameters.
Pezze and Smerzi introduce a quantitative measure of this incompatibility: the parameter R, which quantifies the gap between the attainable precision (the Holevo bound, the tightest known matrix bound for quantum multiparameter estimation) and the QFIM bound. When R = 0, the parameters are compatible and the QFIM bound is achievable. When R > 0, there is a fundamental trade-off, and the Holevo bound describes the actual limit.
This distinction matters enormously in practice. For compatible parameters, multiparameter quantum sensing is "easy" in the sense that no precision is sacrificed by measuring all parameters simultaneously. For incompatible parameters, the challenge is to find states and measurements that minimize the trade-off---coming as close as possible to the QFIM bound for each parameter while respecting the Holevo constraint.
Distributed Quantum Sensing: Four Strategies
One of the most practically important results in the Pezze and Smerzi review concerns distributed quantum sensing (DQS): using entangled networks of sensors to estimate spatially distributed parameters. They classify four main strategies, ranging from simplest to most powerful:
Multiple Single-Parameter Sensors (MSPS): Each sensor estimates one parameter independently. No entanglement between sensors. This is the baseline classical approach.Simultaneous Multiparameter Estimation with Local States (SMELS): A single probe estimates multiple parameters, but uses separable (unentangled) states. Modest improvements over MSPS from optimized measurement design.Simultaneous Multiparameter Estimation with Entangled States (SMEES): Entanglement within each probe enables better simultaneous estimation. This is the strategy demonstrated by Valahu et al.Multiparameter Estimation with Probe Entanglement (MEPE): Entanglement across the sensor network, enabling correlations between spatially separated probes. This achieves an N-times-d improvement in total precision (N probes, d parameters) over the MSPS baseline---a genuine quantum advantage that scales with both the number of probes and the number of parameters.The progression from MSPS to MEPE represents a roadmap for practical quantum sensor networks. Current experiments are at the SMEES stage; MEPE remains largely theoretical but represents the ultimate scaling advantage.
Bayesian Adaptive Optimization
Classical estimation theory assumes many repeated measurements from which frequentist statistics extract parameters. Real quantum sensing experiments often operate in regimes where each measurement is expensive (in time, in the destruction of a fragile quantum state, or in the consumption of entangled resources). Pezze and Smerzi emphasize Bayesian adaptive protocols: after each measurement, update the posterior distribution over parameters, and choose the next measurement setting to maximize expected information gain. This approach naturally handles non-Gaussian posteriors, finite-sample effects, and the sequential optimization of measurement bases---all critical for multiparameter sensing where the optimal measurement depends on the (unknown) parameter values.
The Experiment: A Single Ion Measuring Two Things
Valahu et al. (2024) provide the experimental proof of concept. Their platform is a single Yb-171 ion confined in a Paul trap. The motional state of the ion in the trap is a quantum harmonic oscillator---a system whose position and momentum are canonically conjugate variables, related by the Heisenberg uncertainty relation.
Grid States and Number-Phase States
The key innovation is the preparation of non-Gaussian quantum states tailored for multiparameter sensing. Valahu et al. use two types of engineered states:
Grid states (also called GKP states, after Gottesman, Kitaev, and Preskill) are quantum states whose Wigner function---the quantum analog of a phase-space probability distribution---has a periodic grid structure. This periodicity allows simultaneous sensitivity to both position and momentum displacements, because a shift in either direction produces a detectable change in the state's measurement statistics.
Number-phase states are superpositions of Fock states (states with definite numbers of motional quanta) that provide simultaneous sensitivity to the oscillator's amplitude and phase. These are conceptually analogous to the squeezed states used in gravitational wave detectors, but generalized to provide sub-classical sensitivity in two quadratures simultaneously rather than one.
Results: Sub-Standard-Quantum-Limit on Both Parameters
The central result is that both grid states and number-phase states enable estimation of two non-commuting observables with precision below the standard quantum limit (SQL) for both parameters simultaneously. Specifically:
- Grid states achieved 5.1 dB below the SQL for displacement sensing and 3.1 dB below the SQL for momentum sensing, simultaneously.
- The force sensitivity reached 14.3 yN per root-Hz (yoctonewtons, 10^-24 newtons)---an extraordinary sensitivity achieved in a single trapped ion.
These numbers deserve context. The SQL is the best precision achievable using coherent states (the quantum states closest to classical behavior). Beating it by 5 dB means the measurement variance is about one-third of the classical limit. Doing this for one parameter is routine with squeezed states. Doing it for two non-commuting parameters simultaneously is the breakthrough.
What Makes This Possible
The experiment does not violate the uncertainty principle. The total "uncertainty area" in phase space is constrained by quantum mechanics. What the engineered states do is redistribute the uncertainty: instead of concentrating sensitivity in one direction (as squeezed states do), they create a pattern of sensitivity that provides information about both parameters. The periodic structure of grid states, for instance, means that a small displacement in any direction produces a measurable shift in the interference pattern, even though the total phase-space area is Heisenberg-limited.
The Textbook Foundation: Quantum Sensing Platforms and Protocols
Degen, Reinhard, and Cappellaro (2017) provide the canonical reference for the field of quantum sensing broadly. Their review in Reviews of Modern Physics covers 14 experimental platforms---from nitrogen-vacancy centers in diamond to superconducting circuits to atomic ensembles---and the measurement protocols common to all of them.
Core Protocols
Three measurement protocols underlie most quantum sensing experiments:
Ramsey interferometry: Prepare a superposition state, allow free evolution under the Hamiltonian of interest, and measure the accumulated phase. Sensitivity scales as 1/T, where T is the free evolution time. Limited by decoherence.
Rabi spectroscopy: Drive transitions between quantum states at a frequency near resonance with the parameter of interest. The lineshape of the resonance encodes the parameter value. Less sensitive than Ramsey but more robust to certain types of noise.
Multipulse (dynamical decoupling) sequences: Apply periodic pulses during the sensing interval to selectively filter the noise spectrum. This enables sensing of AC signals while rejecting DC noise, extending coherence times by orders of magnitude.
Entanglement and Quantum Error Correction
Degen et al. also review the role of entanglement in enhancing sensitivity beyond the SQL (the Heisenberg limit, scaling as 1/N rather than 1/sqrt(N) for N particles) and the emerging application of quantum error correction to protect quantum sensors from decoherence. These ideas are now being combined with multiparameter strategies: entangled sensor networks with error-corrected nodes represent the long-term vision for quantum-enhanced sensing infrastructure.
Connecting the Three Perspectives
These three papers form a coherent narrative. Degen et al. (2017) established the single-parameter foundations---platforms, protocols, and the SQL as the benchmark. Valahu et al. (2024) demonstrated that multiparameter sensing is experimentally feasible, achieving sub-SQL precision on two incompatible observables in a single quantum system. Pezze and Smerzi (2025) provide the theoretical framework that explains why this works, what the ultimate limits are, and how to scale it from a single ion to networks of entangled sensors.
The trajectory is clear. Single-parameter quantum sensing is mature and deployed (atomic clocks, magnetometers, gravitational wave detectors). Multiparameter quantum sensing is at the proof-of-concept stage, with the first experiments confirming the theoretical predictions. The next frontier is distributed multiparameter sensing with entangled networks---the MEPE regime---where the quantum advantage scales with both the number of sensors and the number of parameters.
Open Questions
Scalability of non-Gaussian state preparation: Grid states and number-phase states are difficult to prepare. Can they be generated reliably in larger systems---multiple ions, photonic networks, solid-state platforms?Decoherence in multiparameter protocols: The Valahu experiment operates in a well-isolated single-ion system. How does multiparameter advantage degrade in noisier, larger-scale systems? Can quantum error correction preserve it?Optimal measurement design: For a given set of incompatible parameters, what is the best measurement strategy? Bayesian adaptive methods are promising but computationally expensive. Practical heuristics are needed.Applications at scale: The most impactful applications---multi-axis inertial navigation, full magnetic field tensor mapping, multi-point gravitational field measurement---require sensor networks. The gap between SMEES (demonstrated) and MEPE (theoretical) remains wide.What To Watch
The Valahu et al. result is the starting gun for experimental multiparameter quantum sensing. Watch for extensions to more than two parameters, demonstrations in photonic and solid-state systems, and the first proof-of-principle distributed quantum sensor networks. The QFIM/Holevo framework from Pezze and Smerzi provides the theoretical scoreboard against which all future experiments will be judged.
For researchers outside quantum physics, the practical message is this: quantum sensors are moving beyond single-purpose instruments. The same theoretical and experimental advances that enable multiparameter sensing also enable quantum sensor networks for environmental monitoring, navigation, medical imaging, and fundamental physics. The uncertainty principle constrains what a single measurement can reveal, but it does not prevent quantum resources from being distributed across parameters in ways that classical sensors cannot match.