Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Optics Express

Neural network-aided optimisation of a radio-frequency atomic magnetometer

Han Yao, Benjamin Maddox, Ferruccio Renzoni

Published Jul 19, 2023• 6 citations
Publisher

Abstract

Efficient unsupervised optimisation of atomic magnetometers is a requirement in many applications, where direct intervention of an operator is not feasible. The efficient extraction of the optimal operating conditions from a small sample of experimental data requires a robust automated regression of the available data. Here we address this issue and propose the use of general regression neural networks as a tool for the optimisation of atomic magnetometers which does not require human supervision and is efficient, as it is ideally suited to operating with a small sample of data as input. As a case study, we specifically demonstrate the optimisation of an unshielded radio-frequency atomic magnetometer by using a general regression neural network which establishes a mapping between three input variables, the cell temperature, the pump beam power and the probe beam power, and one output variable, the AC sensitivity. The optimisation results into an AC sensitivity of 44 fT/Hz at 26 kHz.

DOI
10.1364/oe.498163

Related papers

Optimisation of a radio-frequency atomic magnetometer: a Uniform Design approach

Han Yao, Benjamin Maddox, Yuval Cohen +1 more

High-sensitivity operation of a radio-frequency atomic magnetometer (RF-AM) requires careful setting of the system parameters, including the lasers intensity and detuning, and the vapour cell temperature. The identification of the optimal operating parameters, which ensures high sensitivity, is typically performed empirically and is often a lengthy process, which is especially labour intensive if frequent retuning of the magnetometer is required to perform different tasks. This paper demonstrates an efficient approach to RF-AM performance optimisation which relies on an open-loop optimisation technique based on Uniform Design (UD). This paper specifically describes the optimisation of an unshielded RF-AM based on a 4-factor-12-level UD of the experimental parameters space. The proposed procedure is shown to lead to the efficient optimisation of the atomic magnetometer at different frequencies, and is applicable to both AC and DC sensitivity optimisation. The procedure does not require any detailed knowledge of the model underlying the operation of the RF-AM and is effective in reducing the number of experimental runs required for the optimisation. It is ideally suited to self-calibration of devices without human supervision.

Optics Express•Dec 27, 2021•9 citations

Radio-frequency atomic magnetometers: an analysis of interrogation regimes

Jenelle Rajroop

An atomic magnetometer is a sensor which is used to measure a magnetic field through its interaction with the atomic sample. Significant research into atomic magnetometry has led to the development of very sensitive atomic sensors capable of matching the sensitivity of the most sensitive magnetometers, superconducting quantum interference devices (SQUIDs). Because SQUIDS require cryogenics to operate, atomic magnetometers provide a sensitive, yet low-cost alternative. They have found use in many areas such as medicine, security, explosives detection and fundamental physics research. One of the primary factors influencing sensitivity is the detuning of the probe beam from the resonant transitions of the atomic ground state. A caesium room temperature radio-frequency (rf) magnetometer is constructed and used to investigate the influence of the probe beam detuning on the magnetometer signal of the F = 3 and F = 4 ground states. The results of probing near and far from resonance revealed an off-resonant regime and two absorptive regimes. In the off-resonant regime, the atomic spins are unperturbed by the probe beam; it is a quantum non-demolition (QND) interaction. The two absorptive regimes, found when the probe beam is in the vicinity of either the 62S1 2 F = 3 → 6 2P3 2 F 0 = 2,3,4 or the 62S1 2 F = 4 → 6 2P3 2 F 0 = 3,4,5 transitions, is characterised as a non-QND interaction in which the probe beam influences the measurement. The sensitivity of the rf magnetometer is determined to be ≈ 1.98 fT/ √ Hz. In addition, the exploration of the relationship between the signal to noise ratio (SNR) and probe beam detuning revealed that the SNR is constant with detuning but the larger the detuning, the higher the probe beam power needs to be to reach the optimum SNR.

•Jun 28, 2018•0 citations

Room temperature femtotesla radio-frequency atomic magnetometer

Witold Chałupczak, R. M. Godun, Szymon Pustelny +1 more

A radio-frequency tunable atomic magnetometer with a sensitivity of about 1 fT/Hz1/2 in a range of 10–500 kHz is demonstrated. The magnetometer is operated in the orientation configuration in which atoms are pumped to the stretched atomic state by a scheme based on indirect optical pumping using only one unmodulated, low-power laser. The magnetometer operates with cesium atoms, which have sufficient vapor pressure near room temperature to enable high magnetometric sensitivities. The technique enables a compact and robust module to be constructed that could become an in-the-field device.

Applied Physics Letters•Jun 11, 2012•87 citations

High-sensitivity operation of an unshielded single cell radio-frequency atomic magnetometer

Han Yao, Benjamin Maddox, Ferruccio Renzoni

Real-world applications of atomic magnetometers require the ability to operate them with high-sensitivity in the presence of magnetic noise. In the present work, high-sensitivity operation of unshielded atomic magnetometers in a magnetically noisy environment is demonstrated. The distinguishing feature of the demonstrated approach is the implementation of active in-situ bias field stabilization using multiple fluxgate magnetometers. This is combined with the use of a counter-propagating pump and triple-pass probe configuration, to maximize the atomic polarization and the probe rotation respectively, so to reach high-sensitivity. The improvement in sensitivity of the unshielded system with respect to previous realizations is fully characterized, with the contributions of the different modifications of the apparatus individually quantified. The presented set-up is suitable for the detection of long-range magnetic fields, where shielding or differential measurements using multi-sensor set-ups do not constitute viable options.

Optics Express•Oct 21, 2022•29 citations

Machine Learning Assisted Vector Atomic Magnetometry

Xin Meng, Youwei Zhang, Xichang Zhang +6 more

Multiparameter sensing such as vector magnetometry often involves complex setups due to various external fields needed in explicitly connecting one measured signal to one parameter. Here, we propose a paradigm of indirect encoding for vector atomic magnetometry based on machine learning. We encode the three-dimensional magnetic-field information in the set of four simultaneously acquired signals associated with the optical rotation of a laser beam traversing the atomic sample. The map between the recorded signals and the vectorial field information is established through a pre-trained deep neural network. We demonstrate experimentally a single-shot all optical vector atomic magnetometer, with a simple scalar-magnetometer design employing only one elliptically-polarized laser beam and no additional coils. Magnetic field amplitude sensitivities of about 100 fT/√Hz and angular sensitivities of about 100∼200μrad/√Hz (for a magnetic field of around 140 nT) are derived from the neural network. Our approach can reduce the complexity of the architecture of vector magnetometers, and may shed light on the general design of multiparameter sensing.

arXiv:2301.05707•Dec 31, 2022•0 citations

A radio-frequency Bose–Einstein condensate magnetometer

Y. Cohen, Benjamin Maddox, Cameron Deans +2 more

We report on a radio frequency magnetometer employing a Bose–Einstein condensate of 87Rb atoms held in a dipole trap. An AC sensitivity of 296 pT/Hz is achieved at a probing volume of 3.2×10−8 cm3, leading to a volume-normalized sensitivity of 53 fT/Hz cm−3. At larger probing volumes with the atoms released from the magnetic trap used in the initial phase of the evaporation sequence, the AC sensitivity is improved to 1.4 pT/Hz, allowing a two-mode approach for applications requiring improved sensitivity. Immediate application in high-resolution electromagnetic induction imaging is expected when compared to performance of other induction-based imaging platforms.

Applied Physics Letters•Apr 18, 2022•10 citations
Efficient unsupervised optimisation of atomic magnetometers is a requirement in many applications, where direct intervention of an operator is not feasible. The efficient extraction of the optimal operating conditions from a small sample of experimental data requires a robust automated regression of the available data. Here we address this issue and propose the use of general regression neural networks as a tool for the optimisation of atomic magnetometers which does not require human supervision and is efficient, as it is ideally suited to operating with a small sample of data as input. As a case study, we specifically demonstrate the optimisation of an unshielded radio-frequency atomic magnetometer by using a general regression neural network which establishes a mapping between three input variables, the cell temperature, the pump beam power and the probe beam power, and one output variable, the AC sensitivity. The optimisation results into an AC sensitivity of 44 fT/Hz at 26 kHz.
High-sensitivity operation of a radio-frequency atomic magnetometer (RF-AM) requires careful setting of the system parameters, including the lasers intensity and detuning, and the vapour cell temperature. The identification of the optimal operating parameters, which ensures high sensitivity, is typically performed empirically and is often a lengthy process, which is especially labour intensive if frequent retuning of the magnetometer is required to perform different tasks. This paper demonstrates an efficient approach to RF-AM performance optimisation which relies on an open-loop optimisation technique based on Uniform Design (UD). This paper specifically describes the optimisation of an unshielded RF-AM based on a 4-factor-12-level UD of the experimental parameters space. The proposed procedure is shown to lead to the efficient optimisation of the atomic magnetometer at different frequencies, and is applicable to both AC and DC sensitivity optimisation. The procedure does not require any detailed knowledge of the model underlying the operation of the RF-AM and is effective in reducing the number of experimental runs required for the optimisation. It is ideally suited to self-calibration of devices without human supervision.
An atomic magnetometer is a sensor which is used to measure a magnetic field through its interaction with the atomic sample. Significant research into atomic magnetometry has led to the development of very sensitive atomic sensors capable of matching the sensitivity of the most sensitive magnetometers, superconducting quantum interference devices (SQUIDs). Because SQUIDS require cryogenics to operate, atomic magnetometers provide a sensitive, yet low-cost alternative. They have found use in many areas such as medicine, security, explosives detection and fundamental physics research. One of the primary factors influencing sensitivity is the detuning of the probe beam from the resonant transitions of the atomic ground state. A caesium room temperature radio-frequency (rf) magnetometer is constructed and used to investigate the influence of the probe beam detuning on the magnetometer signal of the F = 3 and F = 4 ground states. The results of probing near and far from resonance revealed an off-resonant regime and two absorptive regimes. In the off-resonant regime, the atomic spins are unperturbed by the probe beam; it is a quantum non-demolition (QND) interaction. The two absorptive regimes, found when the probe beam is in the vicinity of either the 62S1 2 F = 3 → 6 2P3 2 F 0 = 2,3,4 or the 62S1 2 F = 4 → 6 2P3 2 F 0 = 3,4,5 transitions, is characterised as a non-QND interaction in which the probe beam influences the measurement. The sensitivity of the rf magnetometer is determined to be ≈ 1.98 fT/ √ Hz. In addition, the exploration of the relationship between the signal to noise ratio (SNR) and probe beam detuning revealed that the SNR is constant with detuning but the larger the detuning, the higher the probe beam power needs to be to reach the optimum SNR.
A radio-frequency tunable atomic magnetometer with a sensitivity of about 1 fT/Hz1/2 in a range of 10–500 kHz is demonstrated. The magnetometer is operated in the orientation configuration in which atoms are pumped to the stretched atomic state by a scheme based on indirect optical pumping using only one unmodulated, low-power laser. The magnetometer operates with cesium atoms, which have sufficient vapor pressure near room temperature to enable high magnetometric sensitivities. The technique enables a compact and robust module to be constructed that could become an in-the-field device.
Real-world applications of atomic magnetometers require the ability to operate them with high-sensitivity in the presence of magnetic noise. In the present work, high-sensitivity operation of unshielded atomic magnetometers in a magnetically noisy environment is demonstrated. The distinguishing feature of the demonstrated approach is the implementation of active in-situ bias field stabilization using multiple fluxgate magnetometers. This is combined with the use of a counter-propagating pump and triple-pass probe configuration, to maximize the atomic polarization and the probe rotation respectively, so to reach high-sensitivity. The improvement in sensitivity of the unshielded system with respect to previous realizations is fully characterized, with the contributions of the different modifications of the apparatus individually quantified. The presented set-up is suitable for the detection of long-range magnetic fields, where shielding or differential measurements using multi-sensor set-ups do not constitute viable options.
Multiparameter sensing such as vector magnetometry often involves complex setups due to various external fields needed in explicitly connecting one measured signal to one parameter. Here, we propose a paradigm of indirect encoding for vector atomic magnetometry based on machine learning. We encode the three-dimensional magnetic-field information in the set of four simultaneously acquired signals associated with the optical rotation of a laser beam traversing the atomic sample. The map between the recorded signals and the vectorial field information is established through a pre-trained deep neural network. We demonstrate experimentally a single-shot all optical vector atomic magnetometer, with a simple scalar-magnetometer design employing only one elliptically-polarized laser beam and no additional coils. Magnetic field amplitude sensitivities of about 100 fT/√Hz and angular sensitivities of about 100∼200μrad/√Hz (for a magnetic field of around 140 nT) are derived from the neural network. Our approach can reduce the complexity of the architecture of vector magnetometers, and may shed light on the general design of multiparameter sensing.
We report on a radio frequency magnetometer employing a Bose–Einstein condensate of 87Rb atoms held in a dipole trap. An AC sensitivity of 296 pT/Hz is achieved at a probing volume of 3.2×10−8 cm3, leading to a volume-normalized sensitivity of 53 fT/Hz cm−3. At larger probing volumes with the atoms released from the magnetic trap used in the initial phase of the evaporation sequence, the AC sensitivity is improved to 1.4 pT/Hz, allowing a two-mode approach for applications requiring improved sensitivity. Immediate application in high-resolution electromagnetic induction imaging is expected when compared to performance of other induction-based imaging platforms.