Magnetometry for Agriculture and Animal Systems - From Classical Sensors to Quantum-Enabled Biosensing
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Biosensors |
| Published | 2026-06-01 |
| Authors | Zixuan Wang, Xi Zhang, Kexun Tang, Liming wu, Y N Huang |
| DOI | 10.3390/bios16060316 |
Room-temperature quantum magnetometers now resolve the picotesla-to-femtotesla signals that carry plant and animal physiology. The review concludes that noise suppression, not sensor sensitivity, sets the pace for field use. SERF optically pumped magnetometers reach 1 to 10 fT/√Hz, against 0.01 to 1 nT/√Hz for fluxgate sensors and 0.1 to 1 fT/√Hz for cryogenic SQUIDs. Plant action potentials of roughly 0.1 to 1 pT sit about nine orders of magnitude below the 50 µT geomagnetic background.
The authors assembled the comparison from sensor physics, running from fluxgate saturation to alkali-vapor spin dynamics, then from the ionic mechanisms of plant signaling to the electrophysiology of ruminant cardiology. Two summary tables set sensor sensitivity and bandwidth against the amplitudes of agriculturally relevant biomagnetic signals. A translational framework then maps a route from shielded medical laboratories to unshielded barns, greenhouses, and fields.
Why magnetic fields suit agricultural systems
Section titled “Why magnetic fields suit agricultural systems”Magnetic fields generated by ionic currents travel through soil, tissue, and structural materials with minimal attenuation or distortion. Optical sensing loses penetration depth in scattering and absorbing media, so soil opacity, canopy occlusion, tissue heterogeneity, and changing illumination compromise subsurface root work and deep-tissue physiology. Electrochemical sensors give chemical specificity but need direct contact with soil or fluids, which brings biofouling, electrode degradation, calibration drift, and sensitivity to temperature, pH, ionic strength, and moisture.
Magnetic readout is label-free and non-contact, and it reaches processes buried in opaque media. The review groups the biological sources into four classes: ionic currents, organ-level magnetism, soil magnetization, and magnetic nanoparticles. Membrane depolarization drives ion transport, and the resulting currents generate magnetic fields through the Biot-Savart law.
Sensor classes from fluxgate to NV center
Section titled “Sensor classes from fluxgate to NV center”Fluxgate magnetometers saturate a ferromagnetic core and read the resulting voltage harmonics. They give 10 to 100 pT/√Hz sensitivity from minus 55 °C to plus 175 °C at tens of milliwatts of power, but their footprint and power draw limit use in compact livestock tags. Hall-effect sensors work in the millitesla regime with bandwidth to 100 kHz and tolerate dust and vibration, so they suit machinery position and grain flow monitoring. Magnetoresistive sensors, in AMR, GMR, and TMR generations, reach about 1 nT sensitivity at microwatt power. One TMR array capped with 50 nm SiO2, 25 nm Si3N4, and 50 nm SiO2 stayed biocompatible with cultured neurons.
SQUIDs are the historical sensitivity benchmark at 0.1 to 1 fT/√Hz, but cryogenic operation ties them to laboratories and increases source distance. SERF optically pumped magnetometers reach fT-level sensitivity without cryogenics by suppressing spin-exchange relaxation in optically polarized alkali vapor near zero field. NV centers in diamond read the Zeeman shift of the NV spin through optically detected magnetic resonance, giving nanometer-to-micrometer resolution at room temperature. Weak-measurement-assisted optical readout has been proposed to lift SERF detection contrast, and inverse weak measurement was reported to improve robustness against laser power fluctuations and long-term drift.
The review’s sensor comparison sets the selection criteria side by side.
| Sensor | Typical sensitivity | Bandwidth | Ruggedness | Agricultural use |
|---|---|---|---|---|
| Fluxgate | 0.01 to 1 nT/√Hz | DC to 3 kHz | High | Soil mapping, drainage pipe detection |
| Hall effect | 10 to 10,000 nT/√Hz | DC to 100 kHz | High | Machinery position, grain flow monitoring |
| GMR/TMR | 0.1 to 10 nT/√Hz | DC to MHz | High | Bolus tracking, sensor arrays |
| LF NMR | about 1000 nT/√Hz | Hz to kHz | Moderate | Soil water, root imaging |
| NV center | 0.01 to 1 nT/√Hz | DC to GHz | High | Micro-sensing, tracers |
| SERF OPM | 1 to 10 fT/√Hz | DC to 100 Hz | Low, field and temperature sensitive | Livestock magnetocardiography, plant electrophysiology |
| SQUID | 0.1 to 1 fT/√Hz | DC to GHz | Poor, cryogenic | Laboratory reference, fetal magnetocardiography |
The amplitudes the sensors must resolve
Section titled “The amplitudes the sensors must resolve”The biological amplitude table sets the same comparison for signal sources rather than instruments.
| Biological source | Typical amplitude | Bandwidth | Note |
|---|---|---|---|
| Human brain, MEG | 50 to 500 fT | 0.1 to 100 Hz | Needs extreme sensitivity |
| Plant action potential | 0.5 to 1 pT | 0.1 to 10 Hz | Ion channels, very slow and ultra-weak |
| Fetal heart, cow | 1 to 10 pT | 1 to 50 Hz | Must be separated from the maternal signal |
| Adult heart, cow | 10 to 100 pT | 1 to 100 Hz | Overlaps environmental noise |
| Geomagnetic background | about 50,000,000 pT, or 50 µT | DC | Must be subtracted, six orders of magnitude larger |
The review’s Figure 2B places plant action potentials roughly nine orders of magnitude below that background. A SERF OPM gradiometer has already resolved heat-induced action potential transients in Venus flytrap with 0.5 to 1.0 pT peak-to-peak amplitudes. A plant action potential spans chloride efflux and calcium influx during depolarization, followed by potassium-mediated repolarization, so the magnetic signal mirrors the timing and shape of the electrical event.
Plant systems: fast signals, slow waves, and roots
Section titled “Plant systems: fast signals, slow waves, and roots”Variation potentials dominate most crops, including wheat, maize, and tomato, and they run from near DC to 10 Hz. They are hydraulic-electrical waves. A local perturbation changes xylem and phloem pressure, activates mechanosensitive ion channels, and drives coordinated ion fluxes across vascular membranes. Resolving them needs baseline stability against 1/f noise.
Belowground, nutrient uptake and proton extrusion across root membranes set up local current loops in the rhizosphere. Low-field MRI resolves root architecture in intact soil by exploiting spin-spin relaxation time differences between bulk soil water and root water. Applied fields also change plant behavior. Static fields raise reactive oxygen species while antioxidant enzymes counteract them, and the engineered MagLOV2 protein achieves a 10-fold amplification of magnetic field effects in fluorescence modulation, which the review presents as a quantum sensor for magnetic field detection in biological contexts.
Livestock, aquaculture, and group signals
Section titled “Livestock, aquaculture, and group signals”Cattle magnetocardiography is the strongest livestock application. Ventricular depolarization and repolarization in adult cattle create 10 to 100 pT peaks at the thoracic surface. Fetal magnetocardiography works where fetal electrocardiography fails, because vernix caseosa insulates the fetus and masks surface electrical signals while magnetic fields pass through. Fetal heart rate variability then serves as a marker for hypoxia, distress, and congenital defects.
Neural fields of 50 to 500 fT support objective welfare scoring. An OPM-MEG brain-computer interface driven by 60 Hz flicker reached 92.98% offline accuracy and an information transfer rate of 58.36 bit/min. In the rumen, ingestible magnetic boluses tracked by external fluxgate or MR arrays reveal a contraction cycle of about one event per minute, and hypomotility flags milk fever and displaced abomasum. Endogenous gastrointestinal slow waves from the interstitial cells of Cajal sit at 0.05 to 0.5 Hz and drop out during acidosis, which offers a steadier proxy for subacute ruminal acidosis than drifting bolus pH sensors. Cattle heart beats have been recorded in unshielded barns with OPM gradiometer arrays. In water, PIT tags and magnetic archival tags track individuals, and magnetoreception in salmonids and eels makes magnetic anomalies from cables and cages a measurable stressor.
AI denoising and digital twins
Section titled “AI denoising and digital twins”Target signals are buried in backgrounds orders of magnitude stronger. Convolutional networks and LSTMs trained on paired reference and signal-plus-noise channels produce synthetic gradiometers that the review reports as surpassing traditional adaptive filtering. AI-assisted denoising in OPM magnetoencephalography and in mobile systems improves detection in unshielded settings. For low-field MRI, AUTOMAP maps sensor-domain data straight to image space, reconstructing images from sparse, noisy k-space data.
Foundation models extend the same idea to multimodal data, and the review proposes a four-stage digital twin for dairy-cow ketosis. The input layer fuses bolus displacement, contraction frequency, and inter-contraction intervals with magnetocardiography waveforms and non-magnetic variables such as feed intake, milk yield, and barn temperature. The predictive layer combines recurrent networks with physiological state-space models to infer energy balance and metabolic stress. The output layer produces individualized ketosis-risk scores, validated against blood beta-hydroxybutyrate and milk ketone tests. A SERF OPM has quantified magnetic nanoparticles down to 0.1 mg/mL in 2 µL samples with an error below 11.8 ng.
Where deployment still fails
Section titled “Where deployment still fails”SERF operation requires the residual field to be near zero, and the sensors have limited dynamic range and strong sensitivity to field gradients, orientation, vibration, and platform motion. Active compensation coils extend operation beyond shielded rooms, but their effectiveness is greatest in stationary or quasi-stationary settings. The review therefore places OPMs in controlled barns, restrained-animal measurements, portable shielded stations, and magnetically quiet zones rather than on drones or free-ranging livestock.
System factors add up: laser sources, vapor-cell heating, field stabilization, multi-axis coils, calibration under temperature and humidity variation, dust and shock protection, and alignment and drift correction. Open problems include bandwidth-sensitivity trade-offs across biological timescales, sensor cost and scalability, and the lack of standardized protocols linking magnetic observables to actionable agricultural metrics. The review’s answer is hierarchical deployment, with classical arrays characterizing the environment and quantum sensors operating in the quiet zones inside it. For device selection, the amplitude table is the practical filter. A fluxgate or MR node fits rumen bolus tracking and soil mapping, while only quantum platforms reach fetal cardiac and plant action potential signals.
Materials and support
Section titled “Materials and support”The review’s diamond requirement is specific. NV centers in diamond give room-temperature magnetic sensing at nanometer-to-micrometer resolution for cellular studies of ion transport, tracer tracking, and nutrient uptake, and the paper cites a diamond quantum vector magnetometer demonstrated for deep-sea applications. That platform needs a diamond window with controlled thickness and a polish fine enough for optical readout, and 6CCVD’s diamond materials catalog lists MPCVD single-crystal and polycrystalline plates and wafers up to 125 mm, with thickness from 0.1 to 500 micrometers and polishing to Ra below 1 nm for single-crystal material. For the package and readout integration around the window, engineering support covers the next step.
View Original Abstract
Magnetic sensors offer a physically grounded and non-invasive approach to probing biological processes that remain inaccessible to optical, electrochemical, and radio-frequency techniques in complex agricultural environments. In recent years, advances in both classical and quantum magnetic sensors have enabled the detection of bioelectromagnetic signals across plants, soils, animals, and aquatic systems, spanning spatial scales from ionic currents to organ-level electrophysiology and population-level dynamics, positioning magnetometry as an emerging modality within the broader biosensor landscape. This review surveys the evolution of magnetic sensing technologies for agricultural and animal systems, from robust classical sensors used in navigation and soil mapping to quantum-enabled platforms, including Optically Pumped Magnetometers (OPMs) and Nitrogen-Vacancy (NV) centers, capable of resolving pT to fT biomagnetic signals. We synthesize the characteristic amplitudes, frequency ranges, and physiological origins of agriculturally relevant magnetic signals, and critically assess how techniques originally developed for medical magnetoencephalography, magnetocardiography, and low-field magnetic resonance imaging (LF-MRI) are being translated into field-deployable agricultural applications. Beyond sensing hardware, we highlight the essential role of artificial intelligence in extracting weak biological signals from dominant environmental noise, enabling synthetic gradiometry, low-field image reconstruction, and scalable interpretation in unshielded settings. Finally, we discuss how the integration of magnetic biosensing with digital twins supports predictive, multiscale monitoring of plant, animal, and ecosystem health. Together, these developments position magnetometry as an enabling technology for next-generation biosensors in precision and sustainable agriculture.
Free Standing Polycrystalline Cvd Diamond Membrane
View Specs →
High Concentration Quantum Grade Single Crystal Diamond
View Specs →Products
View Specs →References we cover
Section titled “References we cover”- Precision Magnetometers for Aerospace Applications - A Review
- Optical magnetic detection of single-neuron action potentials using quantum defects in diamond
Original source
Section titled “Original source”References
Section titled “References”- 2022 - Applications of IoT for optimized greenhouse environment and resources management [Crossref]
- 2021 - Internet of things for the future of smart agriculture: A comprehensive survey of emerging technologies [Crossref]
- 2018 - Hyperspectral sensors and imaging technologies in phytopathology: State of the art [Crossref]
- 2023 - Plant optics: Underlying mechanisms in remotely sensed signals for phenotyping applications [Crossref]
- 2022 - A critical review on applications of hyperspectral remote sensing in crop monitoring [Crossref]
- 2010 - Electrochemical biosensors [Crossref]
- 2025 - From point sensing to intelligent systems: A comprehensive review on advanced sensor technologies for soil health monitoring [Crossref]
- 2023 - A movable unshielded magnetocardiography system [Crossref]