The new research, published in Frontiers in the Internet of Things, argues that the solution lies in densely deployed networks of low-cost, IoT-enabled particulate matter sensors. Traditional instruments such as beta attenuation monitors and tapered element oscillating microbalances are highly accurate but bulky and expensive, which makes dense deployment impossible. In Hyderabad, a metropolitan region of more than 6.7 million people spread over 650 square kilometers, the Central Pollution Control Board and the Telangana State Pollution Control Board operate only twelve centralized air quality monitoring stations. Pollutant levels can vary drastically even between smaller blocks within a city, so a dozen stations simply cannot resolve that spatial variability.

To bridge this gap, the team developed an end-to-end low-cost IoT system and deployed 49 devices across approximately 4 square kilometers of the Gachibowli region of Hyderabad. Forty-three devices were built in-house, while six commercial Airveda monitors were added for comparison. The in-house device is built around the TTGO T-Call ESP32 microcontroller, which reads a Nova SDS011 laser light-scattering particulate sensor and an SHT21 temperature and humidity sensor every 30 seconds. The SHT21 is included because light-scattering sensors become unreliable at high temperature and relative humidity levels, allowing unreliable PM values to be filtered out. Data packets of 24 bytes are offloaded to the ThingSpeak cloud platform using the secure MQTTS protocol over Wi-Fi or a 2G cellular network, with an embedded SIM customized for IoT applications to prevent theft-related misuse and avoid carrier lock-in.

Robustness against data loss was a central design goal. Each device is powered by an AC-to-DC adapter with a 1,000 mAh lithium polymer battery that switches on automatically when AC power fails, ensuring uninterrupted operation. If network connectivity drops, sensed data is cached in a portion of the microcontroller’s 540 KB RAM, up to a maximum of 20,000 data points, and transmitted in bulk once the connection is restored. If a prolonged outage fills the cache, stored points are cleared in first-in-first-out order to make room for new data. The complete device, enclosed in a dust-proof and water-resistant IP65 box of 125 by 125 by 125 millimeters, costs about 7,000 Indian rupees, roughly 84 US dollars, a fraction of the price of commercial monitors that retail for up to 64,900 rupees in the Indian market.

Calibration was treated as rigorously as hardware design. Each low-cost sensor was co-located with a precise Aeroqual S500 reference instrument in a ventilated room for a week, generating roughly 20,160 data points per sensor. Before calibration, all sensors followed the reference in trend but differed by unique offsets in absolute value, so simple linear regression was used to learn individual correction parameters for every device. The team chose linear regression because it compensates for offsets well while preserving the trend in the data, and it achieved strong agreement with the reference instrument, with R-squared values of 0.99 for PM2.5 and 0.87 for PM10. Because sensor behavior changes with the seasons, separate calibration functions were computed at the onset of each season.

Data preprocessing removed unreliable readings before analysis. Values recorded at relative humidity above 80 percent, where PM10 readings are unreliable, and raw values outside the sensor range of 0 to 999 micrograms per cubic meter were discarded, accounting for about 0.5 percent of the dataset. Outliers were then identified with the interquartile range method, flagging a further 1.4 percent, and the remaining gaps of about 1.9 percent were filled with simple linear interpolation. Field deployment was completed in July 2021, and usable data were collected over seven months spanning all seasons, from August 2021 to June 2022, aggregating 20.7 million data points. The 4-square-kilometer study area was divided into roughly 42 boxes of 400 by 400 square meters, with an attempt to place one device in each box, covering urban, semi-urban, and green regions as well as busy traffic junctions and roadside poles.

The seven-month dataset revealed striking seasonal and spatial patterns. Mean and variance of PM10 were highest in winter, when temperature inversion traps cold air and particulates near the ground, and lowest during the monsoon, when frequent rainfall settles the particles. Among the 49 devices, the highest mean PM10 values appeared at a traffic junction, a green region undergoing construction activity, and a roadside location with sluggish traffic flow, while the lowest values occurred in a quiet residential area and roadside sites with free-flowing traffic. Inverse distance weighting was used to interpolate PM10 concentrations at unmonitored locations, producing maps at 11 a.m., 2 p.m., and 9 p.m. for each season. Concentrations peaked during morning and evening office and peak traffic hours and dipped in the early afternoon as traffic thinned.

The network’s density proved its worth during the Diwali festival. On the evening of 4 November 2021, widespread firecracker bursting produced a clear peak in PM10 across all devices, with one commercial monitor rising from 40 to 307 micrograms per cubic meter before and after the celebrations. Concentrations fell sharply within a few hours, indicating a temporary, activity-driven spike. When the team simulated sparse deployments of 12 and 4 devices, the results were sobering: interpolation with 12 devices could identify the event but missed local hotspots, while 4 devices missed the event entirely and could be misleading. The root mean square error rose from 32.47 with 12 devices to 59.29 with 4 devices, demonstrating that sparse networks risk erroneous interpretations of actual pollution levels and distribution.

Correlation analysis offered a concrete guideline for future deployments. Using Kendall’s tau on hourly averaged PM10 samples, the team fitted a two-term exponential model showing that the correlation coefficient changes rapidly between devices less than 350 meters apart, after which the decline becomes gradual. This suggests that particulate matter monitors should be deployed at most 350 meters apart to capture spatial variability accurately in a densely packed urban landscape like Gachibowli, where office spaces, highways, residential complexes, and parks coexist within a small area. The researchers also built a responsive web-based dashboard, named AirIoT, built on the Django framework with PostgreSQL and real-time updates every 10 minutes, allowing users to view device locations on an interactive map, examine 48-hour trends, and access data through OAuth-authenticated APIs.

The findings carry direct implications for city planning authorities, who can use the network and its data to guide urban planning, infrastructure development, and targeted pollution mitigation. The team suggests that future work could explore adaptive, non-parametric sensing rates that change dynamically based on recent historical data to make collection more robust and energy efficient, as well as advanced interpolation techniques such as Kriging. For now, the study stands as a compelling demonstration that calibrated low-cost sensors, deployed densely enough, can reveal the street-level texture of air pollution that a dozen centralized stations never could.

Subject of Research: Dense deployment of low-cost IoT particulate matter monitoring network in Hyderabad, India

Article Title: Development of end-to-end low-cost IoT system for densely deployed PM monitoring network: an Indian case study

Article References: Parmar, A., Sara, S., Dwivedi, A. K., Reddy, C. R., Patwardhan, I., Bijjam, S. D., Chaudhari, S., Rajan, K. S., & Vemuri, K. (2024). Development of end-to-end low-cost IoT system for densely deployed PM monitoring network: an Indian case study. Frontiers in the Internet of Things, 3, Article 1332322. https://doi.org/10.3389/friot.2024.1332322

Image Credits: AI Generated

DOI: 10.3389/friot.2024.1332322

Keywords: air pollution, particulate matter, IoT, low-cost sensors, Hyderabad, PM2.5, PM10, sensor calibration, dense deployment, Diwali, spatial interpolation, web dashboard

The core difficulty is that temperature attacks every link in a measurement chain. Many commercial instruments are simply not rated for the temperatures envisioned in advanced molten salt systems. Elevated heat degrades piezoelectric elements and their bonding layers, progressively demagnetizes the permanent magnets at the heart of electromagnetic meters, and changes fluid properties that calibrations depend on. Thermal gradients perturb ultrasonic propagation and conductivity-dependent electromagnetic signals, while thermal expansion quietly alters calibrated geometries. Add corrosion, wetting changes, fouling, and interfacial degradation, and the picture darkens further. In reactors, radiation adds a qualification requirement whose significance depends on the sensing element, its packaging, and its proximity to the core. Long service intervals and restricted access mean that in-situ recalibration, the routine fix for drifting instruments elsewhere in industry, is often simply not available.

The choice of sensing principle is dictated by the fluid itself, and here the review draws a sharp line between two worlds. Liquid metals conduct electricity with conductivities near one million siemens per meter, enabling strong electromagnetic coupling and both contact and contactless sensor configurations. Molten salts, by contrast, are several orders of magnitude less conductive. Experimental studies show that signal levels can become insufficient for conductivity-dependent electromagnetic approaches in salts, a limitation that applies most directly to inductive and Lorentz-force techniques rather than to all electromagnetic flowmeters. Molten salts also impose punishing material compatibility requirements on any sensor interface, and limited optical access constrains another family of techniques. In sodium, the physics is generous; in fluoride and chloride salts, it is grudging.

Electromagnetic flowmeters illustrate both the promise and the pitfalls. In sodium systems, permanent-magnet flowmeters provide circuit-level measurements, while immersed eddy-current flowmeters, with an excitation coil and two receiving coils, deliver local velocity readings in constrained locations, with electronics fast enough for millisecond-class loss-of-flow monitoring. These devices have no moving parts, respond rapidly, and work with opaque fluids. But temperature effects differ sharply by principle. Conventional permanent-magnet flowmeters depend primarily on velocity, magnetic field, and geometry rather than directly on conductivity, whereas inductive and Lorentz-force configurations are sensitive to conductivity changes that alter electromagnetic coupling. Magnetic sources impose their own thermal ceiling: 2:17-type samarium-cobalt magnets demagnetize at elevated temperatures, although optimized compositions extend their usable range, and one neutron irradiation study reported comparatively small changes in Sm2Co17 magnetic properties. Wetted, electrode-based designs in heavy liquid metals face corrosion, wetting instability, and thermoelectric offsets on top of that.

Recent research has responded with a wave of cleverness aimed at drift, contact, and calibration. Phase-shift designs use relative phase information to reduce sensitivity to conductivity variation, and time-of-flight Lorentz-force velocimetry has been proposed to cut conductivity dependence further. Eddy-current methods that estimate velocity and conductivity simultaneously separate the measurement from a major drift variable. Perhaps most striking is transient eddy-current flow metering, which derives velocity from the advection of a transient eddy-current pattern and is, in principle, calibration-free. Fully contactless Lorentz-force velocimetry has advanced from theory to plant-scale trials in liquid aluminum at 780 degrees Celsius, with an estimated relative flow error of 2.34 percent and dynamic online calibration in metallurgical service. These advances strengthen measurement in highly conducting fluids, but they do not erase the signal-amplitude penalty that low-conductivity molten salts impose.

Ultrasonic techniques offer the broadest demonstrated route to conductivity-independent, non-intrusive measurement across both fluid classes. Ultrasonic Doppler velocimetry is widely used to obtain velocity profiles in opaque liquid metals, and transit-time systems can operate in clamp-on or standoff configurations. The high-temperature limitations, however, span the entire signal chain: conventional PZT elements can degrade, bond layers can fail below nominal Curie temperatures, and couplants may dry, creep, or delaminate. Temperature gradients change sound speed and refract the beam, and corrosion or wetting changes at interfaces can destabilize the acoustic path. Mitigations are maturing. Lithium niobate transducers produced signals after 300 hours in lead-bismuth at 350 degrees Celsius, and separate gamma irradiation tests showed no noticeable changes at 22 MGy, though such component tests do not establish performance under combined thermal and radiation exposure. Gallium phosphate guided-wave transducers have been demonstrated on piping near 600 degrees Celsius, and waveguides and buffer rods move sensitive components away from the hot zone, as shown by Doppler measurements in molten copper-tin at 620 degrees Celsius. In FLiNaK fluoride salt, transit-time ultrasonics was cross-validated against pressure-drop measurements at 600 degrees Celsius, establishing feasibility, although acoustic coupling and thermal-interface stability remain constraints on service life.

The remaining principles each carry their own thermal baggage. Differential-pressure devices infer flow from calibrated pressure loss across restrictions, but thermal expansion changes throat areas and pressure taps, temperature-dependent density and viscosity alter the conversion, and in molten salts the impulse lines can freeze or plug below the salt’s melting range, making trace heating integral to the design. Vortex meters avoid impulse lines but depend on bluff-body geometry and Reynolds number, which thermal expansion and viscosity shifts can push out of the linear regime. Coriolis meters measure mass flow directly from vibrating-tube inertia, attractive when density varies strongly with temperature, but temperature-dependent elastic modulus produces zero and span drift, and drive and pickoff assemblies impose their own temperature ceilings; they have been used in heavy-liquid-metal facilities and in sodium comparisons at 240 to 340 degrees Celsius. Thermal meters, which infer flow from imposed heat and convective transport, depend on properties and surface conditions that corrosion and deposits can change. Fiber-optic sensors offer electromagnetic immunity and radiation tolerance, and a regenerated fiber Bragg grating hot-wire sensor has measured gas velocities up to 800 degrees Celsius, but long-duration demonstrations of packaged fiber-optic flowmeters in liquid metals or molten salts remain limited.

Deployment maturity, the review finds, varies dramatically by fluid class. Sodium enjoys the most mature instrumentation: permanent-magnet and eddy-current flowmeters have documented operating experience in reactors such as FBTR and Phénix, with current development focused on temperature capability and reduced calibration rather than basic feasibility. Heavy liquid metals retain multiple viable options but lack a single dominant standard; permanent-magnet flowmeters in the KALLA lead-bismuth program achieved reported accuracy of 2 to 4 percent after calibration and correction, while chloride-salt Venturi tests showed substantial scatter and corrosion of wetted components. Molten salts are the hardest case, with maturity depending on chemistry: nitrate salts have ultrasonic and vortex demonstrations at the Évora Molten Salt Platform, fluoride salts combine historical Venturi service in the MSRE coolant loop with recent ultrasonic measurements but limited lifetime qualification, and chloride salts still rely largely on customized pressure-based solutions.

The authors frame the field’s future around two engineering strategies and a philosophical shift. Sensor hardening keeps the transduction element near the fluid while raising its temperature capability through high-Curie-temperature piezoelectrics, thermally stable bonding, and heat-tolerant electronics, preserving strong direct coupling but tying service life to hot-zone materials. Standoff and contactless designs instead move sensitive components away or eliminate wetted interfaces, trading coupling strength for survivability. Neither is universally superior. More fundamentally, the review argues for degradation-aware measurement systems: redundancy across independent sensing principles, self-validating sensor frameworks that report measurement status and confidence, calibration-free methods like transient eddy-current metering, and machine learning as a supplement to physics-based calibration for detecting drift. Emerging principles, including thermal-wave metering and thermal time-of-flight methods demonstrated at the Andasol III plant, where 94 percent of analyzed operational data fell within plus or minus 5 percent of full-scale flow, point toward instrumentation that diagnoses its own health.

The central message is sobering but clarifying: no single principle is universally optimal, and the real challenge is not measuring flow at high temperature but maintaining traceable measurement validity throughout extended service. Sodium shows what maturity looks like when fluid properties favor the sensing physics and operational experience is deep. Heavy liquid metals show what happens when corrosion and interface stability dominate. Molten salts show how low conductivity and harsh chemistry can strip away options. Future high-temperature flow instrumentation, the authors conclude, must combine an appropriate primary measurement principle with thermal and condition diagnostics, reduced calibration dependence, and independent cross-checks that reveal degradation before undetected flow errors develop, because in a molten salt reactor or a fusion breeding blanket, an instrument that quietly stops telling the truth may be more dangerous than one that fails loudly.

Subject of Research: Flow measurement techniques for high-temperature liquid metals and molten salts in advanced energy systems

Article Title: Flow measurement in high-temperature liquid metals and molten salts: principles, recent advances, and deployment challenges

Article References: Her, G., Oh, Y., Kim, H., & Kim, S. (2026). Flow measurement in high-temperature liquid metals and molten salts: principles, recent advances, and deployment challenges. Frontiers in Chemistry, 14, Article 1977068. https://doi.org/10.3389/fchem.2026.1977068

Image Credits: AI Generated

DOI: 10.3389/fchem.2026.1977068

Keywords: flow measurement, liquid metals, molten salts, molten salt reactors, sodium-cooled fast reactor, electromagnetic flowmeter, ultrasonic flowmeter, Lorentz force velocimetry, sensor calibration, high-temperature instrumentation, concentrating solar power, fusion breeding blankets

The research team, led by Tong Wang of Shenzhen University, collected soil samples from three sites across a subtropical agricultural district in southern China, capturing three distinct textures: sandy loam, loam, and clay loam. Each sample was characterized for a full suite of physical and chemical properties, including bulk density, organic carbon, pH, electrical conductivity, field capacity, and permanent wilting point. The soils were then packed into pots under controlled laboratory conditions and exposed to seven salinity levels ranging from 2 to 24 decisiemens per meter, a span chosen to mirror real conditions in coastal Guangdong, where routine irrigation water sits near the low end of that range and seawater intrusion during dry periods can push shallow groundwater toward the top.

The salinity solutions were prepared synthetically from sodium chloride and calcium chloride in a fixed 2:1 molar ratio rather than drawn from natural water sources, a deliberate choice that eliminated uncontrolled ionic confounding. Each pot was leached with its designated saline solution until the drainage water matched the irrigation water’s electrical conductivity, ensuring uniform salt distribution. Gypsum blocks installed before treatment were read with a moisture meter that converts direct current to alternating current to prevent electrode polarization, producing readings proportional to soil water content from dry soil to saturation.

Calibration proceeded by oven-drying the soil, saturating it with saline solutions, and letting it air-dry while periodically recording both block resistance and gravimetric moisture determined by weighing. This produced paired measurements of electrical resistance, which ranged from roughly 4 to 244 kilo-ohms, and true volumetric water content across matric suctions from 0.1 to 2.5 bar. From these data the team defined a correction factor, the difference between predicted and measured water content, and set out to model how that correction depends on salinity, suction, and texture simultaneously.

The modeling tool of choice was Response Surface Methodology, or RSM, a statistical framework designed to untangle the main, quadratic, and interactive effects of multiple continuous variables with relatively few experiments. Because salinity in decisiemens per meter and suction in bar occupy incomparable numerical scales, both factors were coded into a standardized dimensionless range between minus one and plus one, giving regression coefficients equal weight. Rather than using a classical designed experiment such as a central composite design, the researchers treated their systematic laboratory matrix as historical data and fit polynomial surfaces directly using Design-Expert software, confirming that the two independent variables remained orthogonal with negligible multicollinearity.

The central technical finding concerns model order. Linear and two-factor-interaction formulations captured general trends but showed significant lack-of-fit, meaning simpler equations could not represent the combined salinity-suction behavior. Quadratic models improved the fit but still left systematic residuals. Only a cubic polynomial, one that includes third-order terms for each factor, achieved non-significant lack-of-fit tests across all three textures, with adjusted R-squared values of 0.93 to 0.95 and predicted R-squared values within 0.03 of them, indicating genuine generalization rather than over-parameterization. The physical rationale is that electrical transport in a saline porous block follows a dual-regime mechanism: bulk electrolytic conduction through free pore water competes with surface conduction along the electrical double layers lining the gypsum matrix, and the transition between these regimes is asymmetric, not the symmetric parabolic decay a quadratic model assumes.

The fitted coefficients reveal how the correction behaves. In loam soil, salinity carried the largest positive linear coefficient at 0.19, with negative quadratic terms signaling diminishing returns at high salinity and suction, and a negative interaction term showing that simultaneous increases in both variables suppress the corrected moisture. In clay loam, salinity dominated even more strongly with a coefficient of 0.248, while interaction terms were not significant, suggesting the fine pore structure buffers suction effects and lets salinity act largely independently. Sandy loam showed the smallest coefficients overall, consistent with its low specific surface area and limited cation exchange capacity. Across all textures, salinity and its square contributed most to the correction, followed by suction and its square.

The performance metrics are striking. The two-factor cubic models achieved R-squared values of 0.97 to 0.98 with root mean square errors between 0.015 and 0.022 cubic centimeters of water per cubic centimeter of soil, and a coefficient of residual mass near zero, indicating negligible systematic bias. Compared with uncorrected raw gypsum block readings, the calibration reduced RMSE by more than 95 percent across all three textures. Accuracy ranked clay loam highest, then loam, then sandy loam, tracking water-holding capacity and the fine-porosity films that preserve conductive pathways around the sensor. Constrained numerical optimization identified the largest required corrections in clay loam, up to 0.064 cubic centimeters per cubic centimeter at 17.6 decisiemens per meter and 2.5 bar suction, with desirability indices near 0.9.

Some of the raw data patterns are physically revealing in their own right. In clay loam at 6 decisiemens per meter, apparent water content stayed nearly flat between 0.36 and 0.39 across the entire suction range, an apparent plateau that the authors attribute not to a lack of real drainage but to a balance between the resistance increase caused by capillary drying and the conductivity boost from ions concentrated in the shrinking pore water. At moderate salinity, the ionic conductivity of remaining pore water partially compensates for drainage-induced resistance; at higher salinity, thinning water films and increased tortuosity progressively limit ionic transport, producing the strongly nonlinear, non-monotonic response that only the cubic surface can capture.

The practical implications extend well beyond one province. Gypsum blocks remain the backbone of low-cost, distributed irrigation monitoring in greenhouses and smallholder schemes worldwide, precisely because they are cheap, robust, and easy to automate. This study demonstrates that their salt vulnerability is a calibration problem, not a reason to abandon the technology, and that a compact polynomial correction, parameterized per soil texture, can restore accuracy comparable to far more expensive instrumentation. The authors are careful to note the limits: the models are empirical, validated only within the tested ranges of salinity, suction, and texture, and should not be extrapolated beyond them. Field-scale validation across additional soils remains the necessary next step, but for saline agricultural regions from coastal China to arid irrigation districts everywhere, the message is that a seventy-year-old sensor, armed with the right mathematics, can still compete at the precision frontier.

Subject of Research: Response surface calibration of gypsum block soil moisture sensors under salinity and matric suction

Article Title: A response surface-based calibration framework for accurate soil water content determination under saline conditions: a case study in Guangdong Province, China

Article References: Wang, T., Tong, R., Chen, Y., & Wu, X. (2026). A response surface-based calibration framework for accurate soil water content determination under saline conditions: a case study in Guangdong Province, China. Environmental Earth Sciences, 85(16), Article 426. https://doi.org/10.1007/s12665-026-13168-w

Image Credits: AI Generated

DOI: 10.1007/s12665-026-13168-w

Keywords: soil water content, gypsum blocks, sensor calibration, soil salinity, matric suction, response surface methodology, Guangdong, irrigation management, soil texture, electrical resistance, saline agriculture, soil physics

The stakes are considerable. Nitrogen, phosphorus, and potassium regulate chlorophyll synthesis, root development, photosynthesis, enzyme activation, and osmotic regulation in crops, making them the backbone of any fertilizer program. Yet across much of the world, fertility assessment still depends on labor-intensive laboratory chemistry. In India, the Soil Health Card (SHC) program provides standardized laboratory testing and fertilizer recommendations, but sample transport, processing, and delayed reporting make it poorly suited to frequent monitoring, particularly for resource-constrained farmers. Meanwhile, intensive cultivation and indiscriminate fertilizer use have degraded soils in many agricultural regions, sharpening the need for timely, site-specific nutrient information that precision agriculture demands.

Low-cost electrochemical NPK sensors have long promised a solution, but they carry a well-known flaw: their readings drift with the very conditions they are embedded in. Sensor responses are influenced not only by nutrient concentrations but also by soil texture, moisture content, salinity, pH, ionic mobility, and electrical conductivity. Identical nutrient levels can produce different electrical signatures in a wet clay soil and a dry sandy one. As a result, raw sensor outputs deviate nonlinearly from laboratory measurements, and models trained on one dataset often fail when applied to heterogeneous fields. The Dharwad team’s central insight was that this physicochemical interference should not be treated as noise to be minimized, but as information to be modeled explicitly.

To build that model, the researchers assembled a dataset grounded in real agricultural variability. They collected 150 representative soil samples from upland, midland, and lowland zones of Dharwad district in Karnataka, India—fifty from each physiographic region—drawing from the surface 0–20 centimeter layer that constitutes the active root zone for most crops. Within each field, multiple subsamples were gathered in a zigzag pattern and mixed into composite samples, following standard agronomic practice. Each sample was then measured twice: once in the field with the sensor platform, and once at an authorized Soil Health Card laboratory, creating synchronized pairs of raw sensor observations and laboratory reference values for supervised learning.

The sensing hardware itself is deliberately modest. A commercial 7-in-1 electrochemical probe measures nitrogen, phosphorus, potassium, pH, electrical conductivity, moisture, and temperature simultaneously. It communicates over an RS485 serial link, via a MAX485 transceiver running the Modbus RTU protocol, to an ESP32 microcontroller programmed through the Arduino IDE. In the laboratory the system ran on a regulated 12-volt supply; in the field it drew power from portable 12-volt rechargeable batteries, allowing continuous acquisition without external infrastructure. Sensor outputs, reported in milligrams per kilogram, were converted to kilograms per hectare using a soil-mass conversion factor of 2.24, corresponding to an assumed bulk density of 1.12 grams per cubic centimeter, so that predictions would align with SHC reporting conventions.

The calibration engine is a MultiOutput XGBoost regression model, an ensemble method built on gradient-boosted decision trees. Because standard XGBoost handles only a single target, the team used Scikit-learn’s MultiOutputRegressor wrapper to train three independent XGBoost regressors—one each for nitrogen, phosphorus, and potassium—sharing the same engineered feature matrix and identical hyperparameters. The final configuration, selected through five-fold cross-validation on the training data, used 200 trees, a learning rate of 0.05, a maximum tree depth of 6, a minimum child weight of 3, and subsampling and column-sampling ratios of 0.80, with L2 regularization set to 1.0. Regularization and subsampling are what allow the model to capture complex nonlinear feature interactions while resisting overfitting on a modest dataset.

The most conceptually elegant contribution is a single engineered feature: the product of soil pH and electrical conductivity. The rationale is rooted in soil chemistry. pH governs nutrient solubility, ionic speciation, and phosphorus fixation, while electrical conductivity reflects the dissolved ion concentration that drives current flow during electrochemical measurement. Because the sensor operates by measuring ionic responses, the combined pH–EC effect carries information that neither variable reveals alone. An ablation study confirmed the value of this domain knowledge: adding the pH × EC interaction raised the coefficient of determination from 0.903 to 0.916 and reduced root mean square error from 8.92 to 8.37 kilograms per hectare, a 6.17 percent error reduction under identical training conditions.

On the held-out test set of 30 samples, the calibrated model delivered laboratory-equivalent estimates with R² values of 0.916 for nitrogen, 0.887 for phosphorus, and 0.903 for potassium, with root mean square errors of 27.11, 4.01, and 22.84 kilograms per hectare respectively. Nitrogen performed best, likely because its electrochemical response was more consistent, while phosphorus lagged slightly, consistent with its low mobility and strong adsorption–desorption behavior in soils. The framework outperformed linear regression, decision trees, random forests, and support vector regression on the same data, achieving an average R² of 0.902 against 0.881 for random forest and 0.766 for linear regression. Five-fold cross-validation showed remarkable stability, with standard deviations of only about 0.003 in R², and 95 percent confidence intervals remained narrow across all three nutrients.

Equally important is how the team interrogated their own model. Residual analysis found mean errors close to zero—0.18, −0.05, and 0.24 kilograms per hectare for nitrogen, phosphorus, and potassium—with both positive and negative deviations, indicating no systematic over- or underestimation. Bland–Altman agreement analysis, a technique borrowed from clinical measurement science, showed that differences between calibrated predictions and laboratory values clustered around zero, with most observations falling within the 95 percent limits of agreement. A region-wise breakdown confirmed that performance held up across upland, midland, and lowland soils, with nitrogen R² ranging only from 0.899 to 0.928 across zones—evidence that the calibration generalizes across the physiographic heterogeneity it was designed to capture.

The authors are candid about limitations. The study did not quantify long-term sensor drift from electrode aging and environmental exposure, a known problem for low-cost electrochemical probes; prior work cited in the paper shows that inter- and intra-sensor variability can be substantial and that continuous, sensor-specific calibration reduces measurement error. The laboratory reference values themselves lacked an explicit uncertainty budget, and the model was trained on soils from a single district, so validation across diverse agro-climatic zones will be needed before large-scale deployment. Future work will extend the framework to organic carbon, micronutrients, and cation exchange capacity, and will explore adaptive drift compensation, explainable AI, and edge deployment on embedded IoT platforms. Even so, the demonstration stands: for a few dollars of hardware and a well-designed learning algorithm, the gap between instant field readings and trusted laboratory numbers has been narrowed to a level that could put real-time soil intelligence within reach of every farmer.

Subject of Research: Machine learning calibration of low-cost electrochemical sensors for real-time soil macronutrient estimation

Article Title: Real-time soil nutrient estimation using low-cost NPK sensors and soil-aware machine learning calibration

Article References: Kabbur, A. M., & Pawar, P. (2026). Real-time soil nutrient estimation using low-cost NPK sensors and soil-aware machine learning calibration. Smart Agricultural Technology, 15, Article 102582. https://doi.org/10.1016/j.atech.2026.102582

Image Credits: AI Generated

DOI: 10.1016/j.atech.2026.102582

Keywords: soil sensors, NPK nutrients, machine learning, XGBoost, precision agriculture, IoT, ESP32, Soil Health Card, soil fertility, sensor calibration, electrochemical sensing, India

The research team, led by Elisa Pérez Martín and Juan Francisco Fuentes-Pérez of the University of Valladolid, set out to answer a question that calibration-heavy laboratory studies have largely ignored: which sensors actually work out of the box, under realistic adoption constraints, for a farmer who lacks a soil physics laboratory? The distinction matters enormously. Previous comparisons tended to focus on how well each sensor could be calibrated for a specific soil, treating calibration as a routine step. The Spanish team instead framed the problem as what they call a calibration paradox: inexpensive sensors look attractive on a price list, yet their real-world use can demand substantial time, expertise, and soil-specific correction work that transforms a three-dollar bargain into a hidden technical burden.

To build a fair comparison, the researchers assembled a lineup spanning the entire technological and economic spectrum, from resistive probes selling for around three dollars to a TDR instrument priced above seven hundred dollars. The seventeen devices included three resistive sensors, four capacitive sensors, seven frequency domain reflectometry (FDR) probes, two time domain reflectometry (TDR) systems, and one co-axial impedance dielectric reflectometry (CIDR) sensor. Because these devices speak entirely different electronic languages—analog voltages, I2C digital signals, RS-485 Modbus commands, UART, and SDI-12—the team built a unified acquisition platform in Python, using Arduino microcontrollers and protocol adapters to interrogate every sensor under a common framework. The complete scripts, documentation, and datasets were released in a public GitHub repository archived on Zenodo, an unusually transparent move for a sensor benchmark.

The experimental design was deliberately simple and reproducible. Three agricultural soil textures—sandy, clay loam, and clay—were air-dried, homogenized, and wetted to six moisture levels each, with gravimetric water content determined by oven-drying subsamples at 105 degrees Celsius for twenty-four hours to serve as the reference truth. Sensors were inserted in randomized groups of up to four, separated by at least five centimeters, with multiple readings averaged per measurement cycle. The team then evaluated each device in two complementary ways: first, how well its native, uncalibrated output tracked the reference moisture trend, quantified by the coefficient of determination R²; and second, how much quantitative error remained after applying a single global linear calibration, expressed as mean absolute error and root mean square error.

The results split the field cleanly into two camps. Sensors that report moisture-related outputs directly, or through manufacturer-provided conversions, generally followed the reference trend, while those evaluated through raw voltage or capacitance signals showed non-linear behavior, limited dynamic range, and early saturation. The standout performer was the S-Soil MTEC-02A, an RS-485 industrial sensor from Seeed Studio priced at roughly ninety-four dollars, which achieved an overall R² of 0.96 and the lowest global errors of the entire benchmark, with a mean absolute error of just 1.77 percentage points. It was followed closely by the HydraProbe CIDR sensor, the Acclima TDR-315H, and the METER Group TEROS 11, all premium instruments in the three-hundred-to-five-hundred-dollar range that posted R² values between 0.94 and 0.95.

At the other end of the table, the bargain bin told a cautionary tale. The Adafruit STEMMA capacitive sensor, the Grove Moisture Sensor, the DFRobot SEN0114, and the Waveshare 9527—all costing under five dollars—produced overall R² values as low as 0.24 and root mean square errors approaching or exceeding nine percentage points. The STEMMA sensor even spent much of its measurement range pinned near its upper response limit, while the SEN0193 and SEN0308 showed the perverse behavior of decreasing output as soil got wetter. These resistive and simple capacitive devices are sensitive to salinity, prone to corrosion, and difficult to translate into quantitative moisture estimates without soil-specific calibration work that most farmers will never perform.

Crucially, the team did not stop at raw accuracy. They folded five criteria—durability, farmer-sensor interaction complexity, accuracy, reliability, and economic cost—into a weighted composite index designed to reflect practical deployment, with accuracy and reliability weighted most heavily at 0.30 and 0.25 respectively. The verdict upended the price hierarchy. The mid-priced S-Soil MTEC-02A topped the ranking with a score of 4.80, narrowly beating the forty-dollar ComWinTop CWT-Soil-THCPH-S at 4.75, while the Vegetronix VH400 and the TDR-315H tied at 4.40. The premium HydraProbe and TEROS 11 scored 4.20, dragged down by their cost. The cheapest sensors clustered at the bottom, with the STEMMA scoring just 1.95. Performance, in other words, is not a linear function of price.

Soil texture emerged as an independent source of trouble. Several sensors showed reduced agreement and larger errors in clay soil, which the authors attribute to the complex dielectric behavior of fine-textured materials: higher water retention, bound water on clay mineral surfaces, and elevated electrical conductivity all distort the relationship between electrical response and actual water content, particularly for capacitance and lower-frequency FDR devices. This finding carries a practical warning for anyone deploying sensors in heavy soils, where even good instruments may need site-specific attention. The authors also note that the study used a single unit of each sensor model, so sensor-to-sensor variability within a product family remains untested, and that laboratory-prepared, homogenized soils cannot fully reproduce the stones, air gaps, and compaction heterogeneity of real fields.

The implications reach beyond sensor shopping lists. The authors argue that very low-cost sensors should not be considered substitutes for TDR, CIDR, or robust FDR instruments in quantitative irrigation management, because their poor linearity and large errors conceal a technical burden of custom electronics, signal interpretation, and calibration. Conversely, mid-range FDR-based devices may enable denser monitoring networks, capturing the spatial heterogeneity of soil moisture that sparse, expensive probes inevitably miss. There is also a question of technological sovereignty: sensors with opaque proprietary conversions lock farmers into vendor ecosystems, while devices with transparent interfaces and standard protocols democratize precision irrigation. And even where absolute quantification is uncertain, the study suggests each sensor develops a characteristic soil signature—a relative response pattern that can still support irrigation scheduling and anomaly detection.

For a sector under pressure to do more with less water, the message is refreshingly concrete. The most economically viable path to precision irrigation, this benchmark suggests, runs not through the cheapest gadgets or the most expensive instruments, but through the often-overlooked middle of the market: robust, well-interfaced probes around forty to one hundred dollars that deliver reference-grade trend agreement without demanding a calibration engineer in the cab of every tractor. Before the next wave of smart-farming deployments buries thousands of three-dollar sensors in the world’s fields, this study offers a simple piece of advice backed by data: check what comes out of the box, because the true cost of a sensor is measured in expertise as much as in dollars.

Subject of Research: Comparative benchmarking of commercially available soil moisture sensors for cost-effective precision irrigation

Article Title: Out-of-the-box performance and cost-effectiveness of soil moisture sensors for precision irrigation

Article References: Martín, E. P., Moreno, J. M., de Aza, C. H., Álvarez, F. L., Turrión, M. B., Sanz-Ronda, F. J., & Fuentes-Pérez, J. F. (2026). Out-of-the-box performance and cost-effectiveness of soil moisture sensors for precision irrigation. Smart Agricultural Technology, 15, Article 102564. https://doi.org/10.1016/j.atech.2026.102564

Image Credits: AI Generated

DOI: 10.1016/j.atech.2026.102564

Keywords: soil moisture sensors, precision irrigation, FDR, TDR, capacitive sensors, resistive sensors, sensor calibration, cost-effectiveness, soil texture, smart farming, water management, agricultural technology

The study tackles a distinction that is easy to overlook but central to soil physics. Volumetric soil water content, symbolized as θv, measures the sheer quantity of water stored in the soil, expressed as a percentage of soil volume. Soil water potential, denoted Ψ, measures something subtly different: the energy status of that water, or how much suction a plant root must exert to pull it out. Two soils can hold identical amounts of water while offering dramatically different availability to crops, because the force binding water to soil particles depends on pore size, and pore size depends on texture. Sand, with its large macropores, releases water readily but drains quickly. Clay, packed with micropores, clings to water tenaciously even when quantities look abundant.

To untangle these relationships, the team conducted an outdoor experiment at the Edisto Research and Education Center in Blackville, South Carolina, from late November 2017 to early February 2018. Rather than digging up twelve naturally occurring soils, they constructed the textural spectrum themselves, mixing commercially available sand, silt, and clay in precise proportions defined by the USDA classification system. This yielded twelve soil mixtures representing sand, loamy sand, sandy loam, loam, silt loam, silt, sandy clay loam, clay loam, silty clay loam, sandy clay, silty clay, and clay. Each soil was packed into replicate plastic containers, each holding 700 cubic centimeters, and instrumented with two affordable and widely deployed sensors: the Decagon EC-5, a capacitance-based device that estimates volumetric water content by measuring dielectric permittivity at 70 MHz, and the Watermark 200SS, a granular matrix sensor that gauges soil water potential through electrical resistance within a hydrated gypsum matrix.

The experimental protocol was elegantly simple. The researchers saturated each container with water, then let the soil dry naturally under ambient outdoor conditions while recording sensor outputs and total system weight every morning at nine. Because the container weights were known when dry and wet, the team could compute gravimetrically determined water content at every time point, providing a trusted reference against which to judge both sensors. Particle-size distributions were verified with the hydrometer method, and bulk density was calculated from the oven-dry mass packed into each known volume. Field capacity, permanent wilting point, and available water were then estimated for each texture using the generalized soil water characteristic equations of Saxton and colleagues.

The drying patterns that emerged were starkly texture-dependent. Clay-rich soils began the experiment holding enormous quantities of water, with clay at roughly 50 percent volumetric water content, silty clay at 47 percent, and sandy clay at 45 percent. Sand, by contrast, started at only about 15 percent and loamy sand at 20 percent. As drying progressed, fine-textured soils retained residual water contents of 8 to 12 percent while coarse soils fell to just 2 to 5 percent, a direct consequence of pore-size distribution. The Watermark sensors told the complementary energy story: near saturation, all soils read close to minus 10 kilopascals, but sandy soils plummeted rapidly toward minus 150 to minus 200 kilopascals, the sensor’s practical detection limit, while clay and silty clay lingered between minus 40 and minus 60 kilopascals far longer, releasing their water grudgingly.

Perhaps the most practically valuable result came from the team’s use of segmented regression, a statistical technique that locates breakpoints in nonlinear relationships. Applied to the drying curves, this analysis identified threshold soil water potential values, the points beyond which a small loss of water content triggers a steep drop in water potential and a corresponding crash in plant availability. Across all textures, average thresholds landed at approximately 40 kilopascals for the gravimetric-Watermark pairing and 44 kilopascals for the EC-5-Watermark pairing, but individual textures ranged widely, from minus 18 to minus 52 kilopascals in the gravimetric comparisons. These breakpoints, the authors argue, offer texture-specific reference points for irrigation scheduling that a single universal threshold simply cannot provide.

The calibration performance of the sensors themselves also diverged by texture. The Decagon EC-5 showed outstanding agreement with gravimetric measurements, with coefficients of determination between 0.987 and 0.997 across all twelve soils, and root mean square errors from just 0.29 percent in sand to 4.94 percent in clay. Polynomial models, mostly quadratic or cubic, provided the best fit, including sand at R² of 0.994, sandy clay at 0.992, silty clay at 0.997, and clay loam at 0.995. The higher errors in clay-dominated soils reflect the greater variability in dielectric response that clay content introduces, reinforcing a theme from the broader sensor literature that soil-specific calibration beats factory defaults. Under extremely dry conditions, the EC-5 even produced slightly negative readings in sand, an artifact of diminished dielectric contrast and poor probe-soil contact in nearly waterless coarse material.

The Watermark sensor, meanwhile, proved more texture-sensitive. Its relationship with gravimetrically measured water content ranged from a moderate R² of 0.745 in sand to a strong 0.970 in clay and silty clay, consistent with earlier reports that granular matrix sensors struggle in low-water-retention sandy profiles. Yet comparisons between the Watermark’s potential readings and the EC-5’s content readings remained consistently strong across textures, with R² values from 0.896 to 0.978, suggesting the two sensing principles can be meaningfully linked once soil-specific calibration curves are in place. Such linkage matters because capacitance and resistance sensors answer different questions: one reports how much water is present, the other how available it is to roots.

The authors are candid about the study’s boundaries. The experiment used disturbed, prepared soil mixtures under outdoor container conditions, so real-world complications like soil structure, organic matter, root activity, and weather variability were not captured. The Watermark’s operating range of roughly 0 to minus 200 kilopascals also left the dry end of the retention curve, including the permanent wilting point near minus 1500 kilopascals, outside measurable reach, and hydraulic properties were estimated from texture rather than measured with pressure-plate apparatus. These constraints prevented fitting mechanistic models such as the van Genuchten equation. Still, the empirical relationships developed here, spanning all twelve USDA textural classes under a single unified framework, appear to be the first of their kind reported for South Carolina, a state where irrigated acreage is expanding rapidly across highly heterogeneous soils.

The practical message is resonating in an era when smart irrigation systems promise water savings through automation. As the study concludes, accurate interpretation of soil moisture data demands that soil texture and soil-specific relationships between water content and water potential be considered alongside sensor calibration. A grower reading minus 40 kilopascals in a loamy sand is witnessing a very different soil condition than one reading minus 40 kilopascals in clay, and irrigating both fields identically will inevitably overwater one and shortchange the other. The texture-specific thresholds and calibration curves published in this work offer a concrete starting point for building such nuance into irrigation decision tools, though the authors stress that field validation across diverse crops and climates is still required before widespread deployment.

Subject of Research: Empirical relationships between soil moisture and soil water potential across soil textural classes for irrigation management

Article Title: Influence of soil texture on soil moisture and soil water potential dynamics

Article References: Payero, J. O., & Selvalakshmi, S. (2026). Influence of soil texture on soil moisture and soil water potential dynamics. Discover Soil, 3(1), Article 152. https://doi.org/10.1007/s44378-026-00305-x

Image Credits: AI Generated

DOI: 10.1007/s44378-026-00305-x

Keywords: soil texture, soil moisture, soil water potential, irrigation management, sensor calibration, water retention, Decagon EC-5, Watermark 200SS, field capacity, available water, soil physics, precision agriculture