The stakes could hardly be higher. According to figures cited in the study, air pollution is now linked to as many as 8.1 million deaths globally each year, and fine particulate matter—particles smaller than 2.5 microns, roughly one-thirtieth the width of a human hair—is among the most insidious culprits. Too small to be seen with the naked eye, PM2.5 penetrates deep into the lungs and crosses into the bloodstream, driving respiratory disease, cardiovascular damage, and lung cancer. Thailand’s problem is compounded by a lethal cocktail of sources: vehicle and industrial fuel combustion, the seasonal burning of rice straw and agricultural residues, forest clearing in the north, and household burning of wood and fossil fuels. In provinces such as Chiang Mai, Lamphun, and Phitsanulok, calm winter winds and temperature inversions—cold air trapped beneath warm—lock the dust in the lower atmosphere, sometimes for days.

What separates Cleara-Air from the growing crowd of low-cost air-quality monitors is that it refuses to be a passive observer. Most existing IoT platforms, the researchers note, focus on data collection, visualization, or machine-learning forecasts. Cleara-Air closes the loop. At its heart sits an ESP32 microcontroller wired over the I2C protocol to a Sensirion SPS30 laser-scattering sensor, which measures PM1.0, PM2.5, PM4.0, and PM10 continuously. The controller compares every reading against an operational set-point of 37.5 micrograms per cubic meter—Thailand’s 24-hour PM2.5 standard, a number Thai citizens already recognize as a health warning. When the concentration climbs above that line, the ESP32 sends a HIGH logic signal to two GPIO pins, driving a 4-channel 5V relay that switches on a 220-volt AC ventilation fan and a carbon-brush negative-anion generator. When levels fall back below the threshold, the actuators shut off to save energy.

The actuation mechanism itself is a piece of applied electrostatics. Negative ions released at high voltage from the generator’s discharge tip attach to suspended particles and alter their surface charge, increasing the probability that particles collide and agglomerate. Fine dust that would otherwise linger in the breathing zone clumps into larger, heavier aggregates that settle out of the air more rapidly, while the fan distributes the ion cloud through the room. The team deliberately chose a carbon-brush generator over a conventional corona-discharge module to reduce the potential for ozone formation—a well-known byproduct of traditional ionizers and a genuine safety concern. The researchers are careful, however, to state that no physiological benefits of negative ions are claimed, and that direct ozone concentrations were not measured in this study.

Every fifteen seconds, the ESP32 packages PM1.0, PM2.5, and PM10 readings together with the device’s ON/OFF status and transmits them over Wi-Fi to a PostgreSQL database via HTTP POST. From there, the data flow into a responsive Web-GIS application built with PHP, HTML5, CSS3, JavaScript, and the Leaflet mapping library, which displays sensor locations, real-time concentrations, historical trends, and actuator status on any modern browser—computer, tablet, or smartphone. The database schema links four tables: Devices, which stores installation coordinates; Sensor_Logs, which records measurements and ionizer status as time-series entries; Command_Logs, which captures remote-control commands; and Users. The architecture is entirely open-source, from PostgreSQL 18.1 and Arduino IDE 2.3.10 to QGIS 3.40.6, and the total hardware bill comes to roughly $65, with the SPS30 sensor at $20.91 and the ESP32 board at $11.75.

Engineering the closed loop was not without its traps. Because the 220-volt fan is an inductive load, switching the relay off generates electromagnetic interference that can freeze the sensor’s I2C bus—a failure mode the team calls an I2C hang. Rather than inserting crude delay commands, the firmware uses the millis() function so the processor continuously checks its own status, with a 1,000-millisecond I2C timeout and an error counter: if the sensor fails to respond three times in a row, the microcontroller performs a hard reset within three seconds, clears the bus, and resumes operation without interruption. It is the kind of unglamorous resilience engineering that determines whether a real-world device survives contact with mains electricity.

The experimental evaluation was refreshingly blunt. In fifteen independent trials conducted in a closed room, the researchers deliberately elevated particulate concentrations by introducing powdered flour roughly fifteen centimeters from the sensor inlet. Each trial followed the same scripted sequence: continuous measurement, detection of a threshold crossing, relay activation of the fan and ionizer, transmission of time-stamped data and device status to PostgreSQL at fifteen-second intervals, visualization of the updated status on the Web-GIS interface, and finally actuator deactivation once the reading fell back below the set-point. Database records from the trials capture the drama in miniature—one entry shows the ionizer switching on the instant a particulate event pushed PM2.5 to 75 micrograms per cubic meter, with a later record marking its shutdown. The repeated tests demonstrated that the programmed sequence operated reproducibly under the tested conditions.

The researchers are equally candid about what the trials do not prove. The study evaluates system functionality, not PM2.5 removal efficacy. There was no untreated control condition, no particle-decay analysis, no comparison with established filtration technologies, and no co-location of the SPS30 with a reference-grade instrument—so measurement bias, error metrics, and humidity-dependent uncertainty remain unquantified. Ozone was not directly measured, formal electrical safety certification was not performed, and the bare-comparator threshold rule could cause rapid relay cycling as concentrations fluctuate around the set-point; the team recommends adding a hysteresis dead-band or minimum on-time before any operational deployment. The laboratory tests also used unencrypted HTTP on a local server, meaning secure HTTPS, authentication, and credential protection must precede any field installation.

Conceptually, the team frames Cleara-Air across five dimensions of urban informatics: urban sensing, urban data, urban computing, urban science, and urban applications. The SPS30 provides the sensing layer; time-stamped, georeferenced PostgreSQL records form the data layer; the ESP32 edge controller and server-side processing constitute the computing layer; GIS and Web-GIS supply the spatial representation; and the integrated platform delivers the application. The researchers position the prototype primarily as system-level urban environmental informatics infrastructure rather than a city-scale analysis—one unit cannot characterize how pollution varies across a province. But the framework is designed to scale: a distributed network of calibrated devices with standardized siting criteria and quality-control flags could eventually support investigations of intra-urban PM2.5 variability, population exposure, environmental inequality, and spatially targeted planning interventions.

For a medium-sized regional city in a data-constrained Global South context—Phitsanulok Province, with its roughly 74 percent average relative humidity and patchwork of mountains and low plains, was selected as the target environment—the appeal of a $65, open-source, self-acting air-quality guardian is obvious. Municipalities that cannot afford dense networks of reference-grade monitors could deploy Cleara-Air-class devices to raise public awareness, automate localized mitigation, and feed decision-ready information to policymakers. The next validation phases, the authors say, will include sensor calibration against reference instruments, replicated treatment-versus-control chamber experiments, direct ozone monitoring, and long-duration evaluation of communication reliability and energy consumption. If those hurdles are cleared, the humble flour-dust trials of this prototype may be remembered as the moment urban air-quality informatics stopped merely watching the smog—and started doing something about it.

Subject of Research: IoT and Web-GIS closed-loop system for PM2.5 monitoring and automated ionization actuation

Article Title: Development and experimental evaluation of an IoT-based closed-loop system for PM2.5 monitoring and automated electrical ionization actuation

Article References: Development and experimental evaluation of an IoT-based closed-loop system for PM2.5 monitoring and automated electrical ionization actuation. (n.d.). https://doi.org/10.3389/frsus.2026.1939212

Image Credits: AI Generated

DOI: 10.3389/frsus.2026.1939212

Keywords: Development, experimental, evaluation, IoT-based, closed-loop, system, monitoring, automated, electrical, ionization, actuation, scientific research

The clinical motivation is stark. In Europe, people face roughly a 25 percent lifetime chance of developing hip osteoarthritis, a condition that drives disability, diminished quality of life, and escalating healthcare costs. Acetabular dysplasia, a structural abnormality in which the hip socket fails to adequately cover the femoral head, is recognized as one of the strongest known risk factors for early degenerative changes. Because early detection of such abnormalities could significantly alter disease progression, particularly in young adults, the research team set out to close a conspicuous gap: there are essentially no reference data for three-dimensional hip morphology in this age group, since population studies have focused almost exclusively on older adults.

The technical foundation of the study is an nnU-Net deep learning model, a self-configuring convolutional neural network architecture that has become a benchmark for biomedical image segmentation. The model was trained on just 40 manually segmented hips, drawn from 20 randomly selected participants with equal representation of males and females, using 5-fold cross-validation with both 2D and 3D full-resolution configurations ensembled across folds and then across configurations. Each fold was trained for 1,000 epochs. The model’s inputs were the water-only and in-phase images from a Dixon LavaFlex sequence acquired on a 3 Tesla MRI scanner, with an in-plane resolution of 0.89 by 0.89 square millimeters and a slice thickness of 1.2 millimeters. From these images, the network automatically segmented four key structures in each hip: femoral bone, acetabular bone, femoral cartilage, and acetabular cartilage.

Performance was impressive for bone and respectable for cartilage. On a hold-out test set of 10 hips, the model achieved Dice similarity coefficients of 0.97 for femoral bone, 0.90 for acetabular bone, 0.77 for acetabular cartilage, and 0.77 for femoral cartilage, with mean surface distances smaller than a single voxel for all structures. The femoral bone score matches the highest reported values in the literature for MRI-based bone segmentation. The acetabular bone score is lower, likely because the acetabulum lacks a clear anatomical boundary, requiring the model to learn not only relevant image features but also where segmentation should end. Cartilage segmentation is inherently harder because hip cartilage is thin and vulnerable to partial volume effects, though the authors note that previous studies achieving higher scores often treated cartilage as a single entity rather than separately segmenting femoral and acetabular components as this method does.

Once segmentations were complete, the team extracted a battery of quantitative imaging biomarkers in five categories: basic geometric metrics, cartilage volumes, angular measurements, coverage metrics, and joint space width. The extraction pipeline used Python with the Visualization Toolkit, beginning with discrete marching cubes surface extraction followed by windowed sinc smoothing to eliminate voxelation artifacts. Random sample consensus fitting located the femoral head and acetabular fossa as spheres, while least-squares fitting defined the femoral shaft axis. The femoral head-neck axis was defined by expanding the fitted head sphere by 40 percent, a proportion demonstrated to be the most robust approach, and fitting a circle to the resulting neck cross-section. Acetabular rim extraction used a best-fitting projection plane, 2D Delaunay triangulation, and convex hull refinement. The acetabular clockface was established in 30-degree increments using the acetabular fossa direction vector, the midpoint of the acetabular notch, and the femoral head center.

Coverage metrics quantified how much of the femoral head is overlaid by the acetabular fossa, calculated as whole coverage, upper coverage, and quarter-region coverage defined by planes aligned with the femoral neck axis. Joint space width was measured as mean and minimum values over a region of interest in the acetabular fossa, selected through distance-based Otsu thresholding, and subdivided into eight sections for regional analysis. Alpha angles, which assess femoral head-neck junction morphology, were calculated by projecting the segmented femur onto coronal and axial planes and identifying where the projected surface deviated from a fitted circle by more than 2 millimeters. In total, the pipeline computed 37 biomarkers per hip, all from the segmentation masks alone.

Quality control was layered throughout. An automated check confirmed that each hip segmentation formed a single connected component and that left and right labels were correctly assigned, with incorrect labels automatically swapped and multi-component segmentations excluded. Cases with biomarker values more than 2 standard deviations from the mean were visually reviewed to catch segmentation errors the automated checks missed. A sensitivity analysis on 40 hips simulated two common segmentation deficiencies, incomplete acetabular fossa segmentation and under-segmentation at the cup edges, and found that most biomarkers remained stable, with only tilt and version showing notable shifts. The overall exclusion rate was just 1.8 percent, underscoring the robustness of the automated approach.

The population-level findings are striking. Females had smaller femoral and acetabular radii, reduced cartilage volumes on both femoral and acetabular sides, smaller alpha angles in both coronal and axial planes, and greater acetabular version compared to males. Interestingly, femoral cartilage volume differed significantly between sexes, but the normalized femoral cartilage volume did not, suggesting that normalization by the squared femoral head radius effectively removes size dependency. BMI-related differences were generally small in magnitude, with effect size analysis indicating that statistical significance may partly reflect the large sample size rather than substantial differences. After stratifying by sex, some BMI differences lost significance, and males showed highest cartilage volumes in the normal weight group while females showed a more consistent increase in normalized cartilage volumes with higher BMI.

Correlation analyses revealed coherent relationships among biomarkers. Coverage measurements showed consistent positive correlations with each other, with upper coverage particularly strongly correlated with the 12 o’clock coverage angle at r equals 0.85. Joint space width measures were inversely correlated with coverage, with whole coverage, upper coverage, and the 12 o’clock angle all showing negative correlations with mean and minimal JSW, and a moderate negative association of r equals negative 0.5 between coverage and mean JSW that aligns with previous findings. Tilt and version correlated with coverage angles in ways consistent with their definitions of acetabular orientation. Regional JSW analysis showed that anterior values were consistently larger than posterior values, and strong correlations between overall minimal JSW and regional values in specific areas demonstrated the value of spatially resolved joint space mapping.

The study’s reference values, reported as percentiles for each biomarker, provide a framework for research and exploratory clinical analysis rather than diagnostic thresholds. The authors emphasize that these values are not intended for identifying hip dysplasia or femoroacetabular impingement, and that the relationship between these 3D biomarkers and clinical outcomes remains unknown. Future work should validate them by correlating with clinical assessments, radiographic disease markers, and patient-reported outcomes, both cross-sectionally and longitudinally. Because the biomarkers are calculated from segmentation masks, the method can be applied to other 3D imaging modalities including CT, supporting broader use in research and clinical practice. The authors propose a minimal subset of biomarkers, including normalized cartilage volumes, whole coverage, and regional joint space width measures, as the most promising starting point. All developed code and model weights are openly available, and the standardized approach could be used to investigate morphological changes in longitudinal studies of hip osteoarthritis progression, assess pre- and post-operative changes in hip morphology, and support surgical planning and outcome evaluation.

Subject of Research: Automated 3D Analysis of Hip Morphology in a Young Adult Population

Article Title: Automated 3D Analysis of Hip Morphology in a Young Adult Population

Article References: Kamphuis, M. A., Oei, E. H. G., Runhaar, J., Hanff, D. F., Tolk, J. J., Agricola, R., Bierma-Zeinstra, S. M. A., Klein, S., & Hirvasniemi, J. (2026). Automated 3D Analysis of Hip Morphology in a Young Adult Population. Annals of Biomedical Engineering. https://doi.org/10.1007/s10439-026-04418-w

Image Credits: AI Generated

DOI: 10.1007/s10439-026-04418-w

Keywords: deep learning, hip morphology, MRI segmentation, nnU-Net, osteoarthritis, Generation R cohort, biomedical engineering, Automated, Analysis, Morphology, Young, Adult