Triaxial Sensor Research Articles

586-P: Light Physical Activity Prevents the Progression of Diabetic Nephropathy in Patients with Type 2 Diabetes

Background and aim: Diabetic nephropathy is one of the most important complications of diabetes mellitus. In recent years, aerobic exercise has been reported to be potentially effective in preventing nephropathy. However, the relationship with light physical activity (LPA), activities of daily living, is not clear. In this study, we researched the effect of LPA assessed by tri-axial acceleration sensors on the progression of nephropathy in patients with type 2 diabetes Materials and methods: We analyzed retrospectively the cohort study of type 2 diabetic patients (T2DM) with urinary albumin-to-creatinine ratio (UACR) under 300 mg/gCr who visited Ayabe City Hospital from April 2018 to March 2020 and were observed until May 2022. The progression of nephropathy was defined as an increase in the category of diabetic nephropathy. LPA time was measured using a tri-axial accelerometer ActiGraph GT3X-BT instructed to use the accelerometer for 7 days, except while bathing and sleeping. LPA was defined as 100-1,952 counts per min. Hazard ratio of LPA for the progression of nephropathy was calculated by Cox hazard model after adjusting by age, gender, BMI, hemoglobin A1c (HbA1c), smoking states, alcohol consumption, use of renin-angiotensin system inhibitor at baseline. Results: 167 patients (109 men) including 132 with stage1 and 35 with stage 2, mean age, body mass index, HbA1c, UACR, LPA and median observation period were 72.1 ± 12.3 years, 23.4 ± 3.4 kg/m2, 7.6 ± 0.9 %, 56.8 ± 137.4 mg/gCr, 305.6 ± 85.5 minutes/day and 1246 (406-1283) days. During follow-up, progression of nephropathy was observed in 39.5 % (66/167). The adjusted hazard ratio of LPA per 1 hour of progression of nephropathy was 0.81 (95% confidence interval 0.67-0.98, P = 0.031). Conclusion: Patients with T2DM with low LPA were shown to be at higher risk of developing diabetic nephropathy. This suggests that increasing activities of daily living in patients with T2DM is important in terms of preventing diabetic nephropathy. Disclosure T.Osaka: None. H.Okada: None. M.Fukui: None.

Spatial distribution of electric field measurement system: straight characterization of one or more original electric field sources in association with a new triaxial sensor

Several original point charge like electric field sources and a new tri-axial capacitive based electric field sensor have been designed to be associated together to build a new instrument able to characterize the electrical properties of the environment. The paper presents two fundamental features of the target electric imaging instrument: the characterization of the constituent parts of the system and the validation of the low-level data processing functions. The original electric field sources are composed of several conductive spheres with conical support so that they can be assimilated to point charges. The various arrangements of the sources can generate monopolar or multipolar electric fields. The tri-axial sensor is is integrated in a 2D scanning setup and used to map simultaneously the three components of the electric field generated by original electric field sources composed of several conductive spheres. The measured electric field images are compared to images calculated with an analytical model using the method of image charges. The agreement between measured electric field images and model validates both the original design of the sources and the geometrical features of the sensor which opens the door to application such as the detection of non-conductive buried object and smart detection with autoconfiguration.

Dual-Link Synchronous Acquisition and Transmission System for Cabled Seafloor Earthquake Observatory

Seafloor observatories play a crucial role in acquiring continuous and precise submarine monitoring data, thereby holding significant implications for advancing major scientific advancements in marine science, particularly in the field of seafloor earthquake observation. This work mainly builds a dual-link observation system designed for observing seismic information on the seafloor based on a Zynq7000 system-on-chip and time synchronization module. The system is based on Zynq7000 SoC(MILIANKE; Changzhou, China) and AD7768(Analog Devices, Inc.; Norwood, MA, USA) to achieve eight-channel data (24 bit) synchronous acquisition, and the robustness of the system is improved by upgrading the link to full-duplex transmission and adding node data self-storage function. The P88 1588 PTP time synchronization single board(CoolShark; Beijing, China) is used to provide PPS (Pulse per second) signals for the system to realize microsecond timestamps to support subsequent seismic data inversion. An upper computer system based on the Qt framework is also developed to monitor the network condition in real time while visualizing the data transmission. For the acquisition of seismic signals, we employed triaxial seismic sensors. Additionally, a temperature and humidity monitoring module, along with an attitude detection module, was designed to enable real-time monitoring of the nodes. These modules not only facilitate the real-time monitoring of the nodes but also contribute to seismic data inversion. The experimental results indicate that the system provides a good synchronization of data acquisition, high accuracy, and reliability of inter-node transmission, which has good application prospects.

Force Sensing and Feedback System Based on Novel Triaxial Force Capacitive Sensor for Minimally Invasive Surgical Robot

AbstractMinimally invasive surgery has attracted great attention due to small trauma, light pain, and quick recovery. Currently, most minimally invasive surgical robots lack the ability to force sense, resulting in high risks. Herein, a novel minimally invasive surgical force sensing and feedback system for minimally invasive surgical robot is proposed based on a flexible triaxial force capacitive sensor array. The capacitive force sensors utilize a microstructure electrode and orthogonal triangular pyramid microstructure to tackle the trade‐off between high sensitivity and wide‐detection range, showing 0–3 N detection range for normal force and high sensitivity of 69.19% N−1. Furthermore, the triaxial capacitive force sensors are integrated into the end‐effector of the minimally invasive surgical robot to form the minimally invasive surgical force sensing system. The system can show real‐time position of gripping or touching action, and the magnitude of triaxial force on the display surface. Importantly, the sensing force can further control the movement of the clamp, thus forming a novel force sensing and feedback system. The force sensing and feedback system of this minimally invasive surgical robot lays the foundation for its application in minimally invasive surgery andis expected to improve the safety and success of robotic‐assisted minimally invasive surgery.

Design and simulation of micro piezoelectric fiber triaxial acceleration sensor

Aiming at the demand for passive sonar systems for medium and low-frequency detection, low cost, and small working platform working sensor, a three-axis pressure piezoelectric accelerometer based on piezoelectric ceramic fiber is designed. Firstly, based on the acoustic vibration pick-up principle and extended Hamilton, the action mode of the accelerometer is analyzed theoretically to determine the accelerometer structure size. The accelerometer with a special mass structure is designed by using axially polarized piezoelectric ceramic fiber. The characteristic frequency and frequency domain analysis of the accelerometer are simulated by the FEM method. The results show that the overall size of the piezoelectric accelerometer is 20 mm ×20 mm×35 mm, the first-order natural frequency is 410 Hz, the external field directivity is standard “8” type, and the acceleration sensitivity reaches 112.2 mV/g, which verifies the great potential of the three-axis accelerometer designed by piezoelectric ceramic fiber material in the field of medium and low frequency.

Skill Level Classification in Basketball Free-Throws Using a Single Inertial Sensor

Wearable sensors are an emerging technology, with growing evidence supporting their application in sport performance enhancement. This study utilized data collected from a tri-axial inertial sensor on the wrist of ten recreational and eight professional basketball players while they performed free-throws, to classify their skill levels. We employed a fully connected convolutional neural network (CNN) for the classification task, using 64% of the data for training, 16% for validation, and the remaining 20% for testing the model’s performance. In the case of considering a single parameter from the inertial sensor, the most accurate individual components were upward acceleration (AX), with an accuracy of 82% (sensitivity = 0.79; specificity = 0.84), forward acceleration (AZ), with an accuracy of 80% (sensitivity = 0.78; specificity = 0.83), and wrist angular velocity in the sagittal plane (GY), with an accuracy of 77% (sensitivity = 0.73; specificity = 0.79). The highest accuracy of the classification was achieved when these CNN inputs utilized a stack-up matrix of these three axes, resulting in an accuracy of 88% (sensitivity = 0.87, specificity = 0.90). Applying the CNN to data from a single wearable sensor successfully classified basketball players as recreational or professional with an accuracy of up to 88%. This study represents a step towards the development of a biofeedback device to improve free-throw shooting technique.

Ferrite Shield to Enhance the Performance of Optically Pumped Magnetometers for Fetal Magnetocardiography.

Fetal magnetocardiography (fMCG) has proven to be an important tool for the prenatal monitoring of electrical cardiac activity; however, the high cost of superconducting quantum instrumentation (SQUID) poses a limitation for the dissemination of fMCG as a routine clinical technique. Recently, optically pumped magnetometers (OPMs) operating within person-sized, cylindrical shields have made fMCG more practical, but environmental magnetic interference entering through the shield opening substantially degrades the quality of fMCG signals. The goal of this study was to further attenuate these interferences by placing the OPM array within a small ferrite shield. FMCG recordings were made with and without the ferrite shield in ten subjects inside a person-sized, three-layer mu-metal cylindrical shield. Although the fetal signal was slightly attenuated, the environmental interference was reduced substantially, and maternal interference was also diminished. This increased the signal-to-noise ratio significantly and improved the resolution of the smaller waveform components. The performance improvement was highest in the axial direction and compensated for a major weakness of open-ended, person-sized shields. The ferrite shield is especially beneficial for the deployment of triaxial OPM sensors, which require effective shielding in all directions.

FPGA-Flux Proprietary System for Online Detection of Outer Race Faults in Bearings

Online fault detection in industrial machinery, such as induction motors or their components (e.g., bearings), continues to be a priority. Most commercial equipment provides general measurements and not a diagnosis. On the other hand, commonly, research works that focus on fault detection are tested offline or over processors that do not comply with an online diagnosis. In this sense, the present work proposes a system based on a proprietary field programmable gate array (FPGA) platform with several developed intellectual property cores (IPcores) and tools. The FPGA platform together with a stray magnetic flux sensor are used for the online detection of faults in the outer race of bearings in induction motors. The integrated parts comprising the monitoring system are the stray magnetic flux triaxial sensor, several developed IPcores, an embedded processor for data processing, and a user interface where the diagnosis is visualized. The system performs the fault diagnosis through a statistical analysis as follows: First, a triaxial sensor measures the stray magnetic flux in the motor’s surroundings (this flux will vary as symptoms of the fault). Second, an embedded processor in an FPGA-based proprietary board drives the developed IPcores in calculating the statistical features. Third, a set of ranges is defined for the statistical features values, and it is used to indicate the condition of the bearing in the motor. Therefore, if the value of a statistical feature belongs to a specific range, the system will return a diagnosis of whether a fault is present and, if so, the severity of the damage in the outer race. The results demonstrate that the values of the root mean square (RMS) and kurtosis, extracted from the stray magnetic field from the motor, provide a reliable diagnostic of the analyzed bearing. The results are provided online and displayed for the user through interfaces developed on the FPGA platform, such as in a liquid crystal display or through serial communication by a Bluetooth module. The platform is based on an FPGA XC6SLX45 Spartan 6 of Xilinx, and the architecture of the modules used are described through hardware description language. This system aims to be an online tool that can help users of induction motors in maintenance tasks and for the early detection of faults related to bearings.

Modeling of Sand Triaxial Specimens under Compression: Introducing an Elasto-Plastic Finite Element Model to Capture the Impact of Specimens’ Heterogeneity

This paper follows up on a series of reference papers that inspired MDPI’s Topic “Stochastic Geomechanics: From Experimentation to Forward Modeling”, where global and local deformation effects on sand specimens are fully described from high-resolution boundary displacement fields, as supported by a comprehensive experimental database (which includes varying degrees of specimen’s heterogeneity) that is available to the scientific community for further study. This paper presents an elasto-plastic comparative analysis of different finite element models reproducing different sand specimen heterogeneity configurations as follows: loose, dense, and half-dense half-loose specimens. The experimental conditions for these specimens’ heterogeneity configurations were simulated with an axisymmetric finite element model. To characterize the stress-strain response obtained from the experiments, an elasto-plastic constitutive model with strain-hardening and softening laws was adopted to reproduce the sand specimens’ mechanistic response. An expert-based calibration of the numerical models accounted for both global and local effects by making use of global observations captured by the triaxial point sensors (i.e., axial force and displacement) and by local observations captured by 3D digital image correlation analysis (i.e., 3D boundary displacement fields). Results show that predictions of the proposed numerical models are in good agreement with the experimental observations, both global and local responses. The combined use of global and local observations to calibrate sand triaxial specimens sets the basis for a more comprehensive parameterization process. For the first model set, three experiments were assumed with homogeneous materials. While both dense and loose models showed good agreement with the experiments, the displacement field prediction of the half-dense half-loose layered model identified limitations in reproducing heterogeneous configurations. Afterward, the second set compared and analyzed the half-dense half-loose layered models by implementing a heterogeneous model, showing significantly better model predictions (i.e., after the implementation of the heterogeneous model, which accounts for a transition zone between the upper and lower segments).

Year-round behavioural time budgets of common woodpigeons inferred from acceleration data using machine learning

Accelerometers capture rapid changes in animal motion, and the analysis of large quantities of such data using machine learning algorithms enables the inference of broad animal behaviour categories such as foraging, flying, and resting over long periods of time. We deployed GPS-GSM/GPRS trackers with tri-axial acceleration sensors on common woodpigeons (Columba palumbus) from Hesse, Germany (forest and urban birds) and from Lisbon, Portugal (urban park). We used three machine learning algorithms, Random Forest, Support Vector Machine, and Extreme Gradient Boosting, to classify the main behaviours of the birds, namely foraging, flying, and resting and calculated time budgets over the breeding and winter season. Woodpigeon time budgets varied between seasons, with more foraging time during the breeding season than in winter. Also, woodpigeons from different sites showed differences in the time invested in foraging. The proportion of time woodpigeons spent foraging was lowest in the forest habitat from Hesse, higher in the urban habitat of Hesse, and highest in the urban park in Lisbon. The time budgets we recorded contrast to previous findings in woodpigeons and reaffirm the importance of considering different populations to fully understand the behaviour and adaptation of a particular species to a particular environment. Furthermore, the differences in the time budgets of Woodpigeons from this study and previous ones might be related to environmental change and merit further attention and the future investigation of energy budgets.Significance statementIn this study we took advantage of accelerometer technology and machine learning methods to investigate year-round behavioural time budgets of wild common woodpigeons (Columba palumbus). Our analysis focuses on identifying coarse-scale behaviours (foraging, flying, resting) using various machine learning algorithms. Woodpigeon time budgets varied between seasons and among sites. Particularly interesting is the result showing that urban woodpigeons spend more time foraging than forest conspecifics. Our study opens an opportunity to further investigate and understand how a successful bird species such as the woodpigeon copes with increasing environmental change and urbanisation. The increase in the proportion of time devoted to foraging might be one of the behavioural mechanisms involved but opens questions about the costs associated to such increase in terms of other important behaviours.

Leg Dominance-Surface Stability Interaction: Effects on Postural Control Assessed by Smartphone-Based Accelerometry.

The preferential use of one leg over another in performing lower-limb motor tasks (i.e., leg dominance) is considered to be one of the internal risk factors for sports-related lower-limb injuries. The current study aimed to investigate the effects of leg dominance on postural control during unipedal balancing on three different support surfaces with increasing levels of instability: a firm surface, a foam pad, and a multiaxial balance board. In addition, the interaction effect between leg dominance and surface stability was also tested. To this end, a tri-axial accelerometer-based smartphone sensor was placed over the lumbar spine (L5) of 22 young adults (21.5 ± 0.6 years) to record postural accelerations. Sample entropy (SampEn) was applied to acceleration data as a measure of postural sway regularity (i.e., postural control complexity). The results show that leg dominance (p < 0.001) and interaction (p < 0.001) effects emerge in all acceleration directions. Specifically, balancing on the dominant (kicking) leg shows more irregular postural acceleration fluctuations (high SampEn), reflecting a higher postural control efficiency or automaticity than balancing on the non-dominant leg. However, the interaction effects suggest that unipedal balancing training on unstable surfaces is recommended to reduce interlimb differences in neuromuscular control for injury prevention and rehabilitation.

Machine Learning on Prediction of Relative Physical Activity Intensity Using Medical Radar Sensor and 3D Accelerometer.

One of the most critical topics in sports safety today is the reduction in injury risks through controlled fatigue using non-invasive athlete monitoring. Due to the risk of injuries, it is prohibited to use accelerometer-based smart trackers, activity measurement bracelets, and smart watches for recording health parameters during performance sports activities. This study analyzes the synergy feasibility of medical radar sensors and tri-axial acceleration sensor data to predict physical activity key performance indexes in performance sports by using machine learning (ML). The novelty of this method is that it uses a 24 GHz Doppler radar sensor to detect vital signs such as the heartbeat and breathing without touching the person and to predict the intensity of physical activity, combined with the acceleration data from 3D accelerometers. This study is based on the data collected from professional athletes and freely available datasets created for research purposes. A combination of sensor data management was used: a medical radar sensor with no-contact remote sensing to measure the heart rate (HR) and 3D acceleration to measure the velocity of the activity. Various advanced ML methods and models were employed on the top of sensors to analyze the vital parameters and predict the health activity key performance indexes. three-axial acceleration, heart rate data, age, as well as activity level variances. The ML models recognized the physical activity intensity and estimated the energy expenditure on a realistic level. Leave-one-out (LOO) cross-validation (CV), as well as out-of-sample testing (OST) methods, have been used to evaluate the level of accuracy in activity intensity prediction. The energy expenditure prediction with three-axial accelerometer sensors by using linear regression provided 97-99% accuracy on selected sports (cycling, running, and soccer). The ML-based RPE results using medical radar sensors on a time-series heart rate (HR) dataset varied between 90 and 96% accuracy. The expected level of accuracy was examined with different models. The average accuracy for all the models (RPE and METs) and setups was higher than 90%. The ML models that classify the rating of the perceived exertion and the metabolic equivalent of tasks perform consistently.

Intelligent Cubic-Designed Piezoelectric Node (iCUPE) with Simultaneous Sensing and Energy Harvesting Ability toward Self-Sustained Artificial Intelligence of Things (AIoT)

The evolution of artificial intelligence of things (AIoT) drastically facilitates the development of a smart city via comprehensive perception and seamless communication. As a foundation, various AIoT nodes are experiencing low integration and poor sustainability issues. Herein, a cubic-designed intelligent piezoelectric AIoT node iCUPE is presented, which integrates a high-performance energy harvesting and self-powered sensing module via a micromachined lead zirconate titanate (PZT) thick-film-based high-frequency (HF)-piezoelectric generator (PEG) and poly(vinylidene fluoride-co-trifluoroethylene) (P(VDF-TrFE)) nanofiber thin-film-based low-frequency (LF)-PEGs, respectively. The LF-PEG and HF-PEG with specific frequency up-conversion (FUC) mechanism ensures continuous power supply over a wide range of 10-46 Hz, with a record high power density of 17 mW/cm3 at 1 g acceleration. The cubic design allows for orthogonal placement of the three FUC-PEGs to ensure a wide range of response to vibrational energy sources from different directions. The self-powered triaxial piezoelectric sensor (TPS) combined with machine learning (ML) assisted three orthogonal piezoelectric sensing units by using three LF-PEGs to achieve high-precision multifunctional vibration recognition with resolutions of 0.01 g, 0.01 Hz, and 2° for acceleration, frequency, and tilting angle, respectively, providing a high recognition accuracy of 98%-100%. This work proves the feasibility of developing a ML-based intelligent sensor for accelerometer and gyroscope functions at resonant frequencies. The proposed sustainable iCUPE is highly scalable to explore multifunctional sensing and energy harvesting capabilities under diverse environments, which is essential for AIoT implementation.

Measurement of Frontal Midline Theta Oscillations using OPM-MEG

Optically pumped magnetometers (OPMs) are an emerging lightweight and compact sensor that can measure magnetic fields generated by the human brain. OPMs enable construction of wearable magnetoencephalography (MEG) systems, which offer advantages over conventional instrumentation. However, when trying to measure signals at low frequency, higher levels of inherent sensor noise, magnetic interference and movement artefact introduce a significant challenge. Accurate characterisation of low frequency brain signals is important for neuroscientific, clinical, and paediatric MEG applications and consequently, demonstrating the viability of OPMs in this area is critical. Here, we undertake measurement of theta band (4–8 Hz) neural oscillations and contrast a newly developed 174 channel triaxial wearable OPM-MEG system with conventional (cryogenic-MEG) instrumentation. Our results show that visual steady state responses at 4 Hz, 6 Hz and 8 Hz can be recorded using OPM-MEG with a signal-to-noise ratio (SNR) that is not significantly different to conventional MEG. Moreover, we measure frontal midline theta oscillations during a 2-back working memory task, again demonstrating comparable SNR for both systems. We show that individual differences in both the amplitude and spatial signature of induced frontal-midline theta responses are maintained across systems. Finally, we show that our OPM-MEG results could not have been achieved without a triaxial sensor array, or the use of postprocessing techniques. Our results demonstrate the viability of OPMs for characterising theta oscillations and add weight to the argument that OPMs can replace cryogenic sensors as the fundamental building block of MEG systems.

Development of a Low-Profile Planar Sensor for the Detection of Normal and Shear Forces

Individuals with balance and mobility problems might benefit by the use of devices that detect small changes in ground reaction forces and potentially be used to assist movement. For maximum effectiveness, such sensors must measure pressure in all three dimensions. Impact and shear plantar force are essential variables in inverse dynamics reconstructions of the human joint force. Various force sensors have been proposed to monitor plantar forces of the human foot. Most of them have a single-axis measurement, and few are intended for monitoring normal and shear stress. This article proposes a low-cost, biocompatible triaxial piezoresistive sensor developed using simple fabrication techniques and inexpensive machinery. The sensor can detect pressures from 0-800kPa with high response and recovery with minimum hysteresis and repeatable results of over than 100 cycles.

A New Force‐Decoupling Triaxial Tactile Sensor Based on Elastic Microcones for Accurately Grasping Feedback

Tactile sensor capable of detecting both normal and tangential forces is essential for the next‐generation robotic electronics. However, designing a reliable triaxial tactile sensor to decouple the normal and tangential forces is still a challenge for current research. Herein, a capacitive triaxial force sensor based on a microcone dielectric layer together with the force decoupling method is proposed. The lateral and vertical displacement of the top electrode relative to the bottom electrodes under the action of external forces can be perfectly detected according to the capacitance variations. The microcone‐structured sensor combined with the new decoupling method leads to a high sensitivity of 3.5 kPa−1 for normal force in a low‐pressure range of 0––50 Pa and 0.134 N−1 for tangential force in 0–0.5 N, and a fast response time of 26 ms. The sensors mounted on the fingers of a robotic arm can provide effective tactile information during the grasping task. Finally, a conceptual 3 × 3 tactile sensor array is fabricated and verified for spatial pressure mapping. This work provides new insight into piezocapacitive tactile sensor design and shows its great potential in robotic applications.

Iterative Robust Ellipsoid Fitting Based on M-Estimator With Geometry Radius Constraint

The ellipsoid fitting technique has been popularly used in many application scenarios, especially in triaxial sensor calibration in recent years. The conventional method to achieve ellipsoid fitting is the pseudo-inverse technique with algebraic constraints to obtain proper fitting results. This technique ensures the least mean square (LMS) error relative to all measured data and prevents trivial solutions. However, this approach lacks robustness against outliers resulting in nonideal fitting performance in real calibration situations. The M-estimator technique with several robust cost functions is adopted and applied in this research to attenuate the outlier’s effect to offer a more accurate fitting performance. In contrast, the constraint with geometric meaning is also applied and examined to prevent trivial solutions. These two methods are achieved by using the iterative constrained gradient descent (G.D.) method to approach the proper ellipsoid fitting. The performances are validated through synthesis and real measured data. Through calibrated experiments, the proposed methods are found to be able to increase robustness against outliers and prevent parameters of ellipsoid degeneration toward trivial solutions.

Investigation of a butterfly-arch stress-ribbon pedestrian bridge under ambient excitation: dynamic identification, FE modelling and parametric optimization

The dynamic behavior of a butterfly-arch stress-ribbon pedestrian bridge located in Fuzhou, Fujian, China was investigated under ambient excitation. Highly synchronous tri-axial wireless sensors were used to get an estimate of the model parameters, finding eight stable modes with bending and torsional deformations in the frequency range 3.59–14.92 Hz. Different finite element models of the bridge were developed and a parametric optimization process with a sensitivity-based algorithm was performed to achieve the best model which could be used as baseline for a long-term monitoring of the bridge.

Fabrication of Tri-Axial MEMS Tactile Sensor for Minimally Invasive Medical Application using Wafer-Level Packaging

Fabrication of Tri-Axial MEMS Tactile Sensor for Minimally Invasive Medical Application using Wafer-Level Packaging

Instillation checking using long short‐term memories for ophthalmology patients

AbstractIn this paper, we propose a method of checking the eye lotion instillation for ophthalmology patients. Our method first estimates tilt angles of an eye dropper bottle from acceleration values measured by a triaxial sensor attached to the bottle. It next prepares data each of which is equal to a sequence of standardized tilt values, as data to be presented to a discrimination model. It employs either a long short‐term memory (LSTM for short) or a bidirectional long short‐term memory (B_LSTM for short) to construct the model. Once we present the data to be checked to our model, it produces a certainty degree indicating whether a patient corresponding to the presented data applies eye lotion at the time zone in which a sequence of the tilt values used to prepare the presented data was measured. The final judgement for the instillation depends on thresholding of the certainty degree. It is established, from experimental results using practical data, that adopting B_LSTM‐based models is useful in improving metric values compared to adopting LSTM‐based models, and that our models can cope well with the situation specified by the comparatively long time interval of measuring tilt values while some metric values slightly degrade.