AN EVENT-TRIGGERED, ARTIFICIAL INTELLIGENCE-SUPPORTED MODULAR BIOMETRIC MONITORING AND EARLY WARNING SYSTEM AND METHOD THAT HOLISTICALLY MONITORS ANIMAL AND ENVIRONMENTAL HEALTH.

Abstract

The invention describes the collection of physiological and behavioral data obtained through biometric sensors such as heart rate (HR), heart rate variability (HRV), body temperature, activity and their derivatives, and environmental sensors such as temperature, humidity, air quality and their derivatives, placed on the animal, using an event-triggered architecture and transmitting this data wirelessly to a cloud server; the evaluation of this data by an artificial intelligence and analysis engine based on individual-specific dynamic reference ranges to predict the probability of hunger and thirst, calculate behavioral risk scores, and perform anomaly analyses synchronized with environmental data;This relates to a comprehensive health monitoring system and method that enables real-time, customized, and proactive monitoring of animal health through multidimensional data analysis, by interpreting the obtained analysis results via a decision and warning engine and communicating them to the user as notifications through mobile applications, web panels, or corporate interfaces.

Description

1 TARIFF A holistic, incident-triggered system that monitors animal and environmental health. EMPLOYEE, AI-POWERED MODULAR BIOMETRIC MONITORING AND EARLY WARNING SYSTEM AND METHOD Technical field to which the invention relates: 5 The invention involves a heart rate sensor (HR) implanted in an animal, measuring heart rate variability. The sensor (HRV) measures body temperature, activity and its derivatives, as well as biometric data such as temperature and humidity. physiological and physiological information obtained through environmental sensors, which are air quality and related products. Behavioral data is collected via an event-triggered architecture and transmitted wirelessly. the data is transmitted to the cloud server; this data is then processed by artificial intelligence and an analysis engine and presented to the individual in 10 minutes. The probability of hunger and thirst is assessed based on specific dynamic reference ranges. making predictions, calculating behavioral risk scores, and environmental performing anomaly analyses in sync with the data obtained; The analysis results are then interpreted via a decision and warning engine and displayed on mobile devices. The application will notify the user via web panel or corporate interfaces in the form of a notification on 15 enabling the transmission of real-time, customized animal health information through multidimensional data analysis. and a comprehensive health monitoring system and method that enables proactive follow-up. It is related. State of the art: In systems developed under the known state of the art, monitoring animal health is 20 For this purpose, mostly wearable sensors, collar or earring type devices, implants Telemetric units and environmental monitoring nodes are used. In the literature, heart rate, body temperature, activity, movement, rumination, nutrition, and behavior monitoring of physiological and behavioral parameters such as continuous or periodic monitoring It appears that numerous solutions exist for monitoring. Additionally, temperature and humidity are 25. and barn or living space conditions, separate sensors in terms of environmental well-being It is known that they are monitored by these systems. In the current state of the art, monitoring animal health involves multiple sensors and signals obtained from sensors are stored on the device or on the data management server. There are various systems that explain how this is analyzed. Among these systems are 30 2 Temperature, heart rate, acceleration, and similar parameters can be monitored via implantable wireless sensors. Solutions exist where parameters are measured and used for early warning purposes. In addition, multiple health parameters can be monitored through collar units developed for animals. IoT-based health and anger monitoring systems, where data is collected and transmitted to a server. It is known. Similarly, animal data collected from different sources constitutes a 5 There are also approaches based on analysis by comparison with the baseline. The literature also includes studies focusing on artificial intelligence and anomaly detection in animals. Behavioral classification, stress indicator extraction, herd management supporting decisions and automatically detecting unusual situations It appears to be developing in this direction. However, a significant portion of the current solutions are 10 It treats physiological monitoring and environmental monitoring independently; the data It is often collected using a continuous flow logic and individual dynamic reference. A holistic assessment based on these ranges remains limited. In the current state of the art, there are various recommendations for animal health monitoring systems and Although applications have been developed, these improvements are insufficient. For this purpose, 15 Some of the applications related to the developed inventions are given below. Application file number “WO2005104930A1” is in the known state of the art. It has been examined. The invention that is the subject of the application concerns the health status of animals and It is a system that allows its location to be monitored remotely. The system is implanted in the animal. It includes a wearable or portable tracking device, which is housed in a case containing 20 Through the sensors located there, the animal's predetermined physiological or environmental patterns are monitored. It detects their condition. Data obtained from the sensors is stored inside the enclosure. It is processed by a controller and transmitted wirelessly over a distance via a transmitter. It is transmitted to communication devices at that location. Application file number “TR2022013001A2”, which is in the known state of the art, 25 It has been examined. The invention in question concerns the health status of domestic animals. A health status analyzer is described for monitoring and analyzing data. The invention has the ability to connect with mobile devices and transmit data. a measuring device that has a microprocessor, a tracking module and a battery inside It includes 30 measured by a digital display located on the measuring device. The system allows for the display of data, as well as the pet's heart rate, respiration, 3 at least four sensors that detect physiological and behavioral data such as movement The information obtained through this method is transmitted to the device. Systems in the current state of the art are generally only GPS-based animal tracking systems. These are in the form of systems or wearable devices aimed at human health. Behavioral risk, hunger / thirst prediction, individual dynamic reference ranges and pre-disaster 5 It falls short in areas such as detecting behavioral anomalies. In conclusion, due to the negative aspects described above and the current solutions, the subject matter... Due to its shortcomings, an improvement is needed in the relevant technical field. It has been made. Purpose of the invention: 10 The main purpose of the invention is to utilize multiple sensor modules found on animals. by collecting biological and environmental data obtained through an event-triggered architecture The goal is to provide early diagnosis and anomaly detection through AI-powered analysis. Another aim of the invention is to analyze physiological and behavioral data obtained from animals. 15 Performing adaptive analysis by creating dynamic reference ranges specific to the individual. to provide. Another aim of the invention is to measure heart rate, heart rate variability, body temperature, and movement. By analyzing environmental sensor data together, hunger and thirst levels can be determined. The aim is to enable prediction based on probability. Another purpose of the invention is to analyze the movement, sound, and behavior data of animals. The aim is to calculate the behavioral risk score by performing this procedure. In this context... Behavioral data; increased or decreased mobility, restless wandering, repetitive vocalizations. exclusion, sudden tendency to flee, disruption of normal resting patterns, separation from the herd. or may include measurable patterns such as unusual immobility. Another aim of the invention is to obtain 25 from multiple animals found in the same region. By analyzing data in a synchronized manner, regional environmental anomalies can be detected. The aim is to ensure this. In this context, collective behavioral changes; simultaneous direction change, simultaneous increase in unrest, herd escape or clustering, collective 4 increased vocalization, generalized immobility, or simultaneous deviation from the normal activity rhythm These can include herd-based anomalies. Another aim of the invention is to identify risks and anomalies through an AI-based decision engine. The aim is to create an automatic warning mechanism in such situations. The structural and characteristic features and all the advantages of the invention are given in Figure 5 below. And thanks to the detailed explanation written with reference to these figures, it becomes clearer. This will be understood as such. Therefore, the evaluation should also include these forms and detailed explanations. This should be done taking that into consideration. Explanation of the figures: FIGURE -1; The subject of the invention is an event-triggered 10-phase system that comprehensively monitors animal and environmental health. employee, AI-powered modular biometric tracking and early warning system and It is a diagram illustrating the management structure. Explanation of the references in the figures: 1. Sensor module 2. Event-triggered data collection unit 15 3. Wireless data transmission module 4. Cloud server 5. Artificial intelligence and analysis engine 6. Decision and warning engine 7. Interface 20 100. Initiation and data collection. 110. User activation of the sensor module on the animal. 120. Measurement of data by sensor module and event-triggered data acquisition module. transmission of measured data 130. Data from the sensor module is collected by the event-triggered data acquisition module. gathering 200. The collected data is transmitted to the cloud server via a wireless data transmission module. transmission and pre-processing and filtering by the cloud server 210. Noise reduction, processing of missing data and measurement of collected sensor data. 5 making the errors 300. Creation of a dynamic reference profile. 400. Performing time series analysis using artificial intelligence and an analysis engine. 410. Performing dynamic reference range learning. 420. Creating a probability model for hunger and thirst 10 430. Calculation of behavioral risk score using artificial intelligence and analysis engine. 440. Detecting anomalies before a disaster. 500. Threshold comparison performed by the decision and warning engine. 510. Calculated values ​​can be derived from a dynamic reference profile or user input. dynamic reference ranges with predefined risk thresholds 15 comparison 520. The results are divided into two categories. 600. The user can view the analysis results in real time via the interface. display Description of the invention: 20 The invention relates to the measurement of HR, HRV, body temperature, activity, and behavior in animals. biometric sensors and environmental sensors measuring temperature, humidity, air quality and their derivatives. Physiological and behavioral data obtained through sensors are event-triggered. collected by architecture and transmitted to the cloud server via wireless connection (4); 6 Dynamic reference to the individual by artificial intelligence and analysis engine (5) of the data Estimates of the probability of hunger and thirst are made by evaluating these ranges. this involves calculating behavioral risk scores and analyzing environmental data. performing anomaly analyses in a synchronized manner; the resulting analysis The results are interpreted through the decision and warning engine (6) and the mobile application, 5 notification to the user via web panel or corporate interfaces (7) It relates to a comprehensive health monitoring system and method that enables its transmission. The invention relates to heart rate (HR) and heart rate variability (HRV) measurements taken on animals. body temperature, activity and behavior data obtained through environmental sensors It is a system that enables the collection and analysis of acquired data using an event-triggered architecture. 10 This structure allows for monitoring changes in the health status of animals, behavioral risks, and Detecting environmental threats at an early stage and intervening in a timely manner This makes it possible. The system also provides physiological and biological data from animals. By analyzing behavioral data, a dynamic, individual reference point is created for each animal. This creates ranges. Thus, a personalized health and behavior profile for each animal is created. 15 By removing these elements, adaptive analysis is performed, and risk assessments become more accurate and reliable. This can be done in this way. The invention encompasses heart rate, heart rate variability, body temperature, and movement. Hunger and thirst levels are determined by analyzing data from both environmental sensors and other sources. It can be predicted based on probability. In addition, animal movements and sounds 20 Behavioral risk scores are calculated by evaluating behavioral data, and stress, early detection of illness, aggression, or unusual behavioral changes This can be done by obtaining data from multiple animals in the same area. By analyzing them in a synchronized manner, regional environmental anomalies are also identified. They can be identified; thus, disasters such as earthquakes, fires, and gas leaks can be detected 25 days in advance. Early biological warning signs that may occur can be detected. In addition, the system can detect artificial intelligence. automatic decision-making in risk or anomaly situations through an intelligence-based decision engine By creating a warning mechanism, users are informed quickly and the necessary information is provided. It ensures that preventive or intervention actions are taken. The sensor module on the animal (1) monitors the animal's heart rate continuously or during events 30 A trigger-activated biosensor that measures in response to stress, illness, hunger, or abnormalities. The heart rate sensor (HR) is used for the early detection of physiological conditions; heart 7 analyzing the time differences between the shots and the animal's stress level, metabolic Heart rate variability provides information about status and autonomic nervous system activity. the HRV sensor, which measures the core or surface body temperature of the animal and detects fever, body image monitors used to track infection, environmental stress, and metabolic changes. temperature sensor; animal's direction of movement, acceleration and activity intensity 5 an activity and motion sensor that enables behavioral analysis by measuring; By detecting and analyzing the sounds or behavioral signals the animal makes, stress, sounds and behaviors that allow the detection of signs of restlessness or illness its sensor monitors the environmental conditions of the animal's surroundings (ambient temperature, humidity, It includes environmental sensors that measure air quality, gas density, etc. 10 Here, behavioral data includes increased or decreased mobility, and recurring restlessness. movements, sudden tendency to flee, disruption of normal resting patterns, separation from the herd, measurable behaviors such as increased frequency of vocalization, unusual immobility, etc. These represent patterns. Environmental sensors, on the other hand, measure ambient temperature, humidity, and air quality. 15 capable of measuring (CO₂, NH₃, VOC), gas density, pressure, light and noise levels. It is structured. Event Triggered Data Acquisition Unit (2) collects sensor data from sensor module (1) continuously monitored, only when certain threshold values ​​are exceeded, sudden changes occur, or Data collection process initiated when statistical deviations are detected. It is an addition module. 20 The wireless data transmission module (3) receives sensor data from the Event Triggered data acquisition unit (2). It is the communication module that transmits data to the cloud server (4). Data transmission; short distance and low energy consumption Bluetooth Low Energy (BLE), long range and low energy Low-consumption LoRa or wide-area coverage provided via mobile operator infrastructure. This can be achieved using at least one of the NB-IoT technologies. 25 The cloud server (4) is the server where sensor data is stored, processed and analyzed. Data is pre-processed on this server; noise reduction, missing data... completion, data validation, and measurement error correction processes is being carried out. Artificial intelligence and analysis engine (5), HR, HRV, body 30 obtained from sensor module. Time series of temperature, activity / motion, sound / behavior, and environmental sensor data. 8 processing data; creating dynamic reference ranges based on this data; fasting and calculates the thirst probability model, generates a behavioral risk score, and multiple By evaluating animal data in a synchronized manner, it detects anomalies before a disaster occurs. It is an analysis engine. This engine uses statistical methods, rule-based decision structures, 5 capable of working with supervised or unsupervised machine learning algorithms It can be configured. Artificial intelligence and analysis engine (5), from animal, user animal species, age, usage scenario can be selected by the system or by the operator. mobile data that can be automatically determined based on data continuity and historical data volume. learning window with last 24 hours, last 3 days, last 7 days, last 30 days or application According to the scenario, the time obtained in a specific time period, which can include different durations, is 10. Statistical calculations such as mean, variance, and standard deviation are performed using the series data. methods and machine learning algorithms are constantly updated for each animal enabling the acquisition of a specific adaptive model and dynamic reference ranges. It is the engine that determines this. In addition, heart rate, HRV, activity and movement data are taken together. a multivariate scale created by analyzing specific data combinations 15 hunger, which allows for the classification of hunger according to the likelihood of hunger and thirst. and normalizes the data from the sensors to create the thirst probability model. by analyzing, weighting, and generating a total score, the risk score is calculated and Time synchronization of data from multiple animals in the same region By conducting and analyzing these studies, anomaly detection before a disaster occurs. 20 The behavioral risk score is a dynamic reference of physiological and behavioral data belonging to the animal. It is a risk indicator that quantitatively expresses the degree of deviation from the specified ranges. This score is calculated as HR. deviation, HRV change, activity deviation, voice / behavioral deviation, and optional normalization and weighting of parameters such as environmental impact factor It can be calculated as follows. In an example application, Risk Score = (HR deviation × 25) (w1) + (HRV change × w2) + (activity deviation × w3) + (voice / behavioral deviation × w4) The formula + (environmental impact factor × w5) can be used. However, the invention is not based on this. but not limited to rule-based scoring, statistical indexing, or machine learning. Risk modeling based on fundamental principles can also be used. Pre-disaster anomaly detection, same 30 timestamped sensor data from multiple animals in the region This is achieved through simultaneous analysis. In this context, the data are first analyzed jointly. aligning with the time axis, establishing a regional baseline, and then collectively Behavioral changes are being investigated. Collective behavioral changes involve simultaneous direction. change, increased mass unrest, herd escape, unusual clustering, widespread 9 inactivity, simultaneous deviation from the normal activity rhythm, and increased collective vocalizations, etc. This refers to swarm-based patterns. The identified deviations are obtained from environmental sensors. sudden temperature increase, deterioration of air quality, increase in toxic gas concentration, with environmental threat data such as pressure changes, fire or gas leak indicators By synchronizing, a pre-disaster anomaly score is created. Hunger and thirst 5 Probability model, animal heart rate, HRV, body temperature, activity / movement, Feature extraction via voice / behavior and optional environmental sensor data. It is created by performing this process. Within this scope, the obtained data are normalized, Short-term and long-term trends are compared, and specific data combinations are used. These data combinations are converted into feature vectors; for example, 10 Increased vocalization frequency with increased activity, restlessness with decreased HRV. Circulation, along with changes in body temperature, indirectly relates to water consumption behavior. two such as movement patterns or physiological deviations associated with prolonged immobility or it is based on the principle of evaluating more than one parameter together. The obtained Feature vectors, threshold-based classifier, probabilistic model, decision tree, logistic 15 Hunger and thirst detected using regression or similar machine learning-based classifiers. It is converted into a probability score. Decision and warning engine (6), risk scores from artificial intelligence and analysis engine (5), Dynamic reference profile or user with probability estimates and anomaly outputs. comparing the threshold values ​​defined by; as a result of this comparison, health 20 warning, behavioral risk warning, environmental risk warning or early warning before a disaster It is the module that generates and transmits the warnings to the user via the interface (7). The interface (7) allows the user to view the animal's health and behavior data in real time. mobile application, web that can view, receive alerts and respond quickly It is a panel or institutional structure. 25 An event-triggered, artificial intelligence system that holistically monitors animal and environmental health. Supported modular biometric monitoring and early warning management involves the following steps: includes; - Initiation and data collection (100),  The sensor module on the animal must be activated by the user. 30 (110),  Data measurement by the sensor module and event-triggered data acquisition module. Transmission of measured data (120), That data:  Heart rate (HR)  Heart rate variability (HRV) 5  Body temperature  Activity / movement  Voice / behavior  Environmental sensors (temperature, humidity, air quality)  The event-triggered data acquisition module collects data from the sensor module 10 collection (130). - The collected data is transmitted to the cloud server via a wireless data transmission module. and pre-processing and filtering by the cloud server (200),  Noise reduction, processing of missing data, and measurement of collected sensor data. making the errors correct (210). 15 - Creation of a dynamic reference profile (300), Dynamic reference profile, time obtained from the animal's sensor module. stamped physiological, behavioral and environmental data within a specific learning window individual reference created through analysis throughout the process and continuously updated This is a dataset. The learning window in question consists of 20 user-defined elements. or animal species, age, use case and data determined by the system Automatically selectable based on duration: last 24 hours, last 3 days, last 7 days or It can include at least one of the last 30-day moving averages. This Within this scope, data is pre-processed to remove noise and missing data. 25 from completion, measurement verification and outlier suppression processes These are then processed; followed by the mean, median, variance, standard deviation, trend, and By extracting seasonal pattern parameters, the normal behavior specific to each animal can be determined. and physiological ranges are determined. - Performing time series analysis with artificial intelligence and analysis engine (400), 11 o Learning of the dynamic reference range (410),  Time series obtained from an animal over a specific period of time using the data, such as mean, variance and standard deviation statistical methods and machine learning algorithms using automated normal behavior and physiological ranges 5 Dynamic reference range learning is performed through this learning process. Normal ranges are those observed during the learning period of the animal in question. Normal heart rate band, HRV range of variation, body temperature band, typical daily activity rhythm, rest period, vocalization 10 key factors specific to the individual, such as intensity and environmental exposure levels It refers to behavioral and physiological patterns.  Time series data; periodic or event-triggered. Heart rate recorded in this format and associated with a timestamp, HRV measures body temperature, activity intensity, acceleration, and change of direction. duration of inactivity, frequency of vocalization, ambient temperature, humidity, CO₂, 15 These are measurement sequences such as NH₃, VOC, pressure, light, and noise level. o Creation of a probability model of hunger and thirst (420),  HR, HRV, activity and movement, body temperature, voice and behavior, and Feature extraction from optional environmental parameters to be done, 20  Sensor data from sensors with a timestamp recording of heart rate, HRV, body temperature, activity, movement, sound / behavior and environmental sensor data alignment to a common time axis,  Noise reduction, missing data completion, erroneous measurement 25 Data preprocessing with extraction and outlier suppression. to be done  Short-term and long-term change for each data type calculating trends,  Average value for heart rate, sudden rise / fall, 30 subtracting the deviation and rate of change from the reference range, 12  Level of variability, decrease or increase for HRV data. Calculation of tendency and stress-related deviations,  A sudden deviation from the normal temperature band for body temperature. temperature increase / decrease and related to ambient temperature Identifying changes, 5  Activity and movement data include motion intensity, acceleration, change of direction, period of immobility, restless wandering, sudden escape tendency or disruption of normal resting patterns removal of existing behavioral characteristics,  Frequency of vocalization, vocalization 10 from voice and behavior data intensity, unusual increase in voice volume, or behavioral changes Identifying patterns of restlessness,  Environmental sensors provide ambient temperature, humidity, CO₂, NH₃, VOCs, pressure, light, noise, and air quality changes removal, 15  The animal's characteristics are calculated, determined, and extracted. Creating the feature vector,  The generated feature vector represents the probability of hunger / thirst. for behavioral risk scoring or environmental anomaly detection Transferring it to artificial intelligence and analysis engine. 20  Normalizing the extracted features,  The time of sensor data obtained from sensors matching with the stamp,  Creating a dynamic reference profile for each animal,  Instantaneous sensor values ​​are based on the animal's own dynamic 25 Comparison with reference values,  Deviation from the reference value for each parameter calculation,  Converting the calculated deviations to a common scale, 13  In this transformation, z-score, min-max scaling, Durable based on median and interquartile range Scaling or risk score between 0-1 / 0-100 At least one of the scaling methods can be used.  Behavioral analysis is determined by weighting the normalized values. risk score, hunger / thirst probability or anomaly score calculation.  Converting data combinations into feature vectors,  Data combinations; for example, high activity, increased vocalization, low HRV, low activity, high body 10 temperature, prolonged inactivity, increase in ambient temperature, access to water along with a high respiratory stress indicator such as a decrease in two or more sensor parameters It refers to simultaneous or sequential patterns.  Hunger detection with threshold-based or machine learning-based classifiers 15 and the steps involved in converting it into a possibility of thirst It includes. behavioral risk score with artificial intelligence and analysis engine calculation (430),  Sensor data (HR, HRV, motion, sound) together 20 by evaluating and weighting each parameter weighted linear modeling through mathematical model creation combination (weighted sum: Risk Score=∑ (𝑥 ⋅ 𝑤 )) model Risk Score = weighted sum of parameters It is calculated using the (weighted sum model) formula. 25 o Detection of anomalies before disaster (440).  Synchronizes timestamped data from multiple animals by establishing a regional baseline,  Environmental factors through the identification of collective behavioral changes Early identification of threats. 30 14  Collective behavioral changes; more than one in the same area Simultaneous increase or decrease in activity in the animal, direction movement in unison, clustering, breaking away from the herd, mass escape, These are unusual vocalizations and widespread immobility patterns. Environmental threats include biological anomalies before the earthquake. indicators include fire, gas leak, excessive temperature, and sudden weather events. quality of life deterioration, exposure to toxic gases, and similar factors. These are the risks in the area. - Threshold comparison by decision and warning engine (500),  Calculated values ​​are based on a dynamic reference profile or user input. 10 dynamic reference ranges with predefined risk thresholds comparison (510),  The results were divided into two categories (520). 1. Normal – requires no warning and is recorded.  By comparing real-time data with dynamic reference ranges 15 It is normal as long as the specified threshold value is not exceeded. 2. Risk / Anomaly – requires notification to the user.  By comparing real-time data with dynamic reference ranges Risk or anomaly if the determined threshold value is exceeded. in its detection: 20  Warning (Health warning, behavioral risk warning, environmental risk warning) (warning, early warning before disaster) production,  The generated alerts are displayed as notifications to the user via the interface. transmitted through. - The user can view the analysis results in real time via the interface 25 Viewing (600).

Claims

REQUESTS 1. An event-driven, artificial intelligence system that holistically monitors animal and environmental health. It is an intelligent-supported modular biometric tracking and early warning system, the feature of which is;  Sensors placed on the animal measure sensor data. module (1), 5  Monitoring sensor data from sensor module (1), predefined Exceeding threshold values, sudden changes, or statistical deviations. If detected, it initiates data collection and transmits the data wirelessly. Event Triggered data transmission module (3) transmits data to the cloud server (4) data collection unit (2), 10  The cloud receives sensor data from the Event Triggered data collection unit (2). Wireless data transmission module (3) transmitting to the server (4),  Sensor data received via the wireless data transmission module (3) cloud server where it is stored and pre-processed (4),  Using time series analysis and machine learning algorithms 15 Creating dynamic reference ranges to estimate the probability of hunger and thirst. doing, calculating behavioral risk scores and analyzing data from multiple animals It performs anomaly detection by synchronous analysis and analysis. an artificial intelligence and analysis engine that transmits its results to a decision and warning engine. (5), 20  Dynamic reference of the results of the artificial intelligence and analysis engine (5) by comparing them with other ranges, and Decision and warning system that transmits the warnings it produces to the user via interface (7) engine (6),  The user can view the analysis results in real time, 25 It includes the interface (7) through which it receives warnings.

2. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 1. employee, AI-powered modular biometric tracking and early warning system Its characteristic is; 16  Measures the animal's heart rate continuously or triggered by incidental triggers. biosensors that detect early physiological states such as stress, illness, and hunger. heart rate sensor used in detection,  Analyzes the time differences between heartbeats and heart rate Heart rate variability sensor, which provides information about variability, 5  Measures the core or surface body temperature of the animal and detects fever. monitoring metabolic changes that occur during infection and environmental stress body temperature sensor used,  Measures the animal's direction of movement, acceleration, and activity intensity. Activity and motion sensor, 10  Detecting and analyzing sounds or behavioral signals emitted by the animal voice and behavior sensors that detect behavioral states,  Environmental measurement, which measures the environmental conditions of the animal's environment. It includes a sensor module (1) containing sensors.

3. Event-triggered 15 comprehensive monitoring of animal and environmental health in accordance with Claim 2. employee, AI-powered modular biometric tracking and early warning system Its characteristics include air quality, measured by ambient temperature, humidity, CO₂, NH₃, and VOCs, as well as gas levels. environmental sensors including density, pressure, light and noise sensors It includes.

4. Event-triggered 20 comprehensive monitoring of animal and environmental health in accordance with Claim 2. employee, AI-powered modular biometric tracking and early warning system Its characteristic feature is environmental factors such as ambient temperature, humidity, air quality, and gas density. It includes the conditions.

5. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 1. employee, AI-powered modular biometric tracking and early warning system 25 Its feature is to send sensor data to the cloud server (4), Bluetooth, LoRa, NB-IoT use at least one low-energy communication technology, wireless It includes a data transmission module (3).

6. Comprehensive, event-triggered monitoring of animal and environmental health in accordance with Claim 1. employee, AI-powered modular biometric tracking and early warning system 30 Its feature is that it includes mobile, web panel and corporate-based interface (7). 17 7. An event-driven, artificial intelligence system that holistically monitors animal and environmental health. It is an intelligent-supported modular biometric tracking and early warning method, the feature of which is;  Making the start and collecting data (100),  The collected data is transmitted to the cloud via a wireless data transmission module.

5. Transmission to the server and preprocessing and filtering by the cloud server to be done (200),  Creation of dynamic reference profile (300),  Performing time series analysis with artificial intelligence and analysis engine (400),  Threshold comparison by the decision and warning engine (500),  The user can view the analysis results in real time via the interface. 10 Its display includes (600) processing steps.

8. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 7. employee, AI-powered modular biometric tracking and early warning method Its feature is; the initiation and data collection process step (100);  The sensor module on the animal must be activated by the user. 15 bringing (110),  Data measurement by the sensor module and event-triggered data collection Transmission of measured data to the module (120),  Event-triggered data acquisition module receives data from the sensor module The collection of data (130) includes the steps of the process. 20 9. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 8. employee, AI-powered modular biometric tracking and early warning method Its feature is; data measurement by the sensor module and event-triggered data. Transmission of measured data to collection module (120) mentioned in the process step data including heart rate, heart rate variability, body temperature, activity and 25 motion, sound and behavior, temperature, humidity and air quality environmental sensors It is about having data.

10. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 7. employee, AI-powered modular biometric tracking and early warning method Its feature is that the collected data is transmitted to the cloud via a wireless data transmission module. transmitted to the server and pre-processed and filtered by the cloud server. (200) steps of the process; 18  Noise reduction, processing of missing data, and processing of collected sensor data. Correction of measurement errors (210) process step It includes.

11. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 7. Employee, AI-powered modular biometric tracking and early warning method 5 Its feature is that it performs time series analysis using artificial intelligence and an analysis engine. (400) steps of the process;  Learning the dynamic reference range (410),  Creation of a probability model of hunger and thirst (420),  Calculating behavioral risk score using artificial intelligence and analysis engine 10 (430),  Steps for detecting anomalies before a disaster (440) It includes.

12. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 11. Employee, AI-powered modular biometric tracking and early warning method 15 its feature is; Dynamic reference range learning (410) process step;  From the animal within a time period predetermined by the user Statistical methods are used with the time series data obtained, and normal behavior and physiological 20 with machine learning algorithms Automatic learning of ranges and dynamic reference range learning It includes the steps involved in the process.

13. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 12. employee, AI-powered modular biometric tracking and early warning method Its characteristic is; a time series obtained from an animal over a specific period of time. 25 statistical methods and machine learning algorithms using data through automatic learning of normal behavior and physiological ranges The process step mentioned in performing dynamic reference range learning Statistical methods include mean, variance, and standard deviation.

14. Event-triggered 30 comprehensive animal and environmental health monitoring systems in accordance with Claim 12. employee, AI-powered modular biometric tracking and early warning method 19 Its characteristic is; a time series obtained from an animal over a specific period of time. statistical methods and machine learning algorithms using data through automatic learning of normal behavior and physiological ranges The process step mentioned in performing dynamic reference range learning Time series data, heartbeats 5 acquired and associated with timestamps. speed, HRV, body temperature, activity intensity, acceleration, change of direction, immobility duration, frequency of sounding, ambient temperature, humidity, CO₂, NH₃, VOCs, pressure, light and The measurement sequences have a noise level.

15. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 12. Employee, AI-powered modular biometric tracking and early warning method 10 Its characteristic is; a time series obtained from an animal over a specific period of time. statistical methods and machine learning algorithms using data through automatic learning of normal behavior and physiological ranges The process step mentioned in performing dynamic reference range learning normal behavioral and physiological ranges during the animal's learning period are 15 Observed normal heart rate band, HRV range of variation, body temperature. band, typical daily activity rhythm, rest period, vocal intensity and individual-specific basic behavioral and physiological changes based on environmental exposure levels It is the presence of patterns.

16. Event-triggered 20 comprehensive monitoring of animal and environmental health in accordance with Claim 12. employee, AI-powered modular biometric tracking and early warning method Its feature is the creation of a probability model of hunger and thirst (420) process step;  HR, HRV, activity and movement, body temperature, voice and behavior, and discretion. Feature extraction from related environmental parameters, 25  Normalizing the extracted features,  Converting data combinations into feature vectors,  Hunger and hunger can be detected using threshold-based or machine learning-based classifiers. It involves the steps involved in transforming the possibility of thirst into reality.

17. Event-triggered 30 comprehensive animal and environmental health monitoring systems in accordance with Claim 16. employee, AI-powered modular biometric tracking and early warning method Its features include; HR, HRV, activity and movement, body temperature, voice and behavior, and The process of extracting features from optional environmental parameters. step;  Recording sensor data from sensors with a timestamp. and heart rate, HRV, body temperature, activity, movement, voice / behavior and Alignment of environmental sensor data to a common time axis, 5  Noise reduction, missing data completion, erroneous measurement debugging, and anomaly detection. Data preprocessing with value suppression  Short-term and long-term trend changes for each data type calculation,  Average value for heart rate, sudden rise / fall, reference 10 subtracting the deviation from the range and the rate of change,  Variability level, decreasing or increasing trend, and stress for HRV data. Calculation of related deviations,  Deviation from the normal temperature range for body temperature, sudden temperature change. increase / decrease and determination of changes related to environmental temperature, 15  Activity and movement data include motion intensity, acceleration, and change of direction, period of immobility, restless wandering, sudden tendency to flee, or normal Removal of behavioral characteristics that disrupt rest patterns,  Frequency and intensity of vocalizations from voice and behavioral data, unusual vocal increase or behavioral restlessness patterns 20 determination,  Environmental sensors measure ambient temperature, humidity, CO₂, NH₃, VOCs, and pressure, Extraction of light, noise and air quality variations,  Animal attributes with calculated, determined, and extracted characteristics vector creation, 25  The generated feature vector includes hunger / thirst probability, behavioral characteristics. Artificial intelligence and analysis for risk scoring or environmental anomaly detection. It includes the steps involved in transferring the data to the engine.

18. Including holistic, event-triggered monitoring of animal and environmental health in accordance with Claim 16. employee, AI-powered modular biometric tracking and early warning method 30 and its characteristic is; the normalization process step of the extracted features;  Sensor data obtained from sensors with a timestamp matching, 21  Creating a dynamic reference profile for each animal,  Instantaneous sensor values ​​are based on the animal's own dynamic reference point. Comparison with their values,  Calculating the deviation from the reference value for each parameter,  Converting the calculated deviations to a common scale, 5  Behavioral risk score obtained by weighting the normalized values, calculation of hunger / thirst probability or anomaly score process It includes the steps.

19. Including event-triggered holistic monitoring of animal and environmental health in accordance with Claim 16. Employee, AI-powered modular biometric tracking and early warning method 10 Its characteristic is the process of converting data combinations into feature vectors. The data combinations mentioned in step [number] indicate high activity, increased vocalization, low HRV, low activity, high body temperature, prolonged inactivity, increased environmental temperature, high respiratory stress indicator 15 two or more sensor parameters that have reduced access to water This refers to the presence of simultaneous or sequential patterns.

20. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 12. employee, AI-powered modular biometric tracking and early warning method Its feature is that it calculates behavioral risk scores using artificial intelligence and an analysis engine. The risk score mentioned in step (430) of calculation is 20 The calculation is based on the formula Risk Score = weighted sum of parameters. It is calculated using...

21. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 12. employee, AI-powered modular biometric tracking and early warning method Its feature is; Detection of anomalies before disaster (440) process step; 25  By synchronizing timestamped data from multiple animals. Establishment of a regional baseline,  Identifying changes in collective behavior and environmental threats It includes early detection process steps.

22. Event-triggered 30 comprehensive monitoring of animal and environmental health in accordance with Claim 21. employee, AI-powered modular biometric tracking and early warning method 22 Its characteristic feature is the identification of collective behavioral changes and environmental factors. The collective behavior mentioned in the early threat identification process step. changes resulting in simultaneous increase in activity in multiple areas in the same region or decrease, movement in unison, clustering, breaking away from the herd, mass escape, This includes unusual vocalizations and widespread immobility patterns. 5 23. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 21. employee, AI-powered modular biometric tracking and early warning method Its characteristic feature is the identification of collective behavioral changes and environmental factors. In the early identification of threats process step, the environmental threats mentioned are: Pre-earthquake biological anomaly indicators include fire, gas leak, extreme temperature, 10 sudden deterioration in air quality, living space risks due to toxic gas exposure. It is the fact that.

24. Event-triggered holistic monitoring of animal and environmental health in accordance with Claim 7. employee, AI-powered modular biometric tracking and early warning method Its feature is that the threshold comparison is performed by the decision and warning engine. 15 (500) steps of the process;  The calculated values ​​are based on a dynamic reference profile or user-defined values. Dynamic reference with predefined risk thresholds as input. Comparison with intervals (510),  The division of results into two categories (520) includes the steps of the process. 20 25. Including holistic, event-triggered monitoring of animal and environmental health in accordance with Claim 24. employee, AI-powered modular biometric tracking and early warning method The feature is that the results are divided into two categories (520) of the process step;  Determined by comparing real-time data with dynamic reference ranges. Risk and anomaly detection, which is the situation where the threshold value is not exceeded. 25 If not, it is categorized as not requiring any warning and recording,  Determined by comparing real-time data with dynamic reference ranges. by detecting a risk or anomaly, which is the situation where the threshold value is exceeded. The generation of warnings and the notification of these warnings to the user are 30. The process involves transmitting the information through an interface, including the necessary steps. 23 26. Including event-triggered holistic monitoring of animal and environmental health in accordance with Claim 25. employee, AI-powered modular biometric tracking and early warning method Its characteristic feature is the comparison of real-time data with dynamic reference ranges. risk and anomaly detection, which is the situation where the determined threshold value is exceeded. The generation of warnings and the notification of these warnings to the user. 5 The warnings mentioned in the process of transmitting them via the interface, health warning, behavioral risk warning, environmental risk warning, early warning before disaster It is the fact that.