ES2655544B1 - Method and system for monitoring living things - Google Patents

Abstract

Method and system for monitoring living things # The present invention proposes a method and system for monitoring people or animals efficiently. For them, sensors are used in electronic devices of internet of things (1) that process and send the information optimally from the point of view of the energy consumption of the device to a central server (3) that analyzes the information coming from the devices (1) and information from external data sources, open or proprietary (5). From this information the server generates patterns of individual, collective and social behavior (6), generates notifications (7) when any of the patterns is not met and generates indicators (8) that facilitate the understanding of the information to the end user of the solution.

Description

5

10

fifteen

twenty

25

30

DESCRIPTION

Method and system for monitoring living beings TECHNICAL FIELD OF THE INVENTION

The present invention relates to a method and system for the monitoring and generation of behavior patterns applied to fields such as extensive livestock (or other animals) inhabiting unconfined spaces, as well as people in different fields of application. This is done using electronic devices, preferably internet devices of things (with limitation in terms of the transfer of information, for maximizing the life of the battery and with different sensors). These devices collect data (for long periods of time) send the information, directly (without the need for intermediate gateways), to a server in the cloud where different adaptive algorithms are executed, using the information from those devices and other External data sources generate behavioral patterns and notifications when those patterns are not met. The objective is on the one hand the generation of behavior patterns for the detection of behavioral anomalies based on the joint analysis of the information coming from multiple sensors; and on the other, to optimize the energy consumption in internet devices of things that work in bands with restrictions on sending data.

BACKGROUND OF THE INVENTION

Currently there are different methods, systems and devices that are responsible for monitoring living things (people, animals, plants) or objects. Among them, in relation to the methods, there is a high variability in their operation. These methods perform the information processing as follows:

• Or they are methods that propose the detection of anomalies based on the comparison of thresholds in the device itself and send a notification when the threshold is exceeded.

• Or they are methods that send the information without any processing to the user of the solution.

5

10

fifteen

twenty

25

30

35

• Or they are methods that send the information without any processing to the server and it is on the server where the information is processed for the generation of notifications.

As for the way of proceeding from these systems, they are either methods that use devices that require an intermediate gateway or server that reads the information of the sensors present in the monitoring devices, or they are methods that use frequency bands without restrictions as to the amount of information transmitted.

At present, a series of innovative communications technologies designed for object communication and advanced mathematical tools based on the use of a lot of information that allow the generation of new applications not possible until now have emerged. These new applications have a series of requirements in terms of functionalities, optimization requirements of the energy consumed and requirements in terms of the amount of information transmitted, among others that require new methods of handling and processing the information. The current state of the art does not solve, or does it inefficiently, problems such as the detection of patterns in extensive livestock, patterns of behavior of older people or patterns of behavior of operators, among others. These applications require the analysis of large amounts of data on the server side and the use of other external data sources that provide information relevant to the application. On the other hand, the devices used in these applications have strong restrictions in terms of energy consumption and the amount of information to be transmitted.

Among the state of the art, the following inventions can be highlighted:

US20110181422. In said invention, reference is made to a method for estimating the position, the state of the person carrying the device, of notifying emergency systems automatically and the determination of patterns associated with a device based on the information of acceleration collected by the device. The method makes use of a telephone line to send the information. In the same way, the system consisting of a device and some algorithms, among others, is protected to estimate the status of the person carrying the device. This patent cannot solve the problems that arise in the present invention for various reasons. First, it focuses on

5

10

fifteen

twenty

25

30

35

the estimation of patterns based solely on the acceleration information while the present invention makes use of the information coming from multiple sensors present in the device. Secondly, the information processing does not take into account information from other devices or external data sources. Finally, the method of this patent uses the telephone line as communication technologies, understood as 2G / 3G communication, which does not impose any restrictions regarding the sending of information.

US20130231574 proposes a system for monitoring the status of people based on a series of physiological variables obtained through a series of sensors arranged on a device. The system affects the measurement of different physiological parameters and the integration with the health services of a hospital or medical center. In no case in the present patent is a method for obtaining patterns proposed by the joint analysis of different devices or the use of external data sources nor does it refer to the optimization of the energy balance by the device.

US8830068. This patent refers to a method to monitor the state of animals and people by measuring physical quantities of the body to be monitored and of magnitudes of the environment in extreme environments. The described method consists of monitoring the state based on the establishment of a non-adaptive threshold. The generation of alarms is based on the fact that one or more variables exceed an established threshold, at no time does it refer to the thresholds being adaptive, or to study a set of parameters of the same animal together, or the study of the parameters of different animals together. In the same way, it does not use location information for anomaly detection. This patent does not establish any criteria in relation to the device that sends the measurements to the server.

US9202193. This patent describes a method for the early detection of diseases on animals. This patent is focused on monitoring through RFID devices that have a completely different field of application to that developed in the proposed invention.

US2002 / 0010390A1. This patent describes a system for monitoring animals in closed environments. It refers to a system composed of a sensor associated with the animal capable of measuring a physiological parameter of the animal, a transmitter coupled to the sensor

5

10

fifteen

twenty

25

30

and a receiver who receives the information. The invention does not refer to the way of processing information for anomaly detection.

US2005 / 0145187A1. This patent refers to a method for animal monitoring and for farm management. The monitoring method consists of sending a beacon signal through satellite communications, although it subsequently indicates that the location and other parameters are also sent. The management method consists of a semi-automatic method of data collection and monitoring of the status of the operation. At no time does the present invention refer to the processing of animal data.

There is then a need for a method and system for monitoring, pattern generation and detection / notification of behavioral anomalies, which resolves the above problems efficiently.

SUMMARY OF THE INVENTION

The objective of the present invention is to develop a method and system that, through electronic devices and, for example, adaptive algorithms and collective intelligence, generates behavior patterns in the bodies that carry the device such as people , animals (or living beings in general). Through these patterns, the method will generate notifications when such behavior patterns are not verified. The devices used can be Internet of Things (IoT) devices that work with networks in frequency bands with limited data transfer such as Sigfox, Lora, NB-IoT, among others.

The proposed method may consist of several different processes such as (this is only an example of the possible processes and not all of these processes are necessarily part of the proposed method):

• Treatment of the information on the Internet of Things (IoT) monitoring device used, and sending it.

• Collection in a single server of all the information from all monitoring devices and obtaining other external data sources (to the server) such as, among others, meteorological information, satellite images, traffic conditions, etc.

5

10

fifteen

twenty

25

30

• Analysis of data from devices and external data sources for the generation of behavior patterns.

• Generation of markers associated with said data. So the user who accesses the information does not have the need to interpret the data since they are shown in a friendly way.

• Generation of notifications when one or more behavior patterns are not verified.

By the proposed method and system, the present invention can:

• Generate patterns based on the joint analysis of information from different data sources such as IoT devices and data from external data sources. Unlike other solutions that analyze the information coming from the devices individually, the present method can analyze the data of the different IoT devices associated to different bodies together and including information coming from other data sources. Based on the data collected, it determines patterns of social and individual behavior among the loT devices. The method establishes a mechanism of “collective intelligence” that consists in establishing behavior patterns taking into account all the devices in the same environment and the additional information that may affect those devices such as the weather, traffic conditions , etc.

• Generate friendly indicators based on the information collected to inform potential users of the solution of the status of the monitored element. Unlike other solutions that show the raw data to the user of the solution, the present method generates, based on the information analyzed, a series of indicators that are presented to said users in a friendly way, such as with color coding (green / yellow / red), with a numerical scale (from 1 to 5, or from 1 to 10 or from 1 to 100), among others. Simultaneously to the generation of these indicators, the method generates notifications when the status of one or more indicators exceeds or falls below a threshold set automatically or by the user of the solution.

• Optimization of the energy consumed in IoT devices. The two elements with the highest energy consumption on any internet device of things are the GPS sensor and the communications modem. With the proposed method, energy consumption is optimized by several procedures:

5

10

fifteen

twenty

25

30

o Unlike other solutions, the proposed method minimizes the amount of information sent to the server through the communications modem. Other solutions send the information collected by the sensor / is available in the monitoring devices grossly, that is, the complete signal. This has a number of advantages since the calculation capacity and the amount of information available in the processor of the IoT device is less than the calculation capacity and the amount of information available in a cloud server. However, for certain applications, battery optimization is a critical factor and therefore it is necessary to limit the amount of information to be sent. In the present method, a mechanism for preprocessing information is proposed in a basic way by calculating basic statistical parameters of each monitored variable and comparing it with basic adaptive thresholds calculated on the same processor. In this way, the device only sends the information when any of the statistical parameters exceeds the threshold. In addition, instead of sending the complete information of the monitored signal, only the information associated with the calculated statistics is sent.

or Additionally, to minimize the energy consumption in the information transmitted by the modem of the IoT device, the information is compressed.

o In the same way, when the internet device of things

With a GPS sensor, the acquisition of information from that sensor is carried out on a timed basis or at events to optimize the energy balance in the device.

Specifically, in a first aspect, the present invention proposes a method for monitoring (generating behavior patterns and notifying behavioral anomalies) of living beings (generally a group of people or animals to be monitored), where each living being monitored It carries an electronic monitoring device, where the method comprises the following steps performed by each monitoring device:

5

10

fifteen

twenty

25

30

35

a) Receive information on one or more physical, physiological and / or environmental variables of the living being that carries the monitoring device, measured by one or more sensors associated with the monitoring device;

b) Obtain statistical parameters for each variable in temporary windows by processing said information received;

c) Compare the value of each statistical parameter in the current time window with thresholds previously calculated by the monitoring device, based on at least previous values of said statistical parameters;

d) If, as a result of said comparison, no anomaly is detected in any of the statistical parameters, recalculate the value of each of the thresholds taking into account the current value of each statistical parameter in the current time window;

e) If, as a result of said comparison, any anomaly is detected in any of the statistical parameters (for example, because the value of a parameter is higher or lower, as the case may be, at that threshold with which it is compared) or, if it has been more than a predetermined time since the last sending of information, activate a location sensor (for example, a GPS location system or any other type), obtain information on the current location of the device by means of said location sensor and compress and send information of one or more of said statistical parameters and the current location of the monitoring device (1) to a server. This information that is sent on the statistical parameters can be of different types such as, for example, the values of the parameters in which the anomaly has been detected only or all the parameters of the device, the values in the current time window or in previous windows, an average of the parameters in several temporary windows .... or any other information about said statistical parameters that is of interest.

The method also includes the following steps performed by the server:

f) Collect information received from one or more of the monitoring devices of the group of living beings;

g) Collect information received from one or more external data sources;

5

10

fifteen

twenty

25

30

35

h) Generate behavior patterns for each monitored device based on a statistical processing of the information collected;

i) Upon receiving information sent by a monitoring device, determine whether said information verifies the behavior patterns of said monitoring device, previously generated and stored by the server;

j) If it is determined that the information received from any of the monitoring devices does not verify the behavior patterns of said monitoring device or the information received from said monitoring device is not within a previously defined range of values (so automatic or manual by the system user), notify at least one user associated with said monitoring device through a communication network that an anomaly has been detected in the behavior of said monitoring device.

The generation of patterns (including the individual) is based on at least the information received from more than one monitoring device. That is, not only the information received from the device of the living being in question whose pattern of behavior is to be generated is taken into account, but also information received from other devices in the group. Even, in one embodiment, information received from all the devices in the group or from devices that have some characteristic in common (for example, that are in a certain area, that perform the same work, that have the same age, that are same type, genus or species ....).

The statistical parameters obtained in step b) are at least one of the following: mean, variance, standard deviation, maximum, minimum or mode, of the values of each variable in each time window (or any other known statistical parameter).

The method can also comprise a presentation step (by previously sending the information or because the user can access the server) to the user associated with each monitoring device, part or all of the information received from each monitoring device through a user interface (typically this Information will be presented visually clearly and easily to the user, for example, through graphics, colors.).

5

10

fifteen

twenty

25

30

35

In one embodiment, the information to be sent in step e) is also encrypted and / or encrypted in the monitoring device before being sent to the server.

The generation of patterns in step h) can be performed by applying known statistical techniques on the information collected, such as time series, Logistic regressions, Decision Trees, Support Vector Machines, Bayesian Statistical Analysis, Collective Intelligence Algorithms or Grouping These patterns are generated for each monitored device and can be individual, collective or social.

In a second aspect, the present invention proposes a system for monitoring a group of living beings, where the system comprises:

- A monitoring device for each living being in the group (carried by each living being), where each monitoring device comprises:

- One or more sensors configured to measure one or more physical, physiological and / or environmental variables of the living being carrying the monitoring device;

- A location sensor (sub-system);

- A communication module (also called a communication interface) configured to communicate with a server;

- A (micro) processor configured to:

- Obtain statistical parameters for each variable by processing the information received from the one or more sensors in temporary windows;

- Compare the value of each statistical parameter in the current time window with thresholds previously calculated by the monitoring device, based on at least previous values of said statistical parameters;

- If, as a result of said comparison, no anomaly is detected in any of the statistical parameters, recalculate the value of each of the thresholds taking into account the value of each statistical parameter in the current time window;

- If, as a result of said comparison, any anomaly is detected in any of the statistical parameters or, if more than a predetermined time has elapsed since the last sending of information, activate the sensor

location, get information on the current location of the device

10

5

10

fifteen

twenty

25

30

35

using said location sensor and compressing and sending the value of one or more of said statistical parameters and the current location of the monitoring device to a server through the communication module ;,

- The server that includes:

- A first communication module configured to communicate with monitoring devices, a second communication module configured to communicate with external data sources and a third communication module configured to communicate system users (these modules can use the same or different communication technologies and establish communication through the same or different communication networks);

- A (micro) processor configured to:

- Collect information received through the first communication module of one or more of the monitoring devices of the group of living beings;

- Collect information received from one or more external data sources through the second communication module;

- Generate behavior patterns for each monitored device based on a statistical processing of the information collected (from the monitoring devices of the group of living beings and from external data sources);

- Upon receiving information sent by a monitoring device, determine whether said information verifies the behavior patterns of said monitoring device, previously generated and stored by the server;

- If it is determined that the information received from any of the monitoring devices does not verify the behavior patterns of said monitoring device or the information received from said monitoring device is not within a previously defined range of values, notify the less a user associated with said monitoring device by means of the third communication module that an anomaly in the behavior of said monitoring device has been detected.

The generation of patterns (including the individual) is done at least based on the information received from more than one monitoring device. That is, not only the information received from the device of the living being in question whose pattern of behavior is taken into account

you want to generate, but also information received from other devices in the group.

eleven

5

10

fifteen

twenty

25

30

35

The devices can be electronic devices that use Internet of Things Technology with limited data transfer (and that connects directly to the Internet without the need for intermediate gateways). The devices use one of the following Sigfox, Lora, NB-IoT or LTE-M communication technologies to communicate with the server.

Monitored living beings can be, for example, people (such as workers who work in a particular facility, elderly, sensitive or disabled people who want to be monitored for their safety ...) or animals (for example, livestock).

The variables measured by the sensors in each living being can be of any type, such as, for example, surface temperature, ambient temperature, movement, acceleration, rotation, humidity, gases, heart rate, bioimpedance ... and information on sources of External data (external databases) can be for example meteorological information, traffic status information, information from satellite images or information from social networks.

Finally, in a third aspect of the invention there is a computer program comprising instructions executable by computer to implement the described method, when running on a computer, a digital signal processor, an application-specific integrated circuit, a microprocessor , a microcontroller or any other form of programmable hardware. Said instructions may be stored in a digital data storage medium.

For a more complete understanding of the invention, its objects and advantages, reference may be made to the following specification and the accompanying drawings.

DESCRIPTION OF THE DRAWINGS

To complement the description that is being made and in order to help a better understanding of the features of the invention, in accordance with some preferred examples of practical embodiments thereof, a set of drawings is accompanied as an integral part of this description. where, for illustrative and non-limiting purposes, the following has been represented:

5

10

fifteen

twenty

25

30

35

Figure 1 schematically shows the basic architecture of the proposed method and system according to an embodiment of the invention.

Figure 2 schematically shows the operation of the method and system proposed in the part of the device (internet of things) according to an embodiment of the invention.

Figure 3 schematically shows the operation of the method and system proposed in the server part according to an embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION

The present invention proposes a method and system for generating patterns based on the information collected by electronic devices (optimizing the energy consumed by said devices) and based on the information collected through external data sources, open or proprietary, such as data from meteorological services, traffic information, satellite images, social networks, etc.

These devices can be internet of things (IoT) devices. That is, devices with internet connection (with direct connection, without intermediate residential gateways, or any other type) through a communications network. These communications networks can be of any type but they are usually networks that use frequency bands for connection that limit the flow of data to be sent (such as networks that use Sigfox, Lora, NB-IoT or other technologies communication protocols of communication of this type).

The method proposed by the present invention can be said to have two distinct stages, the part that is executed in the monitoring device and the part that is executed in the server.

On the part of the IoT monitoring device, there are currently and increasingly a greater number of devices that allow the monitoring of physical variables relative to the body carrying the device as well as its geo-location. These devices usually operate in free bands such as ISM ("Industrial, Scientific, Medicís", Industrial, Scientific or Medical) or proprietary, and can send the information directly to a server without the need for gateways (residential or otherwise)

13

5

10

fifteen

twenty

25

30

35

intermediate. In addition, these devices have different sensors associated with them such as, among others, location sensors (for example, GPS, "Global Positioning System"), accelerometers, gyroscopes, temperature sensors that measure the surface temperature of the body or ambient temperature, humidity, gas, heart rate, bioimpedance sensors, etc. The biggest challenge of these devices is, in all cases, the optimization of the battery life and in some cases also, the sending of the information itself given the restrictions present in some frequency bands.The method of the present invention, on the side of the device, is executed in the microcontroller (also called processor or microprocessor) that the device has and in one embodiment Understand the following steps (this is just an example and not all of these steps are mandatory in the proposed method):

• Establish continuous temporary windows that can overlap or not, during which the device's microprocessor collects information from the sensors that measure physical variables of the body that carries it and / or environmental variables.

• Infer different statistical parameters of the recorded variables, such as, among others, mean, maximum, standard deviation, during each of the windows independently.

• Temporarily store the statistical parameters of each variable in the device.

• Establish, for each statistical parameter of each variable, an adaptive threshold, which is established by moving averaging or another procedure of the previous values of said variable. The differential factor of this aspect lies in the adaptability of the thresholds that adapts to the natural variability of the bodies that carry the devices.

• Analyze the current value of each statistical parameter of each physical or environmental variable and compare it with the previously calculated threshold.

• Obtain the location information associated with the device if any of the statistical parameters exceeds the thresholds, or on a timed basis. In this way, it is achieved that since the GPS devices are, together with the communications modem, the elements of an IoT device with greater energy consumption, an optimization of the battery life. Thus, the normal state of the GPS sensor is turned off, timed on or to events. Additionally, to optimize the battery, mechanisms are used that, although

5

10

fifteen

twenty

25

30

35

turn off the GPS sensor, maintain the state of the satellite constellations ("hot start"), reducing the time for the acquisition of the geo-location measurement.

• Compress the information to be sent to optimize battery life and to comply with spectrum restrictions in certain frequency bands such as the 868MHz ISM band that use communication protocols such as Sigfox or Lora.

• Optionally, encrypt and encrypt the information for sending securely.

• Send the statistical parameters of the different monitored and geo-location variables, at regular configurable intervals or by events when the threshold value of any of the variables is exceeded. In this way, a battery optimization is also achieved since the communications modem, which is another of the elements that consume more energy in an IoT device, is normally turned off.

The second part of the method refers to the server. Server means a series of computer resources such as processors, memories, hard drives, network cards, etc. that reside in a single physical installation or that is distributed in a set of them and that is connected to the Internet. The method described in an embodiment of the present invention, on the server side consists of (this is only an example and not all of these steps are mandatory in the proposed method):

• Collect the information of all IoT devices associated with the same type of entity, whether animal, person or object. This information contains a series of statistical parameters associated with a series of physical variables of the monitored and / or environmental body.

• Collect information from various external data sources that may be open or proprietary, such as, among others, database information related to the objects of study, meteorological information, satellite images, social networks ...

• Model the information together and adaptively from the physical and / or environmental variables monitored by the IoT devices, and taking into account the information from external data sources, through tools such as:

o Time series. o Logistic regressions.

5

10

fifteen

twenty

25

30

35

o Decision trees. o Support vector machines. o Groupings. o Bayesian statistical analysis.

The fact that the information can be processed together when the monitored bodies have common aspects such as type and area is highlighted.

• Obtain a pattern of behavior associated with the model.

• Evaluate using a friendly scale (color scale, indicator, numerical) the current state of the monitored body in relation to the model.

• Generate notifications when one or more physical or environmental variables do not verify the pattern of behavior generated.

Figure 1 shows the overall architecture of the proposed system, which is composed of several elements:

• Devices that collect and process information from people, animals or objects that the user wishes to monitor. Normally, each of these devices collect information (for example, through sensors) about a specific entity (an object or living beings such as animals, people or plants). The devices are usually Internet of Things (IoT) devices (1) that work in frequency bands where the maximum amount of information to be transmitted is limited by restrictions in the use of the radio spectrum (such as, for example, protocols of communication Sigfox, Lora, NB-IoT or LTE-M or any other known), and where the optimization of energy consumption is essential for its use in applications that may be remote or because the usual replacement or recharge of Battery. These IoT devices are designed to work in both rural and urban environments and do not require the installation of any additional elements. The device is connected directly to the internet.

• On this IoT device, one of the two parts of the method described in the present invention is executed. The device executes (2) the proposed method (the part of the device) analyzing the information coming from the sensors that measure physiological variables of the body they are monitoring or environmental, and calculating a series of basic statistical parameters such as the mean, variance, mode, maximum, minimum or any other statistical parameter. Through these parameters, the device generates thresholds, which if exceeded, or in a timed way

5

10

fifteen

twenty

25

30

35

(that is, if information has not been transmitted to the server in a certain period of time), it sends the previously compressed information to the server (3) of the solution.

• The server (3) executes (4) the part of the server method. To do this, the server analyzes the data, coming from the devices associated with the same body type, together using data analysis techniques such as, among others, time series, decision trees, clusters, vector machine machines support, etc. so that individual, collective and / or social patterns are generated between the devices. Individual patterns affect the behavior of a monitored living being and take into account, among other things, the information received from the device that carries said living being. Collective patterns affect the behavior of a certain group of living beings and take into account, among other things, the information received from the devices carried by each of the group's living beings. The social patterns said living being affect the interaction behavior between living beings of a certain group of living beings and takes into account, among other things, the information received from the devices carried by each of the group's living beings.

In addition, the server uses, for the generation of such patterns, information from external data sources (5), open or proprietary, such as meteorological information, satellite images, traffic conditions, external databases, information from social networks, etc.

• Once the information from the IoT devices (1) and external data sources (5) is analyzed, the server generates a series of individual, collective or social behavior patterns (6) of each of the devices that form part of the solution, or more specifically patterns of behavior of the entities to which each device is monitored. On these patterns, the current information collected by the IoT devices (1) can be compared and when one or more patterns are not verified, a notification (7) is generated to the system user. This notification can be made using any known communication network. In the case of people, this notified user may be the same user to whom the device is monitored; but in general it will be an external user who uses the proposed system to monitor different entities (for example, the owner of a livestock farm if the monitored beings are the livestock of that farm). The non-verification of the patterns is established automatically or manually by the user of the solution. Additionally, and for the compression of the solution by the user of the same, the server can generate

5

10

fifteen

twenty

25

30

35

Indicators (8), which can be represented on an understandable numerical scale or by color code among others, associated with each of the physical, environmental and social magnitudes of each of the devices present in the solution.

The devices used in the present invention to collect (and process) data from the monitored entities, are usually internet devices of things with certain characteristics such as:

• They are devices that work in frequency bands with limitation as to the amount of data to be sent. Unlike other devices that use technologies such as GSM, WiFi, Bluetooth, etc., where there is no limitation as to the amount of data or the type of data (data, images, video, etc.), the devices used in the This method only allows the sending of small amounts of raw data with high latencies and in some cases without confirmation of receipt of the information. Examples of existing technologies with such limitations of data to be sent can be:

or Sigfox. Sigfox technology operates in the ISM frequency bands that are in Europe at 868MHz, while in other regions it focuses on the 915MHz band. Sigfox technology allows sending up to 70 bytes / hour.

o Lora (in English, "long range low power network", long range and low power network). Lora technology operates in the same frequency bands as Sigfox technology. The limitation in terms of sending data by the Lora technology is between 8.3 to 208 bytes / hour.

or NB-IoT (from English "Narrow Band loT", "narrow band loT" in Spanish), LTE-M (from English "Long Term Evolution for Machines", "Long term evolution for machines" in Spanish), etc. . NB-IoT technology offers speeds of up to 250kbps (up and down) and a latency between 1.6 and 10 seconds. Limit upload packet size to 1000 bytes.

• They are devices in which the optimization of energy is a fundamental element since they are used in applications where the usual replacement of the battery or its charge is not possible. For example, extensive livestock monitoring devices where animals are not accessible at all times.

• They are devices that use a communication module (also called a communication interface) that allows them to connect directly to the Internet without the need for intermediate gateways. Therefore, the devices, provided there is coverage of the

5

10

fifteen

twenty

25

30

35

Radio frequency technology used, sends the information directly to a central server (through the communication module) without the need for intermediate gateways that collect the information. It is also understood as devices that use intermediate gateways to devices that use technologies such as Bluetooth, RFID or NFC, among others, connect to a hub or mobile phone or other element for sending information. On this type of devices, the method of the present invention has no application.

• These are devices that have a locator (for example, a GPS system or any other location system) for obtaining geo-location information. This geo-location information is essential for the extraction of behavior patterns.

Figure 2 shows the scheme of operation of the method in the part of the device according to an embodiment of the invention. The device has one or more sensors (in an alternative embodiment some of these sensors are not present but communicate with it through a communication network). Each of these sensors measure one or more physical, physiological or environmental variables in relation to the body (animal, person or object) that monitors the IoT device (1) (in a preferred embodiment the device is usually found in said monitored body) . These sensors provide a continuous flow of information (10) with the measurements made.

The first part of the method consists in generating overlapping or non-overlapping temporary windows (11), so that a smaller subset of information (12) is obtained. On the poisoned information the method executes a statistical calculation process (13) that generates a series of n (n> = 1) basic statistical parameters (14a .... 14n) such as, among others, mean, median, standard deviation, variance, mode, autocorrelation, frequency components, etc. The next step consists in the process of calculation (15a .... 15n) of adaptive thresholds for each of the previous statistics. These adaptive thresholds (16a ... 16n) for each of the previous statistics are calculated by means of the temporal (historical) analysis of the statistics and are updated every time a new window is processed. The threshold comparison process (17) compares the current statistical values (14a .... 14n) of each of the variables with the threshold values (16a ... 16n) and in case some value exceeds the threshold or in the event that the minimum signal sending timing is exceeded (that is, if a certain period of time has elapsed without sending information from the device to the server), the location value is taken, by

5

10

fifteen

twenty

25

30

35

example, from the GPS sensor (18). Next, the integration process (19) collects the statistical parameters to be sent (which may be only the statistical parameters that have exceeded the threshold or all the statistical parameters of that moment) together with the geolocation information (of that moment). Subsequently, the information is compressed (20) and optionally encrypted and encrypted to improve data security. Finally, the process of sending data (21), sends the information to the server.

The purpose of the generation of temporary windows (11) is to limit the temporal range of calculation of the statistical parameters associated with each variable. In this way the statistics of each variable in that time region are obtained and studied. The temporal regions are different from each other although they may overlap. For the calculation of the adaptive thresholds of each of the statistics used, the present method uses techniques such as, among others, moving averaging of the previous samples or threshold selection techniques by statistical ordering of the sample values. Adaptive thresholds are updated with the appearance of each new temporary window. In this way, the data is analyzed in real time.

The threshold comparison process (17) can have the following operation:

• The current value of each statistic result of the current window is compared with the threshold values, to detect anomalies.

o If any parameter exceeds or falls below the threshold (as the case may be, since for some thresholds the anomaly is that it is below the threshold and others above; the threshold may also consist of a range of values and the anomaly it is notified when the parameter is outside that range of values), the information is sent to the central server. In addition the thresholds are updated.

o If none of the parameters exceed or fall below the threshold, the threshold values are recalculated.

• Periodically and regardless of whether an anomaly is detected in any parameter, the threshold generation process (17) generates, through the history of values, a collection of statistical parameters for sending to the server. This compilation may consist of the values corresponding to the statistical parameters that coincide with that temporary moment or averaging of all the values after the last sending of information.

5

10

fifteen

twenty

25

30

35

The information integration process (19) aims to obtain and integrate GPS information. The reason for not using information from GPS continuously is to minimize the energy consumption of the device. In conditions of tracking or obtaining the position, the GPS device consumes between 25-40mA which implies a very significant part of energy for internet devices of small things and where the battery life is a critical element. To optimize the time until obtaining the location, which under normal conditions can be around 30 seconds, this method may use software techniques or ignition mechanisms of the "hot start" type that can reduce the acquisition time of the information. up to 3-5 seconds which implies a consumption reduction of up to 85%.

The information compression module (20) compacts the information prior to sending in order to minimize the energy consumption of the information transmission. This compression of the information can result in a loss of the accuracy of the measurement but that from the energy point of view and taking into account the limitation of the data transmission that involves the use of certain bands of the radio spectrum such as used by Sigfox, Lora or NB-IoT technologies, among others, it is acceptable. By way of example, the method of the present invention would compress the average temperature value with a resolution of 0.3 ° using 8 bits of information. Other solutions use 32 or 64 bits to send the same information, without losses, but obviating that in many cases the accuracy of the sensor is lower than the accuracy of the measurement. With this example, the amount of information to be sent would be reduced by a factor of 4 with the consequent energy savings.

Additionally, the device, prior to sending the information, may encrypt and encode the previously compressed information in order to increase the security of the information sent. The method will apply basic encryption techniques, given the limitations of calculation capacity of the microcontrollers present in the internet of things devices.

Figure 3 schematically shows the operation of the method and system proposed in the server part according to an embodiment of the invention. Information from internet devices of things (1) previously analyzed, processed and compressed in the device (1), as explained above, reaches the server (3) of the proposed solution. The server can also collect information from external, open or proprietary data sources (5) that may be related

with the devices used (these external sources or databases can be for

twenty-one

5

10

fifteen

twenty

25

30

35

examples on servers external to the system with which the server communicates (3) to obtain the data it needs). This information from external data sources can be for example: weather information, satellite images, traffic conditions, schedules / activity planning etc. or any other type of information of interest.

The information from the IoT monitoring devices (1) (the information of the sensors associated with the device) is collected by the data collection process of the IoT devices (22), while the information from the external data sources (5) is compiled by the process of data collection from external sources (23) which also adapts the information for further processing. To communicate with external data sources, the server uses a communication network that can be used by any of the known communication technologies. All this information is sent to a server processor that, by means of an information analysis algorithm (24), collects the information from all the data sources and executes a statistical processing on them by applying techniques such as (this list is only given to by way of example and not intended to have any limiting effect):

• Temporal series. The method applies time series techniques such as ARIMA (integrated self-regressive model of moving average, English "Autoregressive integrated moving average") or any other, on the statistical values from the sensors of the devices. it obtains a time-value relation of each one of the parameters and allows a prediction of future values in the short term.As an example, the surface temperature of an animal has a very marked temporal behavior since the night temperature is lower than the During the day, by means of the temporal analysis of the surface temperature, behaviors such as loss of the monitoring device (1) among others can be inferred.

• Logistic regressions. Through logistic regressions, short-term values of the statistical parameters of the variables can be predicted based on the values of other variables. This type of technique is useful for predicting the value of a variable based on the state of other variables.

• Algorithms of collective intelligence that allow the obtaining of social behaviors of the devices (specifically of the entities that carry them). This type of social behavior information allows obtaining some patterns of

5

10

fifteen

twenty

25

30

35

behavior that are very useful on certain occasions such as affinity relationships in livestock animals in extensive.

• Support vector machines (SVM), through prior training or through training through a user application, the method may use SVM techniques to infer patterns based on the values of the parameters of the different variables.

• Groups. The method may perform groupings of variables and associate them with certain behaviors present in the individuals being monitored.

After this statistical processing, the pattern generation module (25) will produce patterns of individual and / or collective and / or social behaviors (6) of all the monitoring devices of the system (which will correspond to patterns of individual behaviors and / or collective and / or social of the living being associated with said device), based on the history of the data received and information from external databases (in addition it can also take into account apart from design parameters adjustable by the system user) .

If the data received from any of the devices is not within the pattern of behavior (individual, collective or social), the notification generation module (26) will generate notifications (7) and send them through a communication network ( using email, SMS, push notifications or any other known notification technology), to the end user (or users) of the system (the server usually has a list with the users to notify according to the device whose behavior is abnormal). To do this, you will use a communication network that can use any of the known communication technologies (GSM, 3G, LTE, 4G. Or any other). If there are several devices whose behavior is anomalous, a single notification can be sent listing the devices whose behavior is anomalous.

In other words, in the event that some pattern of behavior is not verified, the notification generation module notifies the user. For its part, the indicator generation module (27) generates a series of indicators (8) that are represented in a friendly way (for example, in the form of a color code or numerical scale), in relation to the variables

5

10

fifteen

twenty

25

30

monitored to facilitate the understanding of the information by the user of the solution.

If the behavior of the device is not anomalous (that is, the data corresponding to said device are within the expected behavior pattern), the behavior patterns can be recalculated taking into account the statistics corresponding to said device.

The server will also include (although not explicitly shown in Figure 3) communication modules (also called communication interfaces) that have the technology necessary to allow the server to communicate with both monitoring devices (using for example Sigfox or Lora technology) as with the end users of the system (to send them notification of anomalies or present information to them) and with external data sources.

The proposed method and system could, for example, generate, among others, the following behavior patterns (this listing is only given by way of example and is not intended to have any limiting effect):

• Pattern of social behavior in animals. Analyzing the geo-location information of different animals together, a marker indicating the social relationships of the animals with each other can be established. When this pattern is not verified, this method generates a notification to the farmer indicating a possible anomaly in a given animal.

• Pattern of rumination and feeding in animals of a certain area using information from all the devices mounted on animals (accelerometer and geo-location), from satellite images (images of normalized vegetation index, NVDI), information from sources of meteorological data (temperature evolution, rainfall, etc.) and even land use information where animals are located. Through the generation of this pattern, land use can be optimized for better animal feed.

5

10

fifteen

twenty

25

30

• Intrusion, attack and / or theft detection pattern in the territory where the group or herd is based on the variation, among others, the following patterns: sudden and substantial variation of the group's activity levels; disintegration or dissolution of the group based on geo-location information; isolation of certain individuals from the group; cessation of activity of some or some individuals of the group; variation in temperature levels; location of individuals outside their usual territory geographically delimited.

• Behavior pattern for operators of hazardous facilities, carriers, cleaning operators, etc. Using the geo-location information of the devices associated with the person and the information coming from external data sources such as meteorological information and traffic status. Through these patterns, anomalies can be generated, for example, that an operator is in an unauthorized risk zone in an installation, and optimize the use of resources, since for example an optimal map for maintenance could be generated and cleaning of a certain city.

• Elderly behavior pattern. Through geo-location information and sensors such as accelerometer, gyroscope, gas detectors, etc. and from information from external data sources such as meteorological information, a pattern of routine behavior of older people can be established. In this way, when the data coming from the monitoring device carried by the elderly person is not within the expected routine behavior, the user (nurse, family member ...) can be notified to verify that the older person is well.

• Pattern of nocturnal movement of arch of the legs in people. Through the acceleration and rotation sensors present in the device and other variables from external data sources, a series of markers such as amount of movement, range of motion, etc. can be generated. of the person (valued on a friendly scale). These markers can be used later by a specialist to provide the best treatment or care.

5

10

fifteen

twenty

25

30

35

One of the many novelties of the present invention lies in the use of external data sources that can be open or proprietary. These data sources may be among others (this list is only given by way of example and is not intended to have any limiting effect):

• Geolocated meteorological information at the location of the device. There are many online services, free or paid, that provide weather information based on a position. This information, by way of example, can be used in relation to the behavior of older people, since, for example, if the weather has heavy rains or low temperatures, it will be normal for the person to change their behavior pattern.

• Information from proprietary databases. On certain occasions the use of information from external databases provide fundamental information for the system. As an example, in the event that internet devices of things are carried by operators, information for example, work shifts, work permits, etc. is essential.

• Information from traffic status. There are internet services that provide traffic status information. In this case, this information is useful, for example, when internet devices of things are associated with vehicle fleets.

• Information from satellite images. Information from satellite images provides information in different frequency bands in relation to the state of the earth's surface. In this case, the process of collecting data from external sources (23) processes the information from the satellite images to generate indicators that affect the entities monitored by the internet devices of things. As an example, by processing the information from the normalized vegetation index (NDVI) images, the quality of the grass where the animals are found can be estimated and this information can be related to the location information and other sensors from the devices to infer behavior patterns; since, for example, if analyzing the NVDI images, a low grass quality is estimated, it is normal for the

5

10

fifteen

twenty

25

30

Animals have a greater activity for the search for food. Therefore, it is essential to analyze this satellite information to obtain behavioral patterns.

• Information from social networks. Using data mining techniques, patterns that can influence the monitoring devices that are monitoring a particular entity can be extracted.

Through the collective intelligence algorithms, the social relationships between the different individuals that carry the internet device of things are analyzed. These algorithms analyze the individual information of each of the devices and the relationships established with the information coming from other devices, establishing affinity relationships. As an example, studying the value of the individual and joint location of each of the animals of an extensive livestock farm establishes "friendship" relationships between different animals of the farm. Usually, when these relationships are broken, it is a indicative that an anomaly is occurring in the individual or in the exploitation.

As a recap, we will then summarize the proposed method for obtaining patterns based on monitoring devices and data analysis, for a preferred embodiment.

As indicated, the method in its preferred embodiment (Figure 1) is made up of two differentiated process steps: the method in the device (2) is executed on the monitoring device (1) (normally devices that use "technology" internet of things ”) and analyzes the values related to the physical, physiological (of the animal or object that carries the device) or environmental variables measured by sensors installed in said device. The device can be carried on a collar, ankle brace, in the clothing that the wearer is wearing, in a pocket, in a harness or by any other known means or system.In another embodiment, some of said sensors may not be installed in the device but be external to it and communicate with it with any technology of known communication The analyzed information is sent to a central server (3) depending on whether the statistical parameters associated with these variables exceed a threshold or on a timed basis.

5

10

fifteen

twenty

25

30

35

In the central server (3), the server performs a method (4) by which it analyzes the information coming from the internet devices of things and using information from external, open or proprietary data sources, (5) , generates a series of patterns of individual, collective and social behavior (6) using statistical techniques such as time series, groupings, Bayesian techniques, among others. In the same way, when any of the patterns is not met or exceeds a threshold set automatically or manually by the user of the solution, the server method (4) generates notifications (7) to the users of the tool. Finally, the server (4), after analyzing the data, can generate a series of indicators (8) to facilitate the compression of the information to the end users of the solution. These indicators (8), for example, show on an understandable numerical scale or in color coding, the state of the variable associated with the internet device of things associated with a person, animal or object.

In one embodiment, the present invention is used for monitoring unconfined animals.

In one embodiment, the internet monitoring devices of things (1) are installed on cattle, for example on livestock animals in extensive.

In one embodiment, the variables monitored by the internet of things monitoring device (1) can be: body surface temperature, 3-axis acceleration and GPS location although of course it is possible to monitor other variables by means of the appropriate sensors.

In the embodiment, the sending of information is also carried out in a timed manner regardless of the threshold values set in the device.

In the embodiment, the information is compressed into 12 bytes of information (although any other understanding is possible).

In the embodiment, the central server (3) generates patterns of behavior (6) individual, collective and social, for example the use of Bayesian techniques and time series based on the data received by all devices located in the same region.

In the embodiment, the central server (3) collects information from external data sources (5)

such as meteorological information and satellite information.

28

5

10

fifteen

twenty

25

30

In the embodiment, the central server (3) generates notifications (7) when the behavior patterns (6) are not met based on parameters set automatically based on the preferences of the end user of the application.

In the embodiment, all the information from each device is presented to the end user of the information summarized in several (for example, six) indicators (8) with a color code to facilitate their understanding.

In this text, when we talk about behavior patterns of monitoring devices or behavior patterns of living beings (animals or people) or objects; both expressions refer to the same concept, that is, to patterns of behavior of the living being or object that carries said monitoring device (which are obtained based on, among other factors, information on the physiological and environmental variables of the living being or object collected and sent to the server by the monitoring device). In no case, the present invention makes any kind of diagnosis on the monitored entities.

Although in the present text, there is usually talk of a server (3) (also called a central server), this would be in a preferred embodiment. In other embodiments, the system may have more than one server with which the devices communicate (for example, different groups of devices can communicate with different servers, that is, each server can manage different groups of devices); performing all these existing servers in the system functions analogous to those described for the central server (3).

In this text, the term "comprises" and its derivations (such as "understanding", etc.) should not be understood in an exclusive sense, that is, these terms should not be construed as excluding the possibility that what is described and defined can include more elements, stages, etc.

Some preferred embodiments of the invention are described in the dependent claims that are included below.

Describing sufficiently the nature of the invention, as well as the way to perform it in practice, it should be noted that its different parts can be manufactured in a variety of materials, sizes and shapes, and can also be introduced into its constitution

29

or procedure, those variations that the practice advises, as long as they do not alter the fundamental principle of the present invention.

This detailed description is provided to help a complete understanding of the invention. Therefore, those skilled in the art will recognize that variations, changes or modifications of the embodiments described herein can be carried out without departing from the scope of protection of the invention. Also, the description of functions and elements well known are omitted for clarity and conciseness. The description and drawings simply illustrate the principles of the invention. Therefore, it should be appreciated that those skilled in the art will be able to devise several provisions that, although not explicitly described or shown in this document, represent the principles of the invention and are included within its scope. In addition, all the examples described in this document are provided primarily for pedagogical reasons to help the reader understand the principles of the invention and the concepts contributed by the inventor (s) to improve the technique, and should be considered as non-limiting with respect to such examples and conditions described specifically. In addition, everything stated in this document related to the principles, aspects and embodiments of the invention, as well as the specific examples thereof, encompass equivalences thereof.

Although the present invention has been described with reference to specific embodiments, those skilled in the art should understand that the foregoing and various other changes, omissions and additions in the form and detail thereof can be made without departing from the scope of the invention. as defined by the following claims.

Claims (10)

5 10 fifteen twenty 25 30 35 1. Method for monitoring a group of living beings, where each living being in the group is an extensive livestock animal and carries a monitoring device (1), where the method comprises the following steps performed by each monitoring device (1 ): a) Receive information on one or more physical, physiological and / or environmental variables of the living being that carries the monitoring device (1), measured by one or more sensors associated with the monitoring device (1), where the one or more variables they are at least one of the following: surface temperature of a body of the living being, movement of the living being, acceleration of the living being or rotation of the living being; b) Obtain statistical parameters for each of the one or more variables in time windows by processing said information received, where the statistical parameters are at least one of the following: mean, variance, standard deviation, maximum, minimum or mode of the values of each variable in each time window; c) Compare the value of each statistical parameter in the current time window with thresholds previously calculated by the monitoring device (1), based on at least previous values of said statistical parameters; d) If, as a result of said comparison, no anomaly is detected in any of the statistical parameters, recalculate the value of each of the previously calculated thresholds taking into account the current value of each statistical parameter in the current time window; e) Send information of one or more statistical parameters and a current location information of the monitoring device (1) to the server (3): at regular intervals; I if, as a result of said comparison, any anomaly is detected in any of the statistical parameters or if more than a predetermined time has elapsed since the last submission of information, where the monitoring device (1) uses at least one of the following communication technologies: Sigfox, Lora, NB-IoT or LTE-M to communicate with the server (3), and where the monitoring device (1) activates timed form a location sensor associated with the monitoring device (1) and obtains information on the current location of the monitoring device (1) before sending information to the server (3); where the method also includes the following steps performed by the server 5 10 fifteen twenty 25 30 f) Collect information received from the statistical parameters of the monitoring devices (1) of the group of living beings; g) Collect information received from one or more external data sources (5); h) Generate behavior patterns for each monitoring device (1) based on a statistical processing of the information collected in steps f) and g), where the generation of behavior patterns for each monitored device (1), is based on at least to the information received from more than one monitoring device (1) of the living beings of the group, and where the generation of patterns is carried out by applying at least one of the following techniques on the information collected: Time series, Logistic regressions, Decision Trees, Support Vector Machines, Bayesian Statistical Analysis, Collective Intelligence Algorithms or Clusters; i) Upon receiving information sent by a monitoring device (1), determine whether said information sent verifies the behavior patterns of said monitoring device (1), previously generated and stored by the server (3); j) If it is determined that the information received from any of the monitoring devices (1), does not verify the behavior patterns generated for said any of the monitoring devices (1) or the information received from said monitoring device (1) it is not within a previously defined range of values, to notify at least one user associated with said one of the monitoring devices (1) through a communication network that an anomaly in the behavior of said one of the monitoring devices (1). 2. The method according to claim 1 wherein the monitoring device (1) of each living being is an electronic device using Internet of Things Technology with limited data transfer. 3. Method according to any of the preceding claims wherein the location sensor is a GPS location system. 4. Method according to any of the preceding claims which further comprises: k) Present to the user associated with each monitoring device (1) part or all of the information received from each monitoring device (1) through a user interface. 5 10 fifteen twenty 25 30 35 5. Method according to any of the preceding claims wherein the information to be sent in step e) is also encoded and / or encrypted in the monitoring device (1) before being sent to the server (3). 6. Method according to any of the preceding claims wherein the information of the one or more external data sources (5) is at least one of: meteorological information and information from satellite images. 7. Method according to any of the preceding claims wherein the patterns generated for each monitoring device (1) are individual, collective or social patterns. 8. Method according to any of the preceding claims where for the generation of behavior patterns for each monitoring device (1) in step h), it is based on the information received from all the devices of the group of living beings whose location is find within a given region. 9. System for monitoring a group of living beings, where each living being in the group is an extensive livestock animal, where the system comprises: - A monitoring device (1) carried by each living being in the group, where each monitoring device (1) comprises: - One or more sensors configured to measure one or more physical, physiological and / or environmental variables of the living being that carries the monitoring device (1), where the one or more variables are at least one of the following: surface temperature of a body of the living being, movement of the living being, acceleration of the living being or rotation of the living being; - A location sensor; - A communication module configured to communicate with a server (3) using at least one of the following communication technologies: Sigfox, Lora, NB-IoT or LTE-M; - A processor configured to: - Obtain statistical parameters for each variable by processing the information received from the one or more sensors in time windows, where the statistical parameters are at least one of the following: mean, variance, standard deviation, maximum, minimum or mode of the values of each variable in each time window; 5 10 fifteen twenty 25 30 35 - Compare the value of each statistical parameter in the current time window with thresholds previously calculated by the monitoring device (1), based on at least previous values of said statistical parameters; - If, as a result of said comparison, no anomaly is detected in any of the statistical parameters, recalculate the value of each of the previously calculated thresholds taking into account the value of each statistical parameter in the current time window; - Activate the location sensor timed and obtain information on a current location of the monitoring device (1); Y - Send information of one or more statistical parameters and the current location information to the server (3) through the communication module: - at regular intervals; I - if as a result of said comparison, any anomaly is detected in any of the statistical parameters or, if more than a predetermined time has elapsed since the last sending of information; - The server (3) comprising: - A first communication module configured to communicate with the monitoring devices (1), a second communication module configured to communicate with external data sources (5) and a third communication module configured to communicate with system users; Y - A processor configured to: - Collect information received through the first communication module of one or more of the monitoring devices (1) of the group of living beings; - Collect information received from one or more external data sources (5) through the second communication module; - Generate behavior patterns for each monitoring device (1) based on a statistical processing of the information collected from the monitoring devices (1) of the group of living beings and external data sources; where the generation of behavior patterns for each monitored device (1), is based on at least the information received from more than one monitoring device (1) of the living beings of the group, and where the generation of patterns is performed applying at least one of the following techniques on the information collected: Time series, Logistic regressions, Trees 3. 4 Decision, Support Vector Machines, Bayesian Statistical Analysis, Collective Intelligence Algorithms or Groups; - Upon receiving information sent by a monitoring device (1), determine whether said information sent verifies the patterns of 5 behavior of said monitoring device (1), previously generated and stored by the server (3); - If it is determined that the information received from any of the monitoring devices (1), does not verify the behavior patterns generated for said one of the monitoring devices (1) or the information 10 received from said monitoring device (1) is not within a range of previously defined values, notify at least one user associated with said one of the monitoring devices (1) by means of the third communication module that an anomaly has been detected in the behavior of said one of the monitoring devices (1) . fifteen 10. System according to claim 9 wherein the monitoring devices (1) are electronic devices using Internet of Things Technology with limited data transfer.