Patent Application
20230102979
The present invention relates to an abnormality detecting system.
Generally speaking, a workflow from breeding through marketing of a cow is broadly divided into multiple steps (for example, grass-feeding, fattening, meat-processing, and so forth). Of these, the fattening process is carried out in a certain area called a feedlot, and therefore, a cow is susceptible to a bovine respiratory disease (BRD) or a bovine respiratory disease complex (BRDC).
In this regard, for example, Non-Patent Document 1 below proposes a system for automatically detecting a cow that has actually developed a BRD (or BRDC) among cows infected with the BRD (or BRDC). The system allows earliest identification of the cow that has actually developed the BRD (or BRDC).
In this regard, detecting an early warning of (or some abnormality concerning) a BRD (or BRDC) prior to an onset of the BRD (or BRDC) in a cow that is a monitoring target may reduce various costs required due to the onset of the BRD (or BRDC) or an increase in severity of the BRD (or BRDC).
In one aspect, an object is to provide an abnormality detecting system for detecting an abnormality in an animal that is a monitoring target.
According to an aspect of the present disclosure, an abnormality detecting system includes
a first identification unit configured to identify a state of an animal that is a monitoring target in each time range, based on time-series data from a motion sensor placed on a predetermined portion of the animal that is a monitoring target;
a first calculation unit configured to calculate a transition probability from the state at a predetermined timing of each time range identified by the first identification unit to a next state; and
a determining unit configured to determine that an abnormality of the animal that is a monitoring target is detected when a score calculated based on the transition probability to the next state satisfies a predetermined condition.
An abnormality detecting system for detecting an abnormality in an animal that is a monitoring target can be provided.
Hereinafter, each embodiment will be described with reference to the accompanying drawings. In the present specification and the drawings, for the components having substantially the same functional configurations, the duplicate descriptions are omitted by providing the same reference numerals thereto.
First, the system configuration of the abnormality detecting system and the functional configuration of the server apparatus will be described.
The abnormality detecting system 100 is a system for detecting an early warning of (or some abnormality concerning) a BRD (or BRDC) in any cow in a feedlot before the cow develops the BRD (or BRDC) during a fattening process.
As depicted in
The measuring device 110 is a motion sensor (in the present embodiment, an acceleration sensor) of three dimensions (an X-axis direction, a Y-axis direction, and a Z-axis direction) attached to a predetermined portion of a cow 10 (in the example of
The measuring device 110 measures time-series three-dimensional acceleration data at a predetermined sampling frequency and transmits the data to the gateway device 120.
The gateway device 120 transmits the time-series three-dimensional acceleration data transmitted from the measuring device 110 to the server apparatus 130.
The server apparatus 130 detects an early warning of (or some abnormality concerning) a BRD (or BRDC) in any cow in a feedlot before the cow develops the BRD (or BRDC). An abnormality detecting program is installed in the server apparatus 130, and when the program is executed, the server apparatus 130 functions as a data obtaining unit 131, a reference data calculation unit 132, an analysis unit 133, and an early warning detecting unit 134.
The data obtaining unit 131 stores time-series data of healthy cow among the time-series data representing the three-dimensional acceleration transmitted from the gateway device 120 in the acceleration data storing unit 135.
The reference data calculation unit 132 reads out the time-series data representing the three-dimensional acceleration of healthy cow stored in the acceleration data storing unit 135 and calculates state transition probability data of healthy cow (will be described in detail below). The reference data calculation unit 132 stores the state transition probability data of healthy cow in the reference data storing unit 136 as reference data.
The analysis unit 133 calculates state transition probability data of the cow that is a monitoring target based on time-series data of the cow that is a monitoring target among the time-series data representing the three-dimensional acceleration acquired by the data obtaining unit 131. The analysis unit 133 calculates a score indicating how far the state transition probability data of the cow that is a monitoring target deviated from the state transition probability data of healthy cow (that is, an abnormality of the state transition).
The early warning detecting unit 134 (determining unit) acquires the score calculated by the analysis unit 133 as data indicating the abnormality of the cow that is a monitoring target and determines whether an early warning (or some abnormality) is detected before development of BRD (or BRDC) in the cow that is a monitoring target. When it is determined that an early warning (or some abnormality) has been detected, the early warning detecting unit 134 notifies a user.
In addition, for example, when data indicating abnormality of the cow that is a monitoring target exceeds a predetermined threshold value for a plurality of consecutive days (for example, two days), the early warning detecting unit 134 determines that an early warning (or some abnormality) has been detected.
Next, a hardware configuration of the server apparatus 130 will be described.
The processor 201 includes various arithmetic and logic operation units such as a central processing unit (CPU), a graphics processing unit (GPU), and so forth. The processor 201 reads various programs (for example, an abnormality detecting program, and so forth), writes the programs into the memory 202, and executes the programs.
The memory 202 includes main storage devices such as a read-only memory (ROM), a random access memory (RAM), and the like. The processor 201 and the memory 202 form what is known as a computer, and when the processor 201 executes the various programs read and written onto the memory 202, the computer implements the above-described functions.
The auxiliary storage device 203 stores the various programs and various data used when the various programs are executed by the processor 201. For example, the acceleration data storing unit 135 and the reference data storing unit 136 are implemented in the auxiliary storage device 203.
The I/F device 204 is a connection device that connects a manually-operating device 210, a display device 211, and the server apparatus 130, which are examples of an external apparatus/device. The I/F device 204 receives a manual operation performed on the server apparatus 130 through the manually-operating device 210. The I/F device 204 outputs the result of a process performed by the server apparatus 130 and displays the result through the display device 211.
The communication device 205 is a communication device for communicating with other apparatuses/devices. In the server apparatus 130, the communication device 205 is used to communicate with the gateway device 120 that is another device.
The drive device 206 is a device for setting a recording medium 212. The recording medium 212 is a medium for optically, electrically, or magnetically recording information, such as a CD-ROM, a flexible disk, a magneto-optical disk, or the like. The recording medium 212 may be a semiconductor memory or the like that electrically records information, such as a ROM, a flash memory, or the like.
The various programs installed in the auxiliary storage device 203 are installed, for example, as a result of a distributed recording medium 212 being set in the drive device 206 and the various programs recorded in the recording medium 212 being read out by the drive device 206. Alternatively, the various programs installed in the auxiliary storage device 203 may have been installed by downloading the various programs from a network via the communication device 205.
Next, a specific example of the process of the data obtaining unit 131 among the process of the units executed by the server apparatus 130 will be described.
The data obtaining unit 131 stores a set of time-series data for healthy cow in the acceleration data storing unit 135.
Next, a functional configuration of the reference data calculation unit 132 among the functional configurations of the units implemented by the server apparatus 130 will be described in detail.
The standardization processing unit 401 reads out the set of time-series data for healthy cow from the acceleration data storing unit 135, performs the standardization process for the time-series data in each axis, and notifies the labeling unit 402 of the standardized data.
The standardization process refers to a process of dividing the time-series data of each axis by a predetermined time range (for example, 1 [SEC]) and calculating a variation of the time-series data of each axis in each time range. That is, the standardized data is composed of calculation results of the variation of the time-series data of each axis in each time range.
The labeling unit 402 performs a labeling process on the standardized data. In the labeling unit 402, the labeling process is performed, i.e.,
when the calculation result of the variation of the time-series data of each axis in each time range included in the standardized data is −4σ or less, “1” is assigned to the standardized data;
when the variation is greater than −4σ and less than or equal to −3σ, “2” is assigned;
when the variation is greater than −3σ and less than or equal to −2σ, “3” is assigned;
when the variation is greater than −2σ and less than or equal to −1σ, “4” is assigned;
when the variation is greater than −1σ and less than +1σ, “5” is assigned;
when the variation is greater than or equal to +1σ and less than +2σ, “6” is assigned;
when the variation is greater than or equal to +2σ and less than +3σ, “7” is assigned;
when the variation is greater than or equal to +3σ and less than +4σ, “8” is assigned; and
when the variation is greater than or equal to +4σ, “9” is assigned (see reference numeral 411).
The state identification unit 403 identifies a state of healthy cow in each time range based on the labeling data generated by performing the labeling process on the time-series data of each axis in each time range. For example, in each time range, when the result of the labeling process for the time-series data in the X-axis direction (LX), the result of the labeling process for the time-series data in the Y-axis direction (LY), and the result of the labeling process for the time-series data in the Z-axis direction (LZ),
are (LX=1, LY=1, LZ=1) respectively, the state of the cow is identified as state I;
are (LX=1, LY=1, LZ=2) respectively, the state of the cow is identified as state II;
. . .
are (LX=9, LY=9, LZ=9) respectively, the state of the cow is identified as state DCCXXIX (see reference numeral 412).
The state transition probability calculation unit 404 calculates a transition probability indicating to which state a state of healthy cow at a predetermined timing of each time range is changed in the next time range. For example, when the state of the cow 10 in the current time range is “state I”, the state of the cow 10, in the next time range, transitions to one of 729 states of “state I” to “state DCCXXIX”. The state transition probability calculation unit 404 counts states transitioned from “state I” in the next time range for each state of the transition destination for a certain period of time. The state transition probability calculation unit 404 divides a number of transitions to each state by the total number of transitions to any state. Accordingly, the state transition probability calculation unit 404 calculates the transition probability from the current state to the next state, for example, for the cow 10.
The state transition probability calculation unit 404 stores the state transition probability data generated by calculating the transition probabilities for, for example, all states of the cow 10 as the transition source in the reference data storing unit 136 as reference data.
As is obvious from the above description, it can be said that the state transition probability data generated by the state transition probability calculation unit 404 is data indicating a tendency of movement of healthy cow.
Next, a specific example of the process of the reference data calculation unit 132 among the process of the units executed by the server apparatus 130 will be described.
As shown in
As shown in
In the example shown in
Also, as shown in
For example, in column 541, for the data for a certain period of time indicating states in each time range of healthy cow, the transition probability from “state I” to “state I” calculated by:
counting a number of transitions (individual number of transitions) in which the current state in the time range is “state I” and the state in the next time range remains “state I”;
counting the total number of transitions, in which the current state in the time range is “state I”, and the state in the next time range is transitioned to any state from “state I” to “state DCCXXIX”; and
dividing individual transition probabilities by the total number of transitions, is stored.
Similarly, in column 542, for the data for a certain period of time indicating states in each time range of healthy cow, the transition probability from “state I” to “state II” calculated by:
counting a number of transitions (individual number of transitions) in which the current state in the time range is “state I” and the state in the next time range is transitioned to “state II”;
counting the total number of transitions, in which the current state in the time range is “state I”, and the state in the next time range is transitioned to any state from “state I” to “state DCCXXIX”; and
dividing individual transition probabilities by the total number of transitions, is stored.
Hereinafter, the same process is performed for the transition destination states of from “state III” to “state DCCXXIX”, and for the transition source states of from “state II” to “state DCCXXIX”, so that the state transition probability data 540 of healthy cow is generated. Incidentally, as shown in
The state transition probability data 540 generated for healthy cow is stored in the reference data storing unit 136 as reference data.
Next, the functional configuration of the analysis unit 133 among the functional configurations of units implemented by the server apparatus 130 will be described in detail.
Functions of the respective units of the standardization processing unit 601 to the state transition probability calculation unit 604 are the same as those of the respective units of the standardization processing unit 401 to the state transition probability calculation unit 404 of the reference data calculation unit 132 described with reference to
The score calculation unit 605 acquires the state transition probability data of the cow that is a monitoring target generated by the state transition probability calculation unit 604. The score calculation unit 605 reads out the state transition probability data of healthy cow from the reference data storing unit 136. The score calculation unit 605 calculates, based on the state transition probability data of the cow that is a monitoring target and the state transition probability data of healthy cow, a score using the following equation (1),
score=−log(P∘X). (Equation 1)
In equation (1) above, P represents the state transition probability data of healthy cow, and X represents the state transition probability data of the cow that is a monitoring target. Furthermore, ∘ indicates an Amadal product. Based on the above equation (1), the score calculation unit 605 calculates for each transition probability a score indicating how far the state transition probability data of the cow that is a monitoring target deviates from the state transition probability data of healthy cow (that is, an abnormality of the state transition).
In addition, the score calculation unit 605 extracts the maximum score from among all of the scores calculated for each of the transition probabilities, and outputs it as data indicating the abnormality of the cow that is a monitoring target at the relevant date.
As is obvious from the above description, the state probability data generated by the state transition probability calculation unit 604 can be said to be data indicating a tendency of movement of the cow that is a monitoring target. In this embodiment, under the assumption that a change occurs in the tendency of movement of the cow that is a monitoring target before the development of BRD (or BRDC), the above score is calculated to detect the change, in contrast to the tendency of movement of healthy cow.
Next, a specific example of the process of the analysis unit 133 among the process of the units executed by the server apparatus 130 will be described.
As shown in
a set of time-series data 710,
standardized data 720, and
labeling data 730,
which are acquired or generated for the cow that is a monitoring target by the standardization processing unit 601, the labeling unit 602, the state identification unit 603, and the state transition probability calculation unit 604, are the same as
the set of time-series data 510,
the standardized data 520, and
the labeling data 530,
which are acquired or generated for healthy cow by the standardization processing unit 401, the labeling unit 402, the state identification unit 403, and the state transition probability calculation unit 404. Thus, the detailed description will be omitted here.
As shown in
The score calculation unit 605 reads out the state transition probability data 540 for healthy cow and uses the above equation (1) to calculate the score for each transition probability, as shown in
As shown in
Next, a specific example of the process of the early warning detecting unit 134 among the process of the units executed by the server apparatus 130 will be described.
In addition, the graph 800 shows that, in the case of the date indicated by the reference numeral 801, since the data indicating the abnormality is greater than or equal to the threshold only in one day and is not consecutive, the early warning detecting unit 134 does not determine that an early warning (or some abnormality) has been detected.
On the other hand, the graph 800 shows that, in the case of the date indicated by the reference numeral 803, since the data indicating the abnormality is greater than or equal to the threshold value for consecutive two days, the early warning detecting unit 134 determines that an early warning (or some abnormality) has been detected.
Next, a flow of the reference data calculation process performed by the server apparatus 130 will be described.
In step S901, the data obtaining unit 131 acquires time-series data indicating the three-dimensional acceleration of healthy cow.
In step S902, the standardization processing unit 401 of the reference data calculation unit 132 performs the standardization process for the time-series data representing the three-dimensional acceleration of healthy cow and generates the standardized data.
In step S903, the labeling unit 402 of the reference data calculation unit 132 performs the labeling process for the standardized data, and generates labeling data.
In step S904, the state identification unit 403 of the reference data calculation unit 132 identifies the state of each time range of healthy cow based on the labeling data.
In step S905, the state transition probability calculation unit 404 of the reference data calculation unit 132 generates state transition probability data of healthy cow based on data indicating the state of each time range of healthy cow.
In step S906, the state transition probability calculation unit 404 of the reference data calculation unit 132 stores the generated state transition probability data of healthy cow in the reference data storing unit 136 as the reference data.
In step S907, the state transition probability calculation unit 404 of the reference data calculation unit 132 determines whether the state transition probability data for a certain period of time of healthy cow has been generated. When it is determined in step S907 that the data is not generated (in the case of NO in step S907), the process returns to step S901.
On the other hand, when it is determined in step S907 that the data is generated (in the case of YES in step S907), the reference data calculation process ends.
Next, the flow of the analysis process by the server apparatus 130 will be described.
In step S1001, the data obtaining unit 131 acquires time-series data representing the three-dimensional acceleration of the cow that is a monitoring target.
In step S1002, the standardization processing unit 601 of the analysis unit 133 performs the standardization process for the time-series data representing the three-dimensional acceleration of the cow that is a monitoring target, and generates the standardized data.
In step S1003, the labeling unit 602 of the analysis unit 133 performs the labeling process for the standardized data and generates the labeling data.
In step S1004, the state identification unit 603 of the analysis unit 133 identify the state of each time range of healthy cow based on the labeling data.
In step S1005, the state transition probability calculation unit 604 of the analysis unit 133 generates the state transition probability data of the cow that is a monitoring target based on data indicating the state of the cow that is a monitoring target in each time range.
In step S1006, the state transition probability calculation unit 604 of the analysis unit 133 determines whether one-day state transition probability data of the cow that is a monitoring target has been generated. When it is determined in step S1006 that the one-day state transition probability data has not been generated (in the case of NO in step S1006), the process returns to step S1001.
On the other hand, when it is determined in step S1006 that the one-day state transition probability data has been generated (in the case of YES in step S1006), the process proceeds to step S1007.
In step S1007, the score calculation unit 605 of the analysis unit 133 acquires the one-day state transition probability data of the cow that is a monitoring target, and reads out the state transition probability data of healthy cow as reference data from the reference data storing unit 136.
In Step S1008, the score calculation unit 605 of the analysis unit 133 calculates a score for each transition probability based on the one-day state transition probability data of the cow that is a monitoring target and the state transition probability data of healthy cow read as reference data. Further, the score calculation unit 605 of the analysis unit 133 extracts the maximum score from among the scores calculated for each of the transition probabilities, and outputs it as data indicating the abnormality of the cow that is a monitoring target at the relevant date.
Next, the flow of the early warning detecting process by the server apparatus 130 will be described.
In step S1101, the early warning detecting unit 134 acquires data representing the abnormality of the cow that is a monitoring target, which is output every day from the score calculation unit 605 of the analysis unit 133.
In step S1102, the early warning detecting unit 134 determines whether the acquired data representing abnormality is greater than or equal to a predetermined threshold value. If it is determined in step S1102 that the acquired data representing abnormality is greater than or equal to the predetermined threshold value (in the case of YES in step S1102), the process proceeds to step S1103.
In step S1103, the early warning detecting unit 134 increments a counter i which counts the number of consecutive days. It is assumed that “0” is input as the initial value in counter i.
In step S1104, the early warning detecting unit 134 determines whether the counter i is greater than or equal to a predetermined number (for example, two days or more). When it is determined in step S1104 that the counter i is greater than or equal to the predetermined number (in the case of YES in step S1104), the process proceeds to step S1105.
In step S1105, the early warning detecting unit 134 determines that an early warning (or some abnormality) has been detected, and notifies the user thereof.
On the other hand, when it is determined in step S1104 that the counter i is less than the predetermined number (in the case of NO in step S1104), the process proceeds to step S1007.
When it is determined in step S1102 that the acquired data is less than the threshold value (in the case of NO in step S1102), the process proceeds to step S1106, “0” is input into the counter i, and the process proceeds to step S1107.
In step S1107, it is determined that the early warning detecting unit 134 does not detect an early warning (or some abnormality).
In step S1108, the early warning detecting unit 134 determines whether the early warning detecting process ends, and returns to step S1101 when it is determined that the early warning detecting process is to be continued (in the case of NO in step S1108). On the other hand, in Step S1108, when it is determined that the early warning detecting process is to be ended (in the case of YES in Step S1108), the early warning detecting process ends.
As is obvious from the above description, the abnormality detecting system according to the first embodiment performs the following process.
States of healthy cow in each time range are identified on the basis of time-series data representing a three-dimensional acceleration as measured by an acceleration sensor placed on a neck of healthy cow. State transition probability data is generated by calculating a transition probability from a state at a predetermined timing of each identified time range to a next state.
States of cow that is a monitoring target in each time range are identified based on time-series data representing a three-dimensional acceleration as measured by an acceleration sensor placed on a neck of the cow that is a monitoring target. State transition probability data is generated by calculating the transition probability from a state at a predetermined timing of the identified time range to the next state.
Based on the state transition probability data generated for healthy cow and the state transition probability data calculated for the cow that is a monitoring target, a score representing an abnormality of a state transition is calculated for each transition probability.
The maximum score is extracted from the score calculated for each transition probability, and is output as data representing the abnormality.
It is determined that the abnormality in the cow that is a monitoring target is detected, when the data representing the abnormality satisfies a prescribed condition.
Thus, in the abnormality detecting system according to the first embodiment, a score is calculated based on the state transition probability data of the cow that is a monitoring target and healthy cow to detect a change in a tendency of movement occurring prior to development of BRD (or BRDC) in contrast to the tendency of movement of healthy cow. Therefore, by the abnormality detecting system according to the first embodiment, it is possible to detect an early warning (or some abnormality) before the cow that is a monitoring target develops BRD (or BRDC).
That is, according to the first embodiment, an abnormality detecting system for detecting an abnormality of cow that is a monitoring target can be provided.
In the first embodiment described above, the score was calculated based on the state transition probability data of the cow that is a monitoring target and the state transition probability data of healthy cow.
However, the method for calculating the score is not limited, and, for example, the score may be calculated based on the state transition probability data of the cow that is a monitoring target. In other words, rather than detecting a change in the tendency of movement based on the contrast to the state transition probability data of healthy cow, a change in the tendency of movement may be detected based on the state transition probability data of only the cow that is a monitoring target. Hereinafter, the second embodiment will be described focusing on the differences from the first embodiment described above.
First, a system configuration of the abnormality detecting system and a functional configuration of the server apparatus will be described.
As shown in
As illustrated in
The analysis unit 1211 generates state transition probability data of the cow that is a monitoring target based on the time-series data of the cow that is a monitoring target among the time-series data representing the three-dimensional acceleration acquired by the data obtaining unit 131. The analysis unit 1211 calculates the score indicating
a degree of increase in each transition probability of the cow that is a monitoring target, or
a degree of decrease in each transition probability of the cow that is a monitoring target, based on the state transition probability data of the cow that is a monitoring target (i.e., the abnormality in the state transition).
Next, the functional configuration of the analysis unit 1211 will be described in detail.
The functions of the respective units of from the standardization processing unit 601 to the state transition probability calculation unit 604 are the same as the functions of the units of from the standardization processing unit 401 to the state transition probability calculation unit 404 of the reference data calculation unit 132 described with reference to
The score calculation unit 1301 acquires the state transition probability data of the cow that is a monitoring target generated by the state transition probability calculation unit 604. The score calculation unit 605 calculates the score using the following equation (2) based on the state transition probability data of the cow that is a monitoring target,
score=−log(X). (Equation 2)
In equation (2) above, X represents the state transition probability data of the cow that is a monitoring target.
The score calculation unit 1301 calculates a score indicating the degree of increase or decrease in each transition probability included in the state transition probability data of the cow that is a monitoring target for each transition probability.
Furthermore, the score calculation unit 1301 extracts the maximum score from among the scores calculated for each of the transition probabilities, and outputs the maximum score as data indicating the abnormality of the cow that is a monitoring target at the relevant date.
Next, a specific example of the process of the analysis unit 1211 among the process of the units executed by the server apparatus 1210 will be described.
As shown in
a set of time-series data 710,
standardized data 720,
labeling data 730,
initial state transition probability data 740, and
one-day state transition probability data 740′,
which are acquired or generated for the cow that is a monitoring target by the standardization processing unit 601, the labeling unit 602, the state identification unit 603, and the state transition probability calculation unit 604, are the same as
the set of time-series data 510,
the standardized data 520,
the labeling data 530,
which are acquired or generated for healthy cow by the standardization processing unit 401, the labeling unit 402, the state identification unit 403, and the state transition probability calculation unit 404 obtain or generate for healthy cow, or
the initial state transition probability data 740, and
the one-day state transition probability data 740′,
which are acquired or generated for the cow that is a monitoring target by the state transition probability calculation unit 604. Thus, the detailed description will be omitted here.
Also in
By generating the one-day state transition probability data 740′ of the cow that is a monitoring target, the score calculation unit 1301 calculates the score for each transition probability using the above equation (2) as shown in
As shown in
Next, the flow of the analysis process by the server apparatus 1210 will be described.
Among the process shown in
In step S1501, the score calculation unit 1301 of the analysis unit 1211 calculates the score for each transition probability based on the one-day state transition probability data of the cow that is a monitoring target. The score calculation unit 1301 of the analysis unit 1211 extracts the maximum score from among the scores calculated for each of the transition probabilities, and outputs it as data representing the abnormality of the cow that is a monitoring target at the relevant date.
As is obvious from the above description, the abnormality detecting system according to the second embodiment performs the following process.
States in each time range of cow that is a monitoring target is identified based on time-series data representing a three-dimensional acceleration as measured by an acceleration sensor placed on a neck of the cow that is a monitoring target. State transition probability data is generated by calculating a transition probability from a state at a predetermined timing of an identified time range to a next state.
Based on the state transition probability data generated for the cow that is a monitoring target, a score representing an abnormality of a state transition is calculated for each stage transition probability.
The maximum score is extracted from the score calculated for each transition probability, and is output as data indicating the abnormality.
It is determined that the abnormality in the cow that is a monitoring target is detected, when the data representing the abnormality satisfies a prescribed condition.
Thus, in the abnormality detecting system in accordance with the second embodiment, a score is calculated based on the state transition probability data of the cow that is a monitoring target to detect a change in a tendency of movement occurring prior to development of BRD (or BRDC). Therefore, by the abnormality detecting system according to the second embodiment, it is possible to detect an early warning (or some abnormality) of BRD (or BRDC) before the cow that is a monitoring target develops BRD (or BRDC) as in the first embodiment.
That is, according to the second embodiment, an abnormality detecting system for detecting an abnormality of cow that is a monitoring target can be provided.
In the first and second embodiments described above, an acceleration sensor is placed on the neck portion, but the attachment part of the acceleration sensor is not limited to the neck portion. The acceleration sensor may be placed on other parts.
Although in the above-described first and second embodiments described above an acceleration sensor is attached as a motion sensor, a motion sensor other than the acceleration sensor (e.g., an angular velocity sensor) may be attached.
In the first embodiment described above, the state transition probability data for healthy cow was described as generating state transition probability data for a certain period of time. However, the period may be one day or a plurality of days.
The state transition probability data for healthy cow may be obtained by generating state transition probability data for one healthy cow, or by generating state transition probability data for a plurality of cows and calculating an average of the data.
In the above-described first and second embodiments, it was described that data indicating abnormalities was output once every day for the cow that is a monitoring target, and it was determined whether an early warning (or some abnormality) was detected once every day. However, the output frequency for outputting data indicating abnormality and the decision frequency for determining whether an early warning (or some abnormality) is detected are not limited to one day, but may be less than one day or greater than or equal to one day.
In the above-described first and second embodiments, the case in which time-series data representing a three-dimensional acceleration was used was described. However, the number of axial directions using time-series data is not limited to three axes. For example, time-series data representing an acceleration of any two axes may be used. Alternatively, time-series data representing only a single-axis acceleration may be used.
In the above-described first and second embodiments, it was described that nine types of labels “1” to “9” were used for the labeling process. However, the types of labels used for the labeling process are not limited to nine types.
In the above-described first and second embodiments, the time range at the time of the standardization process is set to “1 [SEC]”, but the time range at the time of the standardization process is not limited to “1 [SEC]”.
In the above-described first and second embodiments, it was described that when the data indicating the abnormality was greater than or equal to the threshold value for two consecutive days, it was determined that an early warning (or some abnormality) was detected. However, the conditions for determining whether an early warning (or some abnormality) is detected are not limited thereto, and other predetermined conditions may be used to perform the determination. In other words, it may be determined based on whether pre-defined conditions are satisfied or not.
In the above-described first and second embodiments, it was described that the animals that is a monitoring target were cow, but animals other than cow may be monitored.
It should be noted that the present invention is not limited to the above-described embodiments. An embodiment where another element is used in a combination, for example, may be covered by the present invention. In these respects, it is possible to change the embodiment in any way without departing from the spirit of the invention, and how to change can be appropriately determined in accordance with what the present invention is applied to.
The present application claims priority to Japanese Patent Application No. 2020-040117, filed Mar. 9, 2020, and the entire contents of Japanese Patent Application No. 2020-040117 are incorporated herein by reference in its entirety.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2020-040117 | Mar 2020 | JP | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/JP2021/008052 | 3/3/2021 | WO |
