Remote monitoring method and system for infant raising

Abstract

The invention relates to the technical field of nursing remote monitoring management, in particular to a remote monitoring method and system for infant nursing, and the method comprises the steps: obtaining a nursing rule containing a triggering condition event and a response event sequence, collecting and analyzing behavior and environment information into an actual event flow with a time identifier, and transmitting the actual event flow to a server; the method comprises the steps of detecting whether a trigger condition event is met or not, generating a trigger event instance, detecting a response event sequence in an actual event stream based on the trigger event instance, and if matching succeeds, generating a verification record containing the trigger condition event, the response event sequence and the occurrence time and the spatial position of the trigger condition event and the response event sequence. The method has the advantage that more complete and accurate information in the nursing data can be analyzed and extracted.

Description

Technical Field

[0001] This application relates to the technical field of remote monitoring and management of care, and more specifically, to a method and system for remote monitoring of infant and toddler care. Background Technology

[0002] In the field of infant and toddler care, institutions generally develop personalized care plans that incorporate complex conditional judgment logic in order to improve service quality and meet the differentiated developmental needs of infants and toddlers.

[0003] However, these complex plans pose challenges to caregivers' professional judgment and execution, and also increase the difficulty for parents to understand. Although remote monitoring systems provide parents with video observation, the cameras record decontextualized images, failing to demonstrate the dynamic decisions made by caregivers based on the plan's conditions and logic. They only present the care "outcome" without the "antecedent" information, easily leading to misunderstandings and questions about service quality from parents. Current technology cannot verify whether behavior conforms to the plan's logic or generate information including the causes and consequences of the behavior, resulting in incomplete and inaccurate care information.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for remote monitoring of infant and toddler care, which has the advantage of being able to analyze and extract more complete and accurate information from care data.

[0006] This application provides a method for remote monitoring of infant and toddler care, the method including: Obtain preset care rules, which include at least one triggering condition event and at least one sequence of response events corresponding to the triggering condition event; Collect behavioral and environmental information during infant and toddler care, and parse the behavioral and environmental information into a time-stamped actual event stream, which includes object identification, event type, occurrence time and spatial location; Detect whether the actual event stream meets the triggering condition event. If a real-time event that meets the triggering condition event is detected, the real-time event is used as the triggering event instance. Based on the trigger event instance, detect whether a sequence of response events corresponding to the trigger event instance occurs in the actual event stream; If a match is successful, a verification record is generated to represent the consistency between the care behavior and the care rule logic. The verification record includes the triggering condition event, the response event sequence, and the corresponding occurrence time and spatial location.

[0007] The above approach can objectively verify whether caregiving behavior conforms to pre-set complex care rules, thereby improving the transparency and traceability of the caregiving process.

[0008] Furthermore, this application also proposes, based on the above-described remote monitoring method for infant and toddler care, detecting whether a sequence of response events corresponding to the trigger event instance occurs in the actual event stream, including: Detect response events in the response event sequence that correspond to the trigger event instance and contain a preset abstract care instruction type, the abstract care instruction type being determined through structured annotation; Detect unacknowledged events that match response events in the actual event stream; Detect confirmation signals, match the temporal and spatial correlation between confirmation signals and events to be confirmed, and associate confirmation signals with corresponding events to be confirmed; If a match is successful, the event to be confirmed is the sequence of response events that occurred in the actual event stream corresponding to the triggering event instance.

[0009] The above scheme refines the detection process of response event sequences and improves matching accuracy by introducing abstract instruction types and acknowledgment signals.

[0010] Furthermore, this application also proposes, according to the above-mentioned remote monitoring method for infant and toddler care, detecting confirmation signals, matching the temporal and spatial correlation between confirmation signals and events to be confirmed, so that the confirmation signals are associated with the corresponding events to be confirmed, including: The detection confirmation signal contains a temporal and spatial identifier. Based on the comparison and analysis of the temporal and spatial identifiers with the occurrence time and spatial location of the event to be confirmed, the temporal and spatial correlation between the confirmation signal and the event to be confirmed is obtained; When the temporal and spatial correlations meet the preset temporal and spatial proximity conditions, the confirmation signal and the associated event to be confirmed are determined.

[0011] The above scheme provides a method for determining the correlation between confirmation signals and events to be confirmed based on temporal and spatial proximity, thereby enhancing the reliability of matching.

[0012] Furthermore, this application also proposes that, according to the above-mentioned remote monitoring method for infant and toddler care, the method further includes: When there are multiple caregivers, establish a correspondence between the caregiver identity information of multiple pending events and the caregiver location information. The temporal and spatial identifiers of the confirmation signal are obtained. Based on the spatial distance between all caregiver location information and the location information in the temporal and spatial identifiers, the caregiver identity information whose spatial distance meets the preset spatial threshold is selected as candidates. Based on the corresponding relationship, match the caregiver identity information associated with the event to be confirmed among the candidates, select the corresponding matched candidate as the target caregiver according to the preset matching rules, and update the caregiver identity information of the target caregiver in the verification record.

[0013] The above solution solves the problem of behavior attribution in multi-caregiver scenarios and can accurately identify caregivers performing specific care behaviors.

[0014] Furthermore, this application also proposes, based on the aforementioned remote monitoring method for infant and toddler care, to preset matching rules, including: Based on the location information of the candidate's caregiver, obtain the candidate's movement trajectory information between the time of the event to be confirmed and the time of the confirmation signal reception; By combining the movement trajectory information, the occurrence time and spatial location of the event to be confirmed, and the temporal and spatial identifiers of the confirmation signal, the spatial trajectory characteristics of each candidate are calculated. By comparing the degree of deviation from the expected behavioral path in the spatial trajectory features and the continuity of the position in the temporal and spatial markers of the arrival of the confirmation signal, a behavioral consistency metric value for each candidate is calculated. The target caregiver is determined based on the consistency metric.

[0015] The above scheme provides a refined matching rule for determining target caregivers based on caregiver movement trajectories and behavioral consistency metrics, further improving the accuracy of attribution.

[0016] Furthermore, this application also proposes, according to the above-mentioned remote monitoring method for infant and toddler care, detecting confirmation signals, matching the temporal and spatial correlation between confirmation signals and events to be confirmed, so that the confirmation signals are associated with the corresponding events to be confirmed, including: Determine the duration of the confirmation signal; When the duration is less than the preset first threshold, the confirmation signal is parsed as a single confirmation signal, and the temporal and spatial correlation between the confirmation signal and the event to be confirmed is matched so that the confirmation signal is associated with the corresponding event to be confirmed. When the duration is greater than or equal to the first threshold, all events to be confirmed that exist within the preset time window and match the temporal and spatial correlation of the confirmation signal are identified as batch confirmation events, and the confirmation signal is parsed as a batch confirmation signal of the batch confirmation events.

[0017] The above scheme distinguishes between single confirmation and batch confirmation scenarios, improving the system's flexibility and efficiency in handling different types of confirmation signals.

[0018] Furthermore, this application also proposes that, according to the above-mentioned remote monitoring method for infant and toddler care, when the duration is greater than or equal to a first threshold, all events to be confirmed that exist within a preset time window and match the temporal and spatial correlation of the confirmation signal are acquired as batch confirmation events. After parsing the confirmation signal as a batch confirmation signal of batch confirmation events, the method further includes: Retrieve the caregiver's identity information and corresponding caregiver function area corresponding to the batch confirmation event; If there are multiple caregiver identity information from adjacent caregiver functional areas in the batch confirmation event, then based on the movement behavior information of each caregiver across each caregiver functional area within a preset time window, the dwell time and dwell location of each caregiver in each caregiver functional area can be obtained. Based on the dwell time and location, determine the caregivers and associated functional areas corresponding to each batch of confirmed events, and update the verification records.

[0019] By using the above scheme, for batch confirmation scenarios, by analyzing the caregiver's cross-regional movement and stay information, it is possible to more accurately identify the executors of batch actions and their associated functional areas.

[0020] Furthermore, this application also proposes that, according to the above-mentioned remote monitoring method for infant and toddler care, the method further includes: When batch confirmation events involve different categories of care rules, obtain the event type for each batch confirmation event; Based on the event type and the duration of the acknowledgment signal, a preset category priority rule is applied; Based on the category priority rule, batch confirmation events of different event types are grouped for confirmation. The batch confirmation signal is parsed into multiple sub-batch confirmation signals, which are matched to the events to be confirmed for the corresponding event types, and the corresponding verification records are updated.

[0021] The above solution introduces priority rules and group confirmation mechanisms for batch confirmation involving multiple types of care rules, thereby improving the accuracy and efficiency of batch confirmation in complex scenarios.

[0022] Furthermore, this application also proposes that, according to the above-mentioned remote monitoring method for infant and toddler care, batch confirmation events of different event types are grouped and confirmed according to category priority rules. The batch confirmation signal is parsed into multiple sub-batch confirmation signals, which are then matched to the events to be confirmed corresponding to the respective event types. After updating the corresponding verification records, the method further includes: Obtain the sequence of task execution behaviors for the corresponding caregiver in each event type that needs to be confirmed; An event chain model is established based on the temporal logic of continuous actions in the task execution sequence and the occurrence time and spatial location of each event to be confirmed. Based on the event chain model, the consistency between the caregiver's execution logic and the care rules is verified, and the verification record is updated.

[0023] By constructing an event chain model, the above approach allows for in-depth analysis of the caregiver's behavioral sequence in performing tasks, further verifying the degree to which their execution logic conforms to care rules.

[0024] Furthermore, this application also proposes a remote monitoring system for infant and toddler care, comprising: The care rule storage module is used to obtain preset care rules, which include at least one triggering condition event and at least one response event sequence corresponding to the triggering condition event. The data acquisition module is used to collect behavioral and environmental information during infant care and to parse the behavioral and environmental information into a stream of actual events with time stamps. The actual event stream includes object identifier, event type, occurrence time and spatial location. The detection module is used to detect whether the actual event stream meets the triggering conditions. If a real-time event that meets the triggering conditions is detected, the real-time event is used as the triggering event instance. The judgment module is used to detect whether a sequence of response events corresponding to the trigger event instance has occurred in the actual event stream, based on the trigger event instance. The verification module is used to generate a verification record that represents the consistency between the care behavior and the care rule logic if the match is successful. The verification record includes the triggering condition event, the response event sequence, and the corresponding occurrence time and spatial location.

[0025] The above scheme provides a system for implementing the above method, providing hardware and software support for the practical application of the method.

[0026] As can be seen from the above, the remote monitoring method and system for infant and toddler care provided in this application solves the problem of insufficient trust caused by incomplete information in existing remote monitoring by transforming abstract personalized care plans into detectable event sequences and comparing and verifying them with actual behavioral event flows. It has the advantages of being able to objectively verify whether care behaviors conform to complex care rules and being able to analyze and extract more complete and accurate information from care data. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating a remote monitoring method for infant and toddler care, provided as one embodiment of this application.

[0028] Figure 2 This is one of the flowcharts illustrating a remote monitoring method for infant and toddler care, provided as another embodiment of this application.

[0029] Figure 3This is a second flowchart illustrating a remote monitoring method for infant and toddler care, provided as another embodiment of this application.

[0030] Figure 4 This is the third flowchart illustrating a remote monitoring method for infant and toddler care, provided as another embodiment of this application.

[0031] Figure 5 This is the fourth flowchart illustrating a remote monitoring method for infant and toddler care, provided as another embodiment of this application.

[0032] Figure 6 The fifth flowchart illustrates a remote monitoring method for infant and toddler care, provided as another embodiment of this application.

[0033] Figure 7 This is a flowchart of another embodiment of the present application of a remote monitoring method for infant and toddler care.

[0034] Figure 8 This is the seventh flowchart illustrating a remote monitoring method for infant and toddler care, provided as another embodiment of this application.

[0035] Figure 9 This is the eighth flowchart illustrating a remote monitoring method for infant and toddler care, provided as another embodiment of this application.

[0036] Figure 10 A flowchart of a remote monitoring system for infant and toddler care provided in another embodiment of this application.

[0037] In the diagram: 1. Care rule storage module; 2. Data acquisition module; 3. Detection module; 4. Judgment module; 5. Verification module. Detailed Implementation

[0038] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0039] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0040] Traditional remote monitoring systems for infant and toddler care suffer from a lack of clarity regarding the decision-making processes behind care behaviors. When care behaviors need to follow personalized care plans with conditional branches, relying solely on video clips captured by fixed-location cameras cannot fully record the preconditions that trigger specific care behaviors or other concurrent events. This can lead to discrepancies between the behavior observed by the user through remote monitoring and the literal description of the pre-set plan, resulting in misunderstandings about the quality of professional care. This incompleteness of information makes it difficult for users to understand the judgments and choices made by caregivers based on the plan's logic in complex situations.

[0041] Reference Figure 1 In response, this application proposes a method for remote monitoring of infant and toddler care, including: S1000: Obtain preset care rules, which include at least one triggering condition event and at least one response event sequence corresponding to the triggering condition event; S2000: Collects behavioral and environmental information during infant and toddler care and parses the behavioral and environmental information into a stream of actual events with time stamps. The actual event stream includes object identifier, event type, occurrence time and spatial location. S3000: Detects whether the actual event stream meets the triggering condition event. If a real-time event that meets the triggering condition event is detected, the real-time event is used as the triggering event instance. S4000: Based on the trigger event instance, detect whether a sequence of response events corresponding to the trigger event instance has occurred in the actual event stream; S5000: If the match is successful, a verification record is generated to represent the consistency between the care behavior and the care rule logic. The verification record includes the triggering condition event, the response event sequence, and the corresponding occurrence time and spatial location.

[0042] In this embodiment, care rules refer to pre-defined norms guiding infant and toddler care behaviors. These rules include at least one triggering event and at least one corresponding sequence of response events. They can be represented using structured text, rule engines, or expert systems, providing a set of verifiable care behavior standards. A triggering event is a prerequisite for initiating the execution of the response event sequence within the care rules. It can be a specific behavioral state of the infant or toddler, a change in the environment, or a specific point in time, primarily used to identify scenarios requiring special attention or intervention. A response event sequence refers to a series of prescribed actions or behaviors that the caregiver should perform after the triggering event is met. It can be a series of continuous or discontinuous care operations, primarily used to ensure that compliant care measures are taken in specific scenarios. Behavioral and environmental information refers to raw data collected during infant and toddler care through various sensors, cameras, and other devices. This data can include the infant's movements, expressions, sounds, and environmental information such as light and temperature, primarily used to obtain raw, objective records of the actual care process.

[0043] A real-world event stream refers to a time-stamped data sequence formed after parsing and structuring the collected behavioral and environmental information. It includes object identifiers, event types, occurrence times, and spatial locations. Its primary purpose is to transform raw data into a machine-readable, processable, and comparable data format. An object identifier is a tag used to uniquely identify individuals or items involved in the real-world event stream; it can be represented by a unique numerical code or name. An event type is a label that classifies and describes the behaviors or states occurring in the real-world event stream; it can be represented by predefined behavior or state categories, primarily to standardize the description of the events. The occurrence time is the timestamp of each event in the real-world event stream, which can be represented by time information accurate to the second or millisecond. The spatial location refers to the geographical location or area information of each event in the real-world event stream, which can be represented by coordinates, area codes, or location tags. A trigger event instance is a real-time event detected in the real-world event stream that meets preset trigger conditions. It is a specific event record that actually occurred and meets the trigger conditions, primarily used to mark the starting point for initiating the detection of subsequent response event sequences. Verification records are records generated after comparing caregiving behaviors with the logical consistency of care rules. They include triggering events, response event sequences, and the corresponding time and location of occurrence. Their main purpose is to provide objective evidence of whether caregiving behaviors comply with or do not comply with care rules.

[0044] This solution first acquires preset care rules and defines expected care behavior patterns as verification standards. It then collects and structures behavioral and environmental information during infant and toddler care, forming an actual event flow. The system detects whether real-time events meeting the triggering conditions in the care rules appear in this event flow and marks them as trigger event instances. It then compares whether corresponding response event sequences occur, thereby verifying whether the behavior conforms to the preset specifications. Upon successful comparison, a verification record containing triggering conditions, response event sequences, and spatiotemporal information is generated. This solves the problem of difficulty in understanding behavioral logic based solely on isolated video clips and has the advantage of being able to analyze and extract more complete and accurate information from care data.

[0045] In some preferred embodiments, this application is implemented as follows: A pre-defined care rule states that when an infant or toddler exhibits irritability, the caregiver should soothe them. The system collects behavioral information such as the infant's facial expressions and body movements, as well as environmental information such as ambient sounds, through a camera. This information is parsed into an actual event stream, such as an event containing an infant object identifier, event type "crying," occurrence time T1, and spatial location L1, and an event containing a caregiver object identifier, event type "picking up," occurrence time T2, and spatial location L1. The system detects that the "crying" event at time T1 meets the "irritability" trigger condition and uses it as a trigger event instance. Subsequently, based on this trigger event instance, the system checks in the event stream whether a corresponding "soothing" response event has occurred, such as the "picking up" event at time T2. If the "picking up" event is detected, the response event sequence is considered to have occurred, and the match is successful. At this point, the system generates a verification record, which records the triggering event (infant crying, T1, L1) and the response event (caregiver picking up, T2, L1), indicating that the caregiving behavior complies with the care rules.

[0046] Reference Figure 2 In another embodiment of this application, step S4000 further includes: S4100: Detects response events in the response event sequence that correspond to the trigger event instance and contain a preset abstract care instruction type, the abstract care instruction type being determined through structured annotation; S4200: Detects unacknowledged events that match response events in the actual event stream; S4300: Detects confirmation signals and matches the temporal and spatial correlation between confirmation signals and events to be confirmed, so that the confirmation signal is associated with the corresponding events to be confirmed. S4400: If the match is successful, the event to be confirmed is the sequence of response events that occurred in the actual event stream corresponding to the trigger event instance.

[0047] In this embodiment, the abstract care instruction type refers to a general description of the caregiver's behavioral intentions or task objectives, rather than specific physical actions. It can be implemented using predefined tags, classification systems, or semantic descriptions. Structured annotation refers to the process of defining and storing the abstract care instruction type in a standardized format. It can use data formats such as XML and JSON, or be described based on ontology, knowledge graphs, etc. The event to be confirmed refers to the candidate event detected in the actual event stream that may match the response event. It can be a set of events obtained through preliminary screening based on event type, object, location, etc. The confirmation signal refers to auxiliary information used to verify whether the event to be confirmed is a valid response. It can take the form of the caregiver's voice commands, gestures, specific device operations, or staying in a specific area. The temporal and spatial correlation refers to the relationship between the confirmation signal and the event to be confirmed in terms of time and spatial location. It can be measured by time difference, spatial distance, trajectory overlap, etc.

[0048] This solution first detects response events within a response event sequence that are explicitly defined by structured annotations and abstract care instruction types. This decouples care rules from specific physical actions, improving versatility. Subsequently, it filters out events from the actual event stream that initially match these response events, narrowing down the matching range. By detecting confirmation signals and precisely matching the events to be confirmed based on temporal and spatial correlation conditions, an external verification mechanism is implemented, effectively distinguishing between accidental events and conscious caregiver responses. When the confirmation signal and the event to be confirmed meet preset conditions, the event is confirmed as an actual response event sequence corresponding to the abstract care instruction. This multi-level matching and verification process, combining abstract intent, event filtering, and confirmation signal verification, overcomes the problem in existing technologies where abstract instructions are difficult to accurately match actual behaviors. By utilizing confirmation signals and temporal and spatial correlations, it achieves more reliable judgments, thereby improving the accuracy of consistency verification between abstract care instructions and actual care behaviors and resolving misunderstandings caused by incomplete information.

[0049] In some preferred embodiments, this application is implemented as follows: For example, when the system detects that an infant is crying (as a triggering event instance), and the response event sequence corresponding to the crying in the care rules includes the abstract care instruction type of "soothing emotions," the system searches the actual event stream for events that may be related to "soothing emotions," such as behavioral events like a caregiver picking up the infant or gently patting the infant's back. Simultaneously, the system detects confirmation signals issued by the caregiver, such as the caregiver's voice command "Baby, don't cry, Mommy will hold you," or specific signals emitted by sensors worn by the caregiver when performing soothing actions. The system compares the time and spatial location of this confirmation signal with the time and spatial location of the event to be confirmed (such as picking up the infant). If the confirmation signal and the event to be confirmed are highly proximate in time and space, for example, if the voice command is issued very shortly after the action of picking up the infant, and the positions of the caregiver and the infant are highly overlapping, then the system determines that the event of picking up the infant is the caregiver's response to the crying event. That is, the event to be confirmed is confirmed as an actual occurrence of a response event sequence corresponding to the abstract command of "soothing emotions".

[0050] Reference Figure 3 In another embodiment of this application, step S4300 further includes: S4310: Detection confirmation signal, which includes a time and space identifier; S4320: Based on the temporal and spatial identification and the comparison and analysis of the occurrence time and spatial location of the event to be confirmed, the temporal and spatial correlation between the confirmation signal and the event to be confirmed is obtained; S4330: When the temporal and spatial correlation meets the preset temporal and spatial proximity conditions, determine the confirmation signal and the associated event to be confirmed.

[0051] In this embodiment, the temporal and spatial identifier refers to the information carried in the confirmation signal that reflects the time of signal generation or transmission and the corresponding location. Specifically, it can be a timestamp and a location coordinate or area identifier. The temporal proximity condition refers to a preset rule or threshold used to determine whether the time of the confirmation signal is sufficiently close to the time of the event to be confirmed, such as a time difference less than a certain value. The spatial proximity condition refers to a preset rule or threshold used to determine whether the location of the confirmation signal is sufficiently close to the spatial location of the event to be confirmed, such as a spatial distance less than a certain value or being located within the same preset area. The purpose is to filter out confirmation signals that are not related to the event to be confirmed in time and space by setting time and spatial limitations, thereby improving the accuracy of the correlation judgment.

[0052] This solution detects confirmation signals with temporal and spatial identifiers and compares them with the temporal and spatial locations of the event to be confirmed, analyzing their temporal and spatial proximity to determine the degree of correlation. Based on a precise matching mechanism across both time and space dimensions, it effectively distinguishes whether a confirmation signal is specific to the event to be confirmed, thereby improving the accuracy of the correlation. This mechanism ensures that only confirmation signals truly corresponding to the event to be confirmed are adopted, significantly improving the accuracy of judging whether caregiving behavior conforms to care rules and avoiding misjudgments or omissions due to incorrect correlations.

[0053] In some preferred embodiments, this application is implemented as follows: For example, when the system detects an event to be confirmed, such as "a caregiver changing an infant's diaper," this event occurs at a specific time and in a specific area. After the caregiver completes the operation, they send a confirmation signal through a smart device they are wearing. This confirmation signal is received and parsed, and contains a timestamp of the signal transmission and the smart device's location information at that time. The system compares the timestamp of the confirmation signal with the occurrence time of the "diaper changing" event to calculate the time difference; simultaneously, it compares the location information of the smart device with the location of the area where the "diaper changing" event occurred to calculate the spatial distance. If the time difference is less than a preset time threshold and the spatial distance is less than a preset spatial threshold, then the confirmation signal and the "diaper changing" event are considered to meet the time proximity and spatial proximity conditions, thereby determining that the confirmation signal was issued for the "diaper changing" event to be confirmed.

[0054] Reference Figure 4 In another embodiment of this application, the method further includes: S6000: When there are multiple caregivers, establish a correspondence between the caregiver identity information of multiple pending events and the caregiver location information. S7000: Acquire the temporal and spatial identifier of the confirmation signal, and select caregiver identity information whose spatial distance meets the preset spatial threshold as candidates based on the spatial distance between all caregiver location information and the location information in the temporal and spatial identifier. S8000: Match caregiver identity information associated with the event to be confirmed among the candidates according to the corresponding relationship, select the corresponding matched candidate as the target caregiver according to the preset matching rules, and update the caregiver identity information of the target caregiver in the verification record.

[0055] In this embodiment, the pre-configured correspondence between the caregiver identity information of the event to be confirmed refers to the associated information pre-configured by the system, indicating which caregivers may be responsible for or perform a specific event to be confirmed. This can be implemented using database tables, configuration files, or rule engines, and its purpose is to provide a preliminary scope for subsequent caregiver identity matching. Caregiver identity information refers to the set of information used to uniquely identify a caregiver, including the caregiver's unique identifier, name, and real-time or near-real-time location information. The caregiver location information can be obtained using technologies such as UWB positioning, Bluetooth positioning, Wi-Fi positioning, or location estimation based on visual analysis, and its purpose is to provide the caregiver's location data in physical space. The temporal and spatial identifier of the confirmation signal refers to the timestamp attached to the confirmation signal when it is generated. Spatial location information, which can be generated by handheld devices, wearable devices, or sensors in specific areas used by caregivers when receiving confirmation signals, aims to accurately record the spatiotemporal information of the confirmation signal occurrence; preset spatial thresholds refer to the upper limit of distance used to determine whether the caregiver's location is sufficiently close to the confirmation signal location, which can be set according to the positioning accuracy of the actual deployment environment and the caregiver's operating habits, with the purpose of filtering out irrelevant caregivers who are too far away through spatial location screening; preset matching rules refer to the criteria for determining the final target caregiver among multiple candidates that meet the spatial conditions, based on specific logic, which can include various rules based on the caregiver's responsibility area, current task status, historical behavior patterns, or more refined spatial trajectory analysis, with the purpose of selecting the one that best matches the actual situation from multiple possible caregivers.

[0056] This application's solution establishes a pre-defined correspondence between the event to be confirmed and the caregiver's identity information, providing fundamental association information for caregiver identification. Subsequently, it compares the spatial distance of the confirmation signal's temporal and spatial identifier with all caregiver location information, filtering caregivers whose spatial distance meets a pre-defined threshold as candidates. This effectively utilizes the location information at the time the confirmation signal occurs, excluding irrelevant caregivers located at a distance. Based on this, according to the pre-defined correspondence and more refined matching rules, it further matches and determines the target caregiver associated with the event to be confirmed from the narrowed pool of candidates. It is precisely this step-by-step filtering and multi-dimensional matching mechanism that enables the system to accurately associate the confirmation signal with specific caregiver identity information even when multiple caregivers are present simultaneously. This differs from previous solutions that only temporally and spatially correlate the confirmation signal with the event to be confirmed; this solution further addresses the identification of the sender of the confirmation signal, thereby ensuring the accuracy of caregiver identity information in the verification record and improving the traceability and reliability of the entire monitoring system.

[0057] In some preferred embodiments, specifically, suppose that a confirmation event, "infant feeding completed," occurs in a certain area of ​​a childcare facility. The system is pre-configured so that feeding events in this area during a specific time period are typically handled by caregivers A and B. Therefore, a correspondence is established between this confirmation event and the identity information of caregivers A and B, which includes their real-time location information obtained through an indoor positioning system. When caregiver A finishes feeding, they click the confirmation button next to the infant table using a handheld device. The system obtains the temporal and spatial identifier of this confirmation signal, such as time T1 and location P1. The system simultaneously obtains the real-time location information of caregivers A, B, and C, as well as other caregivers in the area (such as caregiver C). The system calculates the spatial distance between the respective locations of caregivers A, B, and C and location P1. If the distance between caregiver A and caregiver B and P1 is less than a preset spatial threshold (e.g., 1 meter), while the distance between caregiver C and P1 is greater than this threshold, then caregivers A and B are identified as candidates. Next, the system matches candidates A and B according to preset matching rules. These preset matching rules may include "caregiver closest to the confirmation signal location" and "caregiver responsible for the infant." If caregiver A's location is closest to P1, and the system record shows that caregiver A is responsible for the infant, then caregiver A is selected as the target caregiver. Finally, caregiver A's identity information is updated in the verification record for the "infant feeding completed" confirmation event.

[0058] Reference Figure 5 In another embodiment of this application, the preset matching rules include: A1: Obtain the candidate's movement trajectory information between the time of the event to be confirmed and the time of the confirmation signal reception based on the location information of the candidate's caregiver; A2: Calculate the spatial trajectory characteristics of each candidate by combining the movement trajectory information, the occurrence time and spatial location of the event to be confirmed, and the temporal and spatial identifiers of the confirmation signal; A3: Compare the degree of deviation from the expected behavioral path in the spatial trajectory features and the continuity of the position in the temporal and spatial markers of the arrival of the confirmation signal to calculate the behavioral consistency metric of each candidate. A4: Determine the corresponding target caregiver based on the consistency metric.

[0059] In this embodiment, caregiver location information refers to data that characterizes the caregiver's spatial location at a specific moment. This can be achieved using indoor positioning technologies such as Wi-Fi fingerprinting, Bluetooth beacons, UWB positioning, or wearable devices such as smart bracelets or positioning badges to collect location coordinates. Movement trajectory information refers to a sequence of caregiver location information continuously recorded over a period of time. This can be achieved by arranging the continuously collected location coordinates in chronological order and storing them as trajectory data. Spatial trajectory features refer to quantitative indicators extracted from the caregiver's movement trajectory information that describe their spatial behavior patterns within a specific time period. This can be achieved by calculating the trajectory length, direction changes, dwell points, and movement speed distribution. Expected behavior path refers to the typical movement route that the caregiver might take to execute the event and issue a confirmation signal, based on the type and location of the event to be confirmed and the expected location of the confirmation signal. This can be achieved using a pre-set standard movement path model associated with a specific care behavior. Deviation refers to the degree of spatial alignment or difference between the candidate's actual movement trajectory and the expected behavioral path. It can be achieved by calculating the average distance, maximum distance, or area difference between the actual trajectory and the expected path. Positional continuity refers to the smoothness and consistency of the caregiver's movement trajectory in space and time as they approach or reach the confirmation signal location. It can be achieved by analyzing whether the time intervals and spatial distances between trajectory points conform to normal movement patterns. Behavioral consistency metric is a quantitative score representing the strength of the correlation between the candidate's behavior and the event to be confirmed, obtained by comprehensively evaluating the deviation of the candidate's movement trajectory from the expected behavioral path and the continuity of reaching the confirmation signal location. It can be achieved by calculating the value through weighted summation of indicators such as deviation and positional continuity or by using a machine learning model. Target caregiver refers to the caregiver who, among multiple candidates, is ultimately determined to have performed the specific event to be confirmed and issued the confirmation signal by comparing behavioral consistency metrics. It can be achieved by selecting the candidate with the highest behavioral consistency metric.

[0060] This application's solution, through preset matching rules, initially screens candidate caregivers and further obtains their movement trajectory information between the time of the event to be confirmed and the time of the confirmation signal reception based on the caregiver's location information. This movement trajectory information provides a detailed record of the caregiver's spatial behavior during the key time period. Combining this movement trajectory information, the time and spatial location of the event to be confirmed, and the temporal and spatial identifier of the confirmation signal, the system calculates the spatial trajectory characteristics of each candidate. This process integrates scattered temporal and spatial information into quantitative features describing the caregiver's behavioral patterns. Subsequently, the system compares the degree of deviation of these spatial trajectory characteristics from the preset expected behavioral path, as well as the continuity of the caregiver's arrival at the confirmation signal location. The expected behavioral path represents the typical movement pattern for performing the caregiving behavior, while location continuity reflects the fluency and purposefulness of the behavior. By comprehensively considering the degree of deviation and location continuity, the system calculates a behavioral consistency metric for each candidate. This metric intuitively reflects the degree to which each caregiver's actual behavior matches the expected pattern for performing the specific caregiving behavior. Ultimately, the system determines the corresponding target caregiver based on these consistency metrics, typically selecting the caregiver with the highest metric value. This solution, after initially filtering potentially relevant caregivers through spatial distance in the preliminary steps, further utilizes the caregiver's dynamic movement trajectory information, combined with detailed temporal and spatial information of events and confirmation signals. By calculating and comparing spatial trajectory characteristics with expected behavioral paths, and assessing location continuity, a more refined behavioral consistency metric is calculated. This analysis based on dynamic trajectories and behavioral patterns effectively distinguishes multiple caregivers moving within the same area or over a short period, accurately identifying the one who actually performed a specific caregiving action and issued a confirmation signal. This solves the problem that relying solely on static location or simple temporal and spatial correlation cannot accurately distinguish the responsible party, significantly improving the accuracy of identifying the target caregiver in multi-caregiver scenarios, and ensuring that verification records can be accurately linked to the specific executor.

[0061] In some preferred embodiments, this application is implemented as follows: Assume that in a childcare area, infant A needs feeding care. The system detects a pending confirmation event related to feeding and subsequently receives a confirmation signal from the caregiver. At this time, the system identifies three candidates in the area: caregiver A, caregiver B, and caregiver C. All three are within a preset spatial threshold range of the location of the pending confirmation event or the location of the confirmation signal reception. The system first obtains the movement trajectory information of caregivers A, B, and C from the time of the pending confirmation event to the time of the confirmation signal reception. For example, caregiver A's trajectory shows that they moved from near infant A to the wall where the confirmation button is placed; caregiver B's trajectory shows that they were active on the other side of the room; and caregiver C's trajectory shows that they briefly stayed near infant A before moving to another area. Next, the system combines the occurrence time and spatial location of the pending confirmation event (e.g., near infant A's bedside) with the temporal and spatial identifiers of the confirmation signal (e.g., the location and pressing time of the confirmation button on the wall) to calculate the spatial trajectory characteristics of caregivers A, B, and C. For example, the system calculates the trajectory length, direction changes, and relative positional relationships with the infant A's bedside and the confirmation button location. Then, the system compares these spatial trajectory characteristics with the pre-defined expected behavioral path for feeding care. The expected path might be from the infant A's bedside to the confirmation button location. The system calculates the deviation of caregivers A, B, and C's actual trajectories from this expected path and assesses the continuity of their position when reaching the confirmation button on the wall. For example, caregiver A's trajectory deviates little from the expected path and shows continuous stops or precise arrival near the confirmation button location; caregiver B's trajectory deviates significantly from the expected path and is far from the confirmation button location; caregiver C's trajectory may partially approach the expected path, but shows poor continuity at the confirmation button location. Based on these comparison results, the system calculates the behavioral consistency metric for each caregiver. For example, caregiver A has the highest metric, caregiver B has the lowest, and caregiver C's metric is in the middle. Finally, based on the calculated consistency metric, the system identifies caregiver A, who has the highest metric, as the target caregiver who performed the breastfeeding care. The system then associates and updates the verification record of this breastfeeding care with caregiver A's identity information.

[0062] Reference Figure 6 In another embodiment of this application, step S4300 further includes: S4340: Determine the duration of the confirmation signal; S4350: When the duration is less than the preset first threshold, the confirmation signal is parsed as a single confirmation signal, and the temporal and spatial correlation between the confirmation signal and the event to be confirmed is matched, so that the confirmation signal is associated with the corresponding event to be confirmed. S4360: When the duration is greater than or equal to the first threshold, acquire all events to be confirmed that exist within the preset time window and match the temporal and spatial correlation of the confirmation signal as batch confirmation events, and parse the confirmation signal as a batch confirmation signal of the batch confirmation events.

[0063] In this embodiment, the duration of the confirmation signal refers to the time between the start and end of the detection of the confirmation signal. It can be implemented using a timer or timestamp recording method, and its purpose is to distinguish different types of confirmation intentions. The preset first threshold is the time boundary used to distinguish between single confirmation signals and batch confirmation signals. It can be configured according to actual application scenarios and user habits, and its purpose is to provide a judgment standard. A single confirmation signal is a signal with a short duration, usually corresponding to the confirmation of a single event to be confirmed, and its purpose is to accurately associate a single event. A batch confirmation event refers to multiple events to be confirmed that exist within a preset time window and match the temporal and spatial correlation of the confirmation signal, and its purpose is to identify a set of events that need to be processed in batches. A batch confirmation signal of a batch confirmation event is a signal with a long duration, used to confirm multiple batch confirmation events at once, and its purpose is to trigger the batch processing of multiple events. The preset time window is a time range set before and after the occurrence of the confirmation signal, used to limit the time boundary for searching batch confirmation events. It can be set according to the typical occurrence interval of the events to be confirmed, and its purpose is to narrow the search range and improve matching efficiency.

[0064] This application's solution categorizes confirmation signals into single confirmation signals (with a duration less than a preset first threshold) and batch confirmation signals (with a duration greater than or equal to the preset first threshold) based on their duration. For single confirmation signals, the system parses them as confirmation intentions for a single event and performs temporal-spatial correlation matching with the events to be confirmed, thus accurately associating the confirmation signal with the corresponding single event. This processing method aligns with basic confirmation signal matching logic, ensuring the accuracy of single confirmations. For batch confirmation signals, the system identifies the intention to confirm in batches and, within a preset time window, searches for all events to be confirmed that match the temporal-spatial correlation of the confirmation signal, identifying these event sets as batch confirmation events. Subsequently, the confirmation signal is parsed into batch confirmation signals for these batch confirmation events. This duration-based differentiation allows the system to employ different matching and parsing strategies based on the user's actual operational intention (single confirmation or batch confirmation). By acquiring all relevant events awaiting confirmation within a single time window and associating them with the batch confirmation signal in batch confirmation scenarios, the system avoids processing each event individually, significantly improving efficiency when handling multiple related events. This approach, which dynamically adjusts the matching and parsing strategy based on the duration of the confirmation signal, effectively complements and optimizes existing confirmation signal matching mechanisms. It enables the system to understand and respond to user confirmation operations more accurately and efficiently, especially in complex scenarios requiring the processing of multiple consecutive or related events awaiting confirmation.

[0065] In some preferred embodiments, specifically, a preset first threshold can be set to 500 milliseconds. When the system detects a confirmation signal, such as a signal generated by a user long-pressing a certain area on the monitoring interface, it first measures the duration of the signal from the press to the release. If the duration is less than 500 milliseconds, the system determines that this is a single confirmation signal and, based on the time and location of the signal occurrence, combined with the time and location of the events to be confirmed, searches for and associates the single event to be confirmed that best meets the temporal and spatial proximity criteria. If the duration is greater than or equal to 500 milliseconds, the system determines that this is a batch confirmation signal. In this case, the system sets a preset time window, for example, within 10 seconds before the signal ends, and searches for all events to be confirmed that occur within this time window and have a spatial proximity relationship with the location where the confirmation signal occurred. These found sets of events to be confirmed are identified as batch confirmation events, and the long-press signal is parsed as a batch confirmation signal for these batch confirmation events.

[0066] Reference Figure 7 In another embodiment of this application, it is further proposed that after step S4360, the following method is also included: S4370: Obtain the caregiver's identity information and corresponding caregiver function area corresponding to the batch confirmation event; S4380: If there are multiple caregiver identity information from adjacent caregiver functional areas in the batch confirmation event, then based on the movement behavior information of each caregiver across each caregiver functional area within a preset time window, obtain the caregiver's stay time and stay location in each caregiver functional area; S4390: Based on the dwell time and dwell location, determine the caregivers and their associated functional areas corresponding to the batch confirmation events, and update the verification records.

[0067] In this embodiment, the mobile behavior information refers to the spatial location change trajectory of the caregiver within a preset time window. It can be realized by using location data sequences obtained through technologies such as ultra-wideband (UWB) positioning, Bluetooth beacon positioning, or camera visual analysis. Its purpose is to reflect the caregiver's activities in different functional areas. The dwell time refers to the length of time that the caregiver continuously maintains their presence in a specific caregiver functional area. It can be calculated based on the continuous occurrence time of location data in a certain functional area in the mobile behavior information. Its purpose is to quantify the caregiver's participation level in a certain functional area. The dwell location refers to the main spatial area or specific coordinates where the caregiver is located in a specific caregiver functional area. It can be determined based on the area with the densest distribution of location data in the mobile behavior information or the average location. Its purpose is to accurately locate the spatial correlation of the caregiver when batch confirmation events occur.

[0068] This application's solution, after identifying batch confirmation events, further obtains the caregiver identity information and their responsible functional areas related to these events, laying the foundation for subsequent detailed analysis. When batch confirmation events involve multiple caregivers from adjacent functional areas, the solution no longer relies solely on functional area division but delves into the caregiver's movement behavior across these functional areas within a preset time window. Through analysis of movement trajectories, the system can obtain the caregiver's dwell time and location in each functional area. Based on this dwell time and location information, the solution can determine which caregiver and which functional area each sub-event in the batch confirmation event is associated with. This detailed attribution based on caregiver spatial behavior overcomes the limitation of batch confirmation in distinguishing individual behaviors. By updating this location association information to the verification record, the verification record can reflect the caregiver's operations and responsibilities when performing batch tasks in complex scenarios. This, combined with the established care rule verification framework, jointly improves the accuracy of the entire monitoring method and solves the problems of inaccurate verification records and inability to reflect the consistency between caregiver behavior and care rules raised in the background technology.

[0069] In some embodiments, this application is implemented as follows: In a lunchtime scenario at a childcare facility, assuming there are two caregivers, Caregiver A and Caregiver B, responsible for dining area 1 and dining area 2 respectively. The system detects that Caregiver A sends a confirmation signal with a duration exceeding a first threshold by long-pressing a confirmation button on a handheld device, which the system identifies as a batch confirmation signal. Within a preset time window associated with this confirmation signal, a total of 10 "feeding completed" pending confirmation events occur in dining area 1 and dining area 2. The system first obtains the caregiver identity information (Caregiver A, Caregiver B) and the corresponding functional areas (Dining area 1, Dining area 2) associated with these 10 batch confirmation events. Since these 10 events occur in functional areas and involve multiple caregivers, the system further obtains the movement behavior information of Caregiver A and Caregiver B within the preset time window, such as location trajectory data obtained through an indoor positioning system. Based on this movement behavior information, the system calculates the dwell time and location of caregiver A in dining area 1 and dining area 2, as well as the dwell time and location of caregiver B in dining area 1 and dining area 2. For example, the analysis shows that caregiver A mainly stayed in dining area 1 during this time window, with a higher percentage of dwell time, and their dwell location was close to the location of the 7 "feeding completed" events that occurred in dining area 1; caregiver B mainly stayed in dining area 2, with a higher percentage of dwell time, and their dwell location was close to the location of the 3 "feeding completed" events that occurred in dining area 2. Based on this dwell time and location information, the system determines that the 7 "feeding completed" events in dining area 1 correspond to caregiver A and their associated dining area 1, and the 3 "feeding completed" events in dining area 2 correspond to caregiver B and their associated dining area 2. Finally, the system updated the verification records, recording the caregivers and functional areas corresponding to each of the 10 "feeding completed" events, such as "Event 1 (Dining Area 1, feeding completed) - Caregiver A", "Event 2 (Dining Area 1, feeding completed) - Caregiver A", "Event 8 (Dining Area 2, feeding completed) - Caregiver B", etc.

[0070] Reference Figure 8 In another embodiment of this application, the method further includes: S9000: When batch confirmation events involve different categories of care rules, obtain the event type of each batch confirmation event; S10000: Matches the event type and the duration of the acknowledgment signal according to the preset category priority rules; S11000: Based on the category priority rule, batch confirmation events of different event types are grouped for confirmation. The batch confirmation signal is parsed into multiple groups of sub-batch confirmation signals, which are matched to the events to be confirmed for the corresponding event types, and the corresponding verification records are updated.

[0071] In this embodiment, batch confirmation events involving different categories of care rules refer to the fact that within a preset time window corresponding to a batch confirmation operation, the system identifies multiple different care rule types among the events to be confirmed that are temporally and spatially associated with the batch confirmation signal. For example, a batch confirmation may cover the confirmation of multiple care tasks such as feeding, diaper changing, and soothing to sleep. Obtaining the event type of each batch confirmation event refers to the system extracting the event type identifier from the event information parsed from the actual event stream. When a batch confirmation event contains multiple events to be confirmed, the system needs to obtain the types of all these events. Category priority rules refer to a pre-defined set of rules used to guide the processing order or importance of different types of care events in a batch confirmation scenario. These rules can be set based on the urgency, importance, or specific procedures of the institution. The rules can be represented as a priority list of different event types or an allocation strategy related to the duration of the confirmation signal. Group confirmation refers to decomposing or mapping the batch confirmation signal corresponding to a complete batch confirmation operation into independent confirmation processes for each event type, based on the different event types and their priorities contained within it. A sub-batch acknowledgment signal is a signal segment or logic unit that is parsed or derived from the original batch acknowledgment signal and is specifically used to acknowledge a particular event type or a group of events of the same type.

[0072] This application's solution, when batch confirmation events involve different categories of care rules, first obtains the specific types of these events, which forms the basis for refined confirmation. Subsequently, based on the event type and the duration of the confirmation signal, the system matches preset category priority rules. These rules reflect the importance and processing order of different care tasks, providing a basis for subsequent group confirmation. Based on the matched category priority rules, the system adopts a group confirmation approach for batch confirmation events of different event types, logically parsing the original batch confirmation signal into multiple groups of sub-batch confirmation signals, each group corresponding to one or a category of event types. These sub-batch confirmation signals are matched to the corresponding events to be confirmed for their respective event types, thereby achieving targeted confirmation of different types of events. Finally, the system updates the corresponding verification records to ensure that the records accurately reflect the caregiver's confirmation status for various tasks. This approach, combined with prior solutions, not only solves the caregiver-region association problem in batch confirmation scenarios but also further addresses the accuracy issue when batch confirmation events involve multiple types of tasks, avoiding confusion of confirmation information for different types of events. This makes the verification records more refined and reliable, and more accurately reflects the caregiver's execution and confirmation process for complex care tasks.

[0073] In some preferred embodiments, for example, a caregiver completes three tasks consecutively within a short period: feeding an infant, changing a diaper, and soothing a crying infant, and confirms these three tasks all at once using a long-duration confirmation signal (e.g., long-pressing a confirmation button). The system first identifies this batch confirmation signal and finds three events to be confirmed that are temporally and spatially associated with the signal within a preset time window: feeding completed, diaper changing completed, and soothing successfully. The system obtains the event types of these three events to be confirmed, namely "feeding," "diaper changing," and "soothing." Simultaneously, the system detects the duration of the confirmation signal. According to a preset category priority rule, for example, setting "soothing" to have a higher priority than "feeding," which is higher than "diaper changing," and different priorities may correspond to different allocation ratios or patterns of confirmation signal duration, the system groupes and confirms these three different event types of events according to this priority rule. For example, if the confirmation signal lasts for 10 seconds, the priority rule might instruct the system to logically parse these 10 seconds of signal into sub-signals corresponding to the "soothing" event (e.g., the logical meaning of the first 5 seconds), the "feeding" event (e.g., the logical meaning of the next 3 seconds), and the "diaper changing" event (e.g., the logical meaning of the last 2 seconds). The system matches the sub-batch confirmation signals corresponding to "soothing" to the soothing success event, the sub-batch confirmation signals corresponding to "feeding" to the feeding completion event, and the sub-batch confirmation signals corresponding to "diaper changing" to the diaper changing completion event. Finally, the system updates the verification records, generating three independent verification records that accurately associate the event type and confirmation information, respectively recording the caregiver's confirmation of the soothing, feeding, and diaper changing tasks.

[0074] Reference Figure 9 In another embodiment of this application, step S11000 further includes: S11100: Obtain the sequence of task execution behaviors of the corresponding caregiver for the events to be confirmed under each event type; S11200: Establish an event chain model based on the temporal logic of continuous actions in the task execution behavior sequence and the occurrence time and spatial location of each event to be confirmed; S11300: Based on the event chain model, verify the consistency between the caregiver's execution logic and care rules, and update the verification record.

[0075] In this embodiment, the task execution behavior sequence refers to an ordered set of continuous or related physical actions, system interactions, and positional changes that a caregiver performs during the handling of a specific event to be confirmed. This sequence can be obtained using sensor data, video analysis, system logs, etc. The temporal logic of continuous actions refers to the chronological order of each action in the task execution behavior sequence and the time intervals between actions. The event chain model is a structured representation that constructs the caregiver's task execution behavior sequence, the occurrence time and spatial location of the event to be confirmed, and other relevant contextual information according to temporal order and logical association. This model can be represented using graph structures, sequence models, or other data structures. Verifying the consistency between the caregiver's execution logic and the care rules involves comparing the actual behavior process of the caregiver reflected in the event chain model with the expected behavior sequence, condition judgments, action order, and temporal and spatial requirements specified in the care rules for this type of event, and determining whether the two are consistent.

[0076] The proposed solution, through the aforementioned steps, achieves in-depth analysis and verification of caregiver behavioral logic. After completing the grouping and confirmation of batch confirmation events and updating the preliminary verification records, this solution further obtains the caregiver task execution behavior sequences related to these events to be confirmed. These behavior sequences contain the specific operational details of the caregiver when handling these events. By analyzing the temporal order and logical relationship of continuous actions in the behavior sequences, and combining this with the time and spatial location of the events to be confirmed, an event chain model is constructed. This model links the caregiver's micro-behavior with macro-events, forming a complete view of the behavioral path and decision-making process. Based on this event chain model, the system can compare it with the detailed execution requirements for this type of event in the preset care rules. This comparison not only determines whether an event has occurred or been confirmed, but also verifies whether the steps, order, timing, and location of the caregiver's task execution conform to the logic of the rules. For example, the rules may require first picking up an item, then moving it to a specific location, and then performing the operation. Through the event chain model, it is possible to check whether the caregiver's behavior sequence follows this order and location requirements. If inconsistencies are found, the verification records are updated, and potential non-standard behaviors are marked. This in-depth behavioral analysis compensates for the shortcomings of relying solely on confirmation signals for verification, revealing deeper logical biases in caregiver behavior and thus enabling a more accurate assessment of care quality.

[0077] In some preferred embodiments, this application is implemented as follows: the system has identified the "feeding" event to be confirmed, and the caregiver has confirmed the event through a batch confirmation signal. The system acquires the caregiver's behavioral sequence before and after the feeding event, such as: taking the bottle out of the locker, walking to the water dispenser, adjusting the water temperature, walking to the infant's crib, picking up the infant, feeding, putting the infant back to the crib, and washing the bottle. These behavioral sequences are accompanied by timestamps and location information. Based on this information, an event chain model is established, associating behavioral nodes such as taking out the bottle, walking to the water dispenser, and feeding with the time and location of the feeding event to be confirmed. Care rules may stipulate that steps such as washing hands, checking the milk temperature, and picking up the infant are required before feeding. Based on the event chain model, the system analyzes whether the caregiver's behavioral sequence includes these steps, whether the order is correct, whether the time interval is reasonable, and whether the location meets the requirements. If the behavior sequence shows that the caregiver takes the bottle directly from the locker and starts feeding without washing hands and checking the milk temperature, or if the feeding location is not in the designated area, then the execution logic is determined to be inconsistent with the care rules, the verification record is updated, and the feeding process is marked as non-standard.

[0078] Reference Figure 10 This application further provides a remote monitoring system for infant and toddler care, the system comprising: Care rule storage module 1 is used to obtain preset care rules, which include at least one trigger condition event and at least one response event sequence corresponding to the trigger condition event; The data acquisition module 2 is used to collect behavioral and environmental information during the care of infants and young children, and to parse the behavioral and environmental information into a stream of actual events with time stamps. The actual event stream includes object identifier, event type, occurrence time and spatial location. Detection module 3 is used to detect whether the actual event stream meets the triggering condition event. If a real-time event that meets the triggering condition event is detected, the real-time event is used as the triggering event instance. Module 4 is used to detect whether a sequence of response events corresponding to the trigger event instance has occurred in the actual event stream based on the trigger event instance. Verification module 5 is used to generate a verification record that represents the consistency between caregiving behavior and care rules if a match is successful. The verification record includes the triggering condition event, the sequence of response events, and the corresponding occurrence time and spatial location.

[0079] The care rule storage module is a unit used to store and manage preset care rules. It can be implemented using a database, file system, or memory structure, and its purpose is to provide a rule basis for subsequent event detection and behavior verification. The acquisition module is a unit used to acquire behavioral and environmental information during infant care and transform it into a structured event stream. It can be implemented using various data acquisition devices such as sensors, cameras, and microphones, as well as data processing units, and its purpose is to acquire objective care process data. The detection module is a unit used to analyze the actual event stream and identify events that meet preset trigger conditions. It can use an event-based... The care rule is implemented using a matching algorithm or rule engine to determine the triggering timing of the care rule. The judgment module is a unit used to search and confirm whether there is a corresponding response event sequence in the actual event stream based on the trigger event instance. It can be implemented using a sequence comparison algorithm or state machine model. Its purpose is to verify whether the actual behavior conforms to the expected response of the rule. The verification module is a unit used to generate a record of the consistency between the care behavior and the care rule logic after the response event sequence is confirmed. It can be implemented using data recording, report generation, or a log system. Its purpose is to provide traceable verification information that includes cause and effect.

[0080] The solution in this application obtains preset care rules through a care rule storage module. These rules transform care plans into sequences of triggering condition events and response events. The data acquisition module continuously collects behavioral and environmental information during infant and toddler care and parses it into a time-stamped stream of actual events, providing data for subsequent analysis.

[0081] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for remote monitoring of infant and toddler care, characterized in that, include: Obtain preset care rules, wherein the care rules include at least one trigger condition event and at least one response event sequence corresponding to the trigger condition event; Collect behavioral and environmental information during infant and toddler care, and parse the behavioral and environmental information into a time-stamped actual event stream, which includes object identifier, event type, occurrence time and spatial location; Detect whether the actual event stream meets the triggering condition event. If a real-time event that meets the triggering condition event is detected, the real-time event is used as a triggering event instance. Based on the trigger event instance, detect whether a response event sequence corresponding to the trigger event instance occurs in the actual event stream; If a match is successful, a verification record is generated to represent the consistency between the care behavior and the care rule logic. The verification record includes the triggering condition event, the response event sequence, and the corresponding occurrence time and spatial location.

2. The remote monitoring method for infant and toddler care according to claim 1, characterized in that, Based on the trigger event instance, detect whether a response event sequence corresponding to the trigger event instance occurs in the actual event stream, including: Detect response events in the response event sequence that correspond to the trigger event instance and contain a preset abstract care instruction type, wherein the abstract care instruction type is determined by structured annotation; Detect the unacknowledged event that matches the response event in the actual event stream; Detect confirmation signals, match the temporal and spatial correlation between the confirmation signals and the events to be confirmed, and associate the confirmation signals with the corresponding events to be confirmed; If the match is successful, the event to be confirmed is the sequence of response events that occurred in the actual event stream corresponding to the trigger event instance.

3. The remote monitoring method for infant and toddler care according to claim 2, characterized in that, Detecting a confirmation signal, matching the confirmation signal with the temporal and spatial correlation between the confirmation signal and the event to be confirmed, and associating the confirmation signal with the corresponding event to be confirmed, includes: The confirmation signal is detected, and the confirmation signal includes a time-space identifier; Based on the comparison and analysis of the time and spatial location of the event to be confirmed with the time and spatial location of the time and location of the event to be confirmed, the time and spatial correlation between the confirmation signal and the event to be confirmed is obtained. When the temporal and spatial correlation meets the preset temporal and spatial proximity conditions, the confirmation signal and the associated event to be confirmed are determined.

4. The remote monitoring method for infant and toddler care according to claim 3, characterized in that, The method further includes: When there are multiple caregivers, a correspondence is established between the caregiver identity information of the multiple events to be confirmed, and the caregiver identity information includes caregiver location information. The temporal and spatial identifier of the confirmation signal is obtained. Based on the spatial distance between all the caregiver location information and the location information in the temporal and spatial identifier, the caregiver identity information whose spatial distance meets the preset spatial threshold is selected as candidates. According to the correspondence, the caregiver identity information associated with the event to be confirmed is matched among the candidates. The corresponding matched candidate is selected as the target caregiver according to the preset matching rules, and the caregiver identity information of the target caregiver is updated in the verification record.

5. The remote monitoring method for infant and toddler care according to claim 4, characterized in that, The preset matching rules include: Based on the location information of the candidate's caregiver, obtain the candidate's movement trajectory information between the time of the event to be confirmed and the time of the confirmation signal reception; By combining the movement trajectory information, the occurrence time and spatial location of the event to be confirmed, and the temporal and spatial identifier of the confirmation signal, the spatial trajectory characteristics of each candidate are calculated; By comparing the degree of deviation from the expected behavioral path in the spatial trajectory features with the continuity of the position in the temporal and spatial identifiers of the arrival at the confirmation signal, a behavioral consistency metric value is calculated for each candidate. The corresponding target caregiver is determined based on the consistency metric value.

6. The remote monitoring method for infant and toddler care according to claim 2, characterized in that, Detecting a confirmation signal, matching the confirmation signal with the temporal and spatial correlation between the confirmation signal and the event to be confirmed, and associating the confirmation signal with the corresponding event to be confirmed, includes: Determine the duration of the confirmation signal; When the duration is less than a preset first threshold, the confirmation signal is parsed as a single confirmation signal, and the temporal and spatial correlation between the confirmation signal and the event to be confirmed is matched so that the confirmation signal is associated with the corresponding event to be confirmed. When the duration is greater than or equal to the first threshold, all the events to be confirmed that exist within a preset time window and match the temporal and spatial correlation of the confirmation signal are obtained as batch confirmation events, and the confirmation signal is parsed as the batch confirmation signal of the batch confirmation events.

7. The remote monitoring method for infant and toddler care according to claim 6, characterized in that, When the duration is greater than or equal to the first threshold, all events to be confirmed that exist within a preset time window and match the temporal and spatial correlation of the confirmation signal are identified as batch confirmation events. After parsing the confirmation signal as a batch confirmation signal of the batch confirmation events, the process further includes: Obtain the caregiver identity information and corresponding caregiver function area corresponding to the batch confirmation event; If the batch confirmation event contains multiple caregiver identity information from adjacent caregiver functional areas, then based on the movement behavior information of each caregiver across each caregiver functional area within a preset time window, the dwell time and dwell location of each caregiver in each caregiver functional area are obtained. Based on the dwell time and dwell location, the caregivers and their associated functional areas corresponding to the batch confirmation events are determined, and the verification records are updated.

8. The method for remote monitoring of infant and toddler care according to claim 7, characterized in that, The method further includes: When the batch confirmation events involve different categories of the care rules, obtain the event type of each batch confirmation event; Based on the event type and the duration of the confirmation signal, a preset category priority rule is matched; Based on the category priority rule, batch confirmation events of different event types are grouped for confirmation. The batch confirmation signal is parsed into multiple sub-batch confirmation signals, which are then matched to the events to be confirmed corresponding to the respective event types, and the corresponding verification records are updated.

9. The remote monitoring method for infant and toddler care according to claim 8, characterized in that, Based on the aforementioned category priority rules, batch confirmation events of different event types are grouped for confirmation. The batch confirmation signal is parsed into multiple sub-batch confirmation signals, which are then matched to the events to be confirmed corresponding to the respective event types. After updating the corresponding verification records, the process further includes: Obtain the sequence of task execution behaviors for the corresponding caregiver in each event type that needs to be confirmed; An event chain model is established based on the temporal logic of the continuous actions in the task execution behavior sequence and the occurrence time and spatial location of each event to be confirmed. Based on the event chain model, verify the consistency between the caregiver's execution logic and the care rules, and update the verification record.

10. A remote monitoring system for infant and toddler care, characterized in that, The system includes: A care rule storage module is used to obtain preset care rules, wherein the care rules include at least one trigger condition event and at least one response event sequence corresponding to the trigger condition event; The data acquisition module is used to collect behavioral and environmental information during the care of infants and young children, and to parse the behavioral and environmental information into a stream of actual events with time stamps. The actual event stream includes object identifier, event type, occurrence time and spatial location. The detection module is used to detect whether the actual event stream meets the triggering condition event. If a real-time event that meets the triggering condition event is detected, the real-time event is used as a triggering event instance. The judgment module is used to detect, based on the trigger event instance, whether a response event sequence corresponding to the trigger event instance has occurred in the actual event stream; The verification module is used to generate a verification record that represents the consistency between the care behavior and the care rule logic if the match is successful. The verification record includes the triggering condition event, the response event sequence, and the corresponding occurrence time and spatial location.

Images