Method and System for Accommodation Operation Support Using Hybrid Routing Based on Customer Journey State

Abstract

The present invention relates to an accommodation operation support system that hybrid routes service requests to a cloud-based conversation processing unit and an on-device-based processing unit by combining the customer's journey status, the network quality status of the field terminal, and the type / processing grade of the user request. Before a visit, reservations are supported through cloud-based consultation, and a common identifier is issued upon reservation confirmation. After a visit or in the event of a network failure, the on-device-based processing unit performs offline redemption (verification) of the common identifier to perform check-in and basic guidance. Furthermore, the offline processing history ensures consistency of operational data by synchronizing with a cloud server upon network recovery by applying idempotent processing and collision resolution rules based on the transaction identifier.

Description

Technology Field

[0001] The present invention relates to technology for automating accommodation service operations, and more specifically, to a method and system that dynamically controls cloud-based processing and on-device-based processing according to the customer's journey status, the network quality status of the field terminal, and the type / processing grade of the user request, thereby providing seamless reservation consultation and guidance and check-in transaction processing even in an unstable network environment. Background Technology

[0002] Existing accommodation management systems often operate in a fragmented manner, with online reservation systems (e.g., OTAs, PMS) and on-site kiosk (or tablet / local server) based check-in systems running separately. This results in customer inconvenience, as they must re-enter information on-site that was previously entered online, while increasing manual work for operators to resolve data discrepancies or verify data.

[0003] In addition, cloud-based chatbots or remote consultation systems may cause check-in bottlenecks due to response delays or failures in environments where the on-site network is unstable (e.g., underground, congested areas, islands, mountainous regions, etc.). Conversely, simple kiosks have limitations in performing flexible decision-making, such as complex natural language consultation and handling various exceptions (e.g., early check-in, extension, additional personnel, etc.).

[0004] Therefore, technology is required that integrates the customer journey stages (before / after booking, before / after arrival, etc.), the on-site communication environment, and the nature of the request (consultation-type, transaction-type, exception-handling-type) to optimally distribute service requests between cloud and on-device resources and continuously provide core functions even in the event of network failures. The problem to be solved

[0005] The objectives of the present invention are as follows.

[0006] (1) A technology is provided that automatically determines the customer journey status and combines the type / processing grade of user requests and network quality status to optimally distribute consultation and guidance tasks between a cloud-based processing unit and an on-device-based processing unit.

[0007] (2) Even in the event of on-site network quality degradation or disconnection, core transactions (e.g., reservation verification, check-in approval, basic guidance, etc.) are performed in on-device standalone mode to prevent service interruption.

[0008] (3) Ensure operational data consistency by preventing duplicate reflection and conflicts when synchronizing offline processing history with the cloud server after network recovery. means of solving the problem

[0009] A lodging operation support method according to one embodiment of the present invention for achieving the above objective is,

[0010] (a) A step of determining the user’s customer journey status by collecting status information including user input information, reservation-related information and time information;

[0011] (b) a step of calculating the network quality status of the field terminal;

[0012] (c) a step of determining the request type and / or processing level for a user request;

[0013] (d) routing service requests to a cloud-based conversational processing unit or an on-device-based processing unit based on customer journey status, network quality status and request type / processing grade; and

[0014] (e) A step of generating and outputting a response through a routed processing unit;

[0015] Includes

[0016] In addition, in the pre-visit state, a cloud-based conversation processing unit performs a reservation consultation and issues a common identifier upon reservation confirmation, and in the post-visit state or network failure state, an on-device-based processing unit can perform a check-in transaction by offline redempting (verifying) the common identifier. Offline processing history is synchronized based on an idempotent key (transaction identifier) ​​upon network recovery to prevent duplicate reflection, and in the event of a conflict, it can be processed according to pre-configured conflict resolution rules. Effects of the invention

[0017] According to the present invention, by combining the scalability and intelligence of online (cloud) systems (e.g., LLM-based consultation) with the immediacy of offline (on-device) systems (low latency, network independence), it is possible to provide customers with an uninterrupted accommodation experience. Furthermore, since check-in and basic guidance are possible in on-device mode even in the event of a network failure, the operational stability of the accommodation establishment can be improved. Moreover, by applying idempotency and collision resolution during the synchronization process after offline processing, data consistency is ensured, and stable operation is possible even in the event of retransmission or communication errors. Brief explanation of the drawing

[0018] Fig. 1 This is a block diagram showing the overall configuration of a lodging operation support system according to an embodiment of the present invention. Fig. 2 This is a flowchart illustrating a hybrid routing process according to a customer journey state, network quality state, and request type / processing grade according to an embodiment of the present invention. Fig. 3 This is an example diagram illustrating the process of redemption (verification) and check-in processing using a common identifier according to an embodiment of the present invention, and the synchronization process of offline processing history. Specific details for implementing the invention

[0019] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. However, the present invention is not limited to the following embodiments and may be modified in various ways within the scope of the technical concept described in the claims.

[0020] 0. Definition of Terms

[0021] In this specification, “customer journey status” refers to a state classified as at least one of pre-reservation, post-reservation, pre-arrival, imminent arrival, post-arrival, and check-in completed, and can be determined by status information including time information (e.g., difference between scheduled check-in time and current time), location information (e.g., whether within a radius of the accommodation), and reservation processing status (e.g., payment / confirmation / cancellation status).

[0022] “Network quality status” refers to a state classified as at least one of normal mode, restricted mode, or offline mode based on communication quality indicators such as latency (RTT), packet loss rate, connection availability, and retransmission rate.

[0023] “Request type” or “processing level” refers to the result of classifying a user request into at least one of the following: compound natural language consultation type, transaction processing type (check-in / reservation verification / key issuance / facility control, etc.), and exception handling type (policy approval required).

[0024] 1. System Configuration (Refer to Fig. 1)

[0025] Referring to FIG. 1, the system of the present invention may include a cloud server (100), a field terminal (200), and a routing control module (300).

[0026] The cloud server (100) may include a cloud-based conversation processing unit (e.g., LLM-based chatbot, knowledge base integration) for reservation management functions and user consultation and guidance. The field terminal (200) may be implemented as a kiosk, tablet, local server, etc., and may include an on-device based processing unit (210) and a local database (220). The local database (220) may store a list of reservation holders, room status information, facility guidance information, policy summary information, etc.

[0027] The routing control module (300) may include a customer journey status determination unit (310), a network quality calculation unit (320), a request type / processing grade determination unit (330), and a routing policy engine (340). The routing control module (300) routes user requests to appropriate processing resources by controlling the distribution of processing between the cloud server (100) and the field terminal (200).

[0028] The customer journey status determination unit (310) can determine the journey status by combining location information of the user terminal (e.g., GPS), the reserved check-in time, the current time, the reservation status, etc. For example, a certain amount of time prior to the scheduled check-in time can be determined as before arrival, and within a certain amount of time and upon entering the radius of the accommodation facility can be determined as imminent arrival or after arrival, but these criteria can be changed according to a preset threshold value and the present invention is not limited to a specific value.

[0029] 2. Request Type / Processing Class Determination and Routing Policy (See Fig. 2)

[0030] When a user request is received, the request type / processing grade determination unit (330) classifies the request according to the processing purpose of the request and the required processing resources. For example, requests requiring complex natural language understanding, such as inquiries about refund policies, recommendations for nearby tourist attractions, and facility usage guides, can be classified as consultation type. Requests involving state changes, such as check-in approval, reservation verification, room key issuance, door lock authorization, or facility control, can be classified as transaction processing type. Requests that may require policy judgment or manager approval, such as early check-in, extension, or additional personnel, can be classified as exception processing type. Request classification can be performed by a rule-based or machine learning-based classification method, and the present invention is not limited to a specific classification method.

[0031] The routing policy engine (340) evaluates the routing policy using the customer journey status, network quality status, and request type / processing grade as inputs. For example, (i) if the customer journey status is pre-arrival and the request type is consultation type, it routes to the cloud server (100) to perform LLM-based consultation. (ii) if the customer journey status is post-arrival or the request type is transaction processing type, it activates the on-device based processing unit (210) of the field terminal (200) to perform immediate processing based on local data. (iii) if the network quality status is restricted or offline, even for a consultation type request, it may route to provide a standard response (basic guidance / FAQ / policy summary) within the range available on-device. Additionally, the routing policy engine (340) may include priority rules, for example, to prioritize transaction processing types for on-device processing, and to set conditions for employee call or remote consultation transfer if exception handling types fall outside the policy scope.

[0032] 3. Network quality-based mode transitions and feature set control (see Fig. 2)

[0033] Referring to FIG. 2, the system measures network quality indicators (e.g., latency, loss rate, availability, etc.) through a network quality calculation unit (320) to calculate the network quality status (S110). If the calculated result is above a preset threshold, it is set to normal mode, and if some indicators are below the threshold, it is set to restricted mode, or if it is unconnectable or significantly below the threshold, it is set to offline mode (S120).

[0034] Next, the customer journey status determination unit (310) determines the customer journey status using status information (S130), and the request type / processing grade determination unit (330) determines the type / processing grade of the user request (S140). Subsequently, the routing policy engine (340) evaluates the mode (normal / restricted / offline), journey status, and request type / processing grade together to select cloud-based processing or on-device-based processing (S150).

[0035] In normal mode, a set of functions that enable LLM-based consultation and integration with the latest content of the cloud server (100) can be activated. In restricted mode, the frequency of cloud calls can be limited, transaction processing requests can be prioritized for on-device processing, and consultation requests can be replaced with summary responses or local FAQs. In offline mode, the local database (220) and local policy information of the field terminal (200) can be used to configure the core set of functions, such as check-in, reservation confirmation, and basic facility guidance. When the network is restored, the function set can be expanded step by step by transitioning to restricted mode or normal mode.

[0036] 4. Issuance of Common Identifier and Offline Redemption (See Fig. 3)

[0037] The cloud server (100) can generate a common identifier (e.g., QR code, token) upon reservation confirmation and transmit it to the user terminal. The field terminal (200) can verify the validity of the reservation (redemption) and perform a check-in process using the common identifier input or presented from the user terminal.

[0038] In particular, to enable redemption even in offline mode, the field terminal (200) may include a verification procedure that can be performed without a network connection using verification information included in or associated with a common identifier. The verification procedure may include anti-tampering verifications such as hash-based verification, digital signature verification, and Message Authentication Code (MAC) verification, and the present invention is not limited to a specific verification method. Additionally, the common identifier may be configured to include an expiration time, reuse prevention information, etc.

[0039] 5. Synchronization of offline processing history (idempotent / conflict resolution) (see Fig. 3)

[0040] When the field terminal (200) performs transaction processing such as check-in approval, authorization, or room status change in offline mode, a transaction identifier (idempotent key) may be assigned to each processing entry. When the field terminal (200) synchronizes the processing entries to the cloud server (100) after network recovery, the cloud server (100) may be configured to prevent duplicate reflection based on the transaction identifier. For example, even if the same transaction is received multiple times due to retransmission or communication errors, it is reflected only once through a comparison of transaction identifiers.

[0041] Additionally, conflicts regarding reservation status or room status may occur during the synchronization process (e.g., duplicate check-in requests for the same room, processing check-ins for cancelled reservations, etc.), and the cloud server (100) can resolve these conflicts according to pre-configured conflict resolution rules. The conflict resolution rules may include at least one of server priority, site priority, timestamp priority, and policy priority, and different rules may be applied depending on the type of conflict. The site terminal (200) may be configured to store processing details in a log and perform retries / retransmissions in the event of synchronization failure, and may handle exception cases by calling a staff member or transferring to a remote consultation if necessary.

[0042] 6. Personnel Transfer and Exception Handling (Optional)

[0043] If the on-device based processing unit (210) detects a request or processing failure that is outside the policy scope, the routing policy engine (340) can route it to be transferred to an employee call or remote consultation. This ensures operational safety even in offline environments or situations where policy judgment is required. Explanation of the symbols

[0044] 100: Cloud Server 110: Cloud-based dialogue processing unit 120: Reservation Management Department 200: Field terminal 210: On-device based processing unit (On-device concierge / transaction processing unit) 220: Local database 300: Routing Control Module 310: Customer Journey Status Determination Unit 320: Network Quality Calculation Unit 330: Request Type / Processing Grade Determination Unit 340: Routing Policy Engine 400: User terminal (User terminal for reservation confirmation and receiving / presenting common identifier; e.g., app / web / messenger terminal)

Claims

Claim 1 A method for supporting accommodation operations performed by a computing device, comprising: (a) determining the customer journey status of a user as one of pre-reservation, post-reservation, pre-arrival, imminent arrival, post-arrival, and check-in completed, based on status information including location information, reservation processing status, scheduled check-in time, and current time obtained from a user terminal or a field terminal; (b) measuring at least one of the latency, packet loss rate, connection availability, and retransmission rate of the field terminal, and comparing the measurement result with a preset threshold value to determine the network quality status of the field terminal as one of normal mode, restricted mode, and offline mode; (c) determining the user request as one of a complex natural language consultation type, a transaction processing type involving state changes, and an exception processing type requiring policy judgment or administrator approval, based on the processing purpose of the user request and the required processing resources; (d) evaluating the customer journey status, the network quality status, and the type of the user request in combination, such that if the network quality status is normal mode, the customer journey status is pre-arrival, and the user request is a complex natural language consultation type, the user request is routed to a cloud-based conversation processing unit, and if the customer journey status is post-arrival or the user A method for supporting accommodation operation, comprising: a step of routing the user request to an on-device-based processing unit when the request is transaction processing type or the network quality status is restricted mode or offline mode; and (e) a step of generating and outputting a response to the user request by the routed processing unit. Claim 2 A lodging operation support method according to claim 1, wherein the routing step activates the natural language consultation and latest content linkage functions of the cloud-based conversation processing unit in the normal mode, limits the frequency of cloud calls in the restricted mode while prioritizing the conversion of the transaction processing type user request to the on-device-based processing unit and provides a summary response or FAQ stored locally for the complex natural language consultation type user request, performs at least one core function among reservation confirmation, check-in, and basic facility guidance using the local database and local policy information of the field terminal in the offline mode, and transfers the exception handling type user request to a staff call or remote consultation if it falls outside the policy range determined by the local policy information. Claim 3 A method for supporting accommodation operation according to claim 1 or claim 2, further comprising the step of generating a common identifier in response to reservation confirmation and providing it to a user terminal when routed to the cloud-based conversation processing unit, and further comprising the step of receiving the common identifier and performing redemption to perform reservation confirmation or check-in approval when routed to the on-device-based processing unit. Claim 4 A method for supporting accommodation operations according to claim 3, wherein the redemption can be performed without a network connection, comprising at least one cryptographic verification among digital signature verification using verification information included in or associated with the common identifier, message authentication code verification, or forgery or tamper-proof verification equivalent thereto. Claim 5 A method for supporting accommodation operations according to claim 4, wherein an offline mode is set according to the network quality status, a transaction identifier is assigned to processing history generated in the offline mode, and upon network recovery, the processing history is transmitted to a cloud server for synchronization, while preventing duplicate reflection of the same processing history based on the transaction identifier, and resolving the conflict according to a preset priority rule or timestamp standard when a conflict occurs between the reservation status or room status. Claim 6 A lodging operation support device comprising an input / output interface, a network communication unit, and one or more processors, wherein the one or more processors determine the user’s customer journey state as one of pre-reservation, post-reservation, pre-arrival, imminent arrival, post-arrival, and check-in completed based on state information including location information, reservation processing status, scheduled check-in time, and current time obtained from a user terminal or a field terminal; (ii) measure at least one of the latency, packet loss rate, connection availability, and retransmission rate of the field terminal, and calculate the network quality state of the field terminal as one of normal mode, restricted mode, and offline mode by comparing the measurement result with a preset threshold value; (iii) determine the user request as one of a complex natural language consultation type, a transaction processing type involving state changes, and an exception processing type requiring policy judgment or administrator approval based on the processing purpose of the user request and the required processing resources; and (iv) evaluate the customer journey state, the network quality state, and the type of the user request in combination, and if the network quality state is normal mode, the customer journey state is pre-arrival, and the user request is a complex natural language consultation type, the user request is sent to a cloud-based dialogue processing unit. A lodging operation support device configured to route, and if the customer journey status is after arrival or the user request is transaction processing type or the network quality status is restricted mode or offline mode, route the user request to an on-device based processing unit, (v) and generate and output a response to the user request by the routed processing unit. Claim 7 A lodging operation support device according to claim 6, wherein the one or more processors are configured to activate the natural language consultation and latest content linkage functions of the cloud-based conversation processing unit in the normal mode, prioritize the conversion of the transaction processing type user request to the on-device-based processing unit while limiting the frequency of cloud calls in the restricted mode, provide a summary response or FAQ stored locally for the complex natural language consultation type user request, perform at least one core function among reservation confirmation, check-in, and basic facility guidance using the local database and local policy information of the field terminal in the offline mode, and transfer the exception handling type user request to a staff call or remote consultation if it falls outside the policy range determined by the local policy information. Claim 8 A lodging operation support device according to claim 6 or claim 7, wherein one or more processors are configured to generate a common identifier in response to reservation confirmation and provide it to a user terminal when routed to the cloud-based conversation processing unit, and to receive the common identifier and perform redemption when routed to the on-device-based processing unit to perform reservation confirmation or check-in approval. Claim 9 A lodging operation support device according to claim 8, configured such that the redemption can be performed without a network connection by including at least one cryptographic verification among digital signature verification using verification information included in or associated with the common identifier, message authentication code verification, or forgery or tamper-proof verification equivalent thereto. Claim 10 A lodging operation support device according to claim 9, wherein one or more processors are configured to set an offline mode according to the network quality status, assign a transaction identifier to processing history generated in the offline mode, synchronize the processing history with a cloud server upon network recovery, prevent duplicate reflection of the same processing history based on the transaction identifier, and resolve the conflict according to a preset priority rule or timestamp standard when a conflict occurs between the reservation status or the room status.

Images