Intelligent mobile guest room service workstation order listening system and method thereof

Abstract

The invention discloses an intelligent mobile guest room service work station order listening system and method, and relates to the technical field of intelligent service robots, and the system comprises the following components: a front-end application which is used for receiving a voice or text request of a client and supports a voice input mode and a text input mode; and the back-end server comprises a voice recognition module, a natural language processing module and a service request classification module. According to the intelligent mobile guest room service workstation order listening system and the method thereof, built-in hotel map data and an advanced path planning algorithm are utilized to plan an optimal path from the current position to the target location for the service robot, time waste of the robot in the advancing process is reduced, the service execution speed is increased, and the service quality is improved. And meanwhile, the energy consumption is reduced, the service efficiency is improved, and the function requirements of automation and intelligence are met.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent service robots, in particular to an intelligent mobile room service workstation order listening system and a method thereof. Background Art

[0002] With the rapid development of science and technology and the continuous improvement of people's living standards, the hotel industry is facing growing customer expectations and fierce market competition. In the field of room service, the traditional service model mainly relies on manual operation, which has many limitations. In traditional room service, customers usually make service requests to the hotel front desk by calling the room phone. The front desk staff manually records the customer's needs and then assigns the task to the corresponding service personnel through the intercom or internal communication system. This method is inefficient. It often takes a long time from the customer's request to the service staff starting to perform the task, resulting in long waiting time for customers and reduced satisfaction. Manual recording and assignment of tasks are prone to information transmission errors, which may lead to inefficiencies in the process. This will lead to service errors and affect customer experience; in terms of service resource management, the traditional model lacks effective overall planning and real-time monitoring; it is difficult for hotels to accurately grasp the restaurant's order pressure, food inventory, and the service staff's real-time work status and location distribution, and they are unable to reasonably allocate resources according to actual needs in a timely manner; this may lead to certain service demands not being met in time during peak hours, and there is idle waste of resources during low hours; at the same time, the quality stability of manual services is difficult to guarantee; the work efficiency and service quality of service personnel are greatly affected by factors such as personal emotions and experience level, and the services provided by different service personnel may be different, making it difficult to achieve standardized and normalized service processes.

[0003] Therefore, in order to meet the urgent needs of the hotel industry to improve service quality, increase operational efficiency and reduce costs, there is an urgent need for a more advanced, efficient and intelligent room service workstation listening system to overcome the shortcomings of existing technologies and realize the automation, intelligence, precision and personalization of room service. Summary of the invention

[0004] In order to solve the above technical problems, the present invention is implemented through the following technical solutions: an intelligent mobile room service workstation order listening system and method thereof, comprising: a front-end application for receiving voice or text requests from customers, and supporting both voice and text input modes; The back-end server includes a speech recognition module, a natural language processing module and a service request classification module; the speech recognition module is used to convert the customer's speech request into a request text; the natural language processing module is used to understand the content of the request text or text request to obtain the service request; the service request classification module is used to classify the service request and call the corresponding service module for processing; Service robots: Service robots are used to receive task instructions from the backend server. Based on the task instructions, the selected robot performs room service tasks, including meal delivery and cleaning. After the service is completed, the service robot reports the task completion status and abnormal information to the central control system. The feedback module is used to send confirmation information to customers after the task is completed to ensure service satisfaction.

[0005] Preferably, the service request classification module is used to classify service requests and call corresponding service modules for processing. The implementation process is as follows: The service request classification module first obtains the service request parsed by the natural language processing module and verifies the validity and feasibility of the service request, including: checking whether the room number of the target room is legal and within the hotel room number range, whether the service type is a recognizable type preset by the system, and obtaining the room number check result and service type respectively; based on the room number check result, according to the service type, the required resources are determined and evaluated to obtain the resource evaluation result; combined with the internal layout of the hotel and the current location information of each resource, the general process of task execution is planned to obtain the task plan; based on the resource evaluation results and task plan, task instructions are generated; task instructions include a detailed description of the task, resource identification for executing the task, the execution order and time requirements of the task, the task location and task standard details.

[0006] Preferably, whether the service type is a recognizable type preset by the system, the process is as follows: Extract keywords based on the service request to obtain keywords, and then match the extracted keywords with the preset service type keyword library to obtain the service type; determine the service module to be called according to the service type; the service module includes a food delivery robot, a cleaning robot, and a maintenance service scheduling robot; For food delivery services, the food delivery robot corresponding to the restaurant or food delivery is called, and the food delivery robot is responsible for interacting with the restaurant system, arranging meal preparation, meal pick-up route planning, and task setting for the food delivery robot; for cleaning services, the cleaning robot is called, which is responsible for cleaning task scheduling, cleaning tool preparation, and cleaning process control and execution; for maintenance services, the maintenance service scheduling robot is called to contact maintenance personnel, deploy maintenance tools, and arrange maintenance tasks.

[0007] Preferably, the required resources are determined and evaluated based on the service type to obtain a resource evaluation result, and the process is as follows: for food delivery tasks, the restaurant's current order pressure, food inventory, and the status of the available food delivery robots are evaluated, the food delivery robot status includes the current position of the food delivery robot, the current power level of the food delivery robot, and the current load of the food delivery robot, to obtain a food delivery resource evaluation result; for cleaning tasks, the scheduling of the cleaning robot and the availability of cleaning tools are considered to obtain a cleaning resource evaluation result; the food delivery resource evaluation result is combined with the cleaning resource evaluation result to obtain a resource evaluation result.

[0008] Preferably, the service module to be called is determined according to the service type, and the service parameters in the service request are passed to the selected service module, wherein the service parameters include the room number A and the type of food of the target room in the meal delivery service; the room number B and the cleaning requirements of the target room in the cleaning service; and the room number C and the repair items of the target room in the maintenance service, so that the service module can accurately perform the task; then the service module activates the room service task and feeds back the task start information to the service request classification module; After the service module activates the room service task, it intelligently selects the most suitable service robot to perform the task based on the service request and the current location and load of the service robot, obtains the selected robot, and then assigns the task. After the task assignment is completed, the service request classification module sends the task instruction to the selected robot and updates the order status to "assigned"; After receiving the task instruction, the selected robot obtains the latest internal map data of the hotel from the hotel's central control system, and uses the path planning algorithm to calculate the optimal path from the current location to the target room; based on the optimal path, the selected robot uses built-in sensors to detect the surrounding environment in real time, avoid obstacles, and ensure safe driving; after the selected robot arrives at the target room, it notifies the guest of its arrival through voice prompts or screen display information, and interacts with the guest according to the type of service to complete the task; after the service is completed, the robot asks the guest about satisfaction and records feedback; after the selected robot completes the task, it returns to the charging station or standby area according to the return instruction of the central control system; the central control system updates the order status as completed or incomplete, and triggers a customer satisfaction survey, or triggers a task remediation plan to re-perform room service until the central control system updates the order status to completed.

[0009] Preferably, the service request classification module continuously monitors the execution status of the room service tasks. If an abnormality occurs, such as a food delivery robot malfunctioning or a cleaning robot failing to arrive on time, an abnormal alarm is generated and an abnormal dispatch instruction is generated. The abnormal dispatch instruction is sent to the logistics personnel, who manually handle the abnormal situation to ensure that the service can be completed smoothly or remedial measures are taken.

[0010] Preferably, the process of selecting a robot to perform room service tasks is as follows: For meal delivery tasks: determine the dish and drink information to be delivered according to the task instructions, and use the built-in hotel map data and path planning algorithm to plan the meal pickup path and meal delivery path from the current location to the restaurant. The path planning algorithm is the A* algorithm. The hotel map data contains the layout information of each floor of the hotel: the location of the rooms, the direction of the corridors, the distribution of stairs and elevators, and the specific coordinates of the restaurant, power distribution room, and storage area. For cleaning tasks: determine whether the tools and consumables required for cleaning are complete and can be used normally. If there are any missing or faulty tools, generate a cleaning tool missing report and report it to the backend server to obtain replenishment or repair arrangements. Consumables include brooms, mops, and detergents. Then, the path planning algorithm is also used to plan the travel path from the current position to the target room in combination with the target room, avoiding crowded areas and areas where other operations are being carried out. Then, the built-in sensors are used to perceive the surrounding environment information in real time, and the surrounding environment information is integrated and corrected with the hotel's internal map data to construct an environmental map, which is used to select the robot to avoid obstacles and follow the delivery path or travel path to the target room; the built-in sensors include lidar and cameras; ensuring that the selected robot reaches the target room safely and efficiently; obstacles include static furniture, walls, and dynamic personnel and other service equipment.

[0011] Preferably, the built-in hotel map data and path planning algorithm are used to plan the meal pickup path from the current location to the restaurant. The process is as follows: Based on the Wi-Fi signal strength fingerprint positioning, the current location is clarified, including the current floor of the hotel and the specific coordinate point nearby; Parse the specific location information of the restaurant from the task instruction sent by the backend server, or determine the target location based on the restaurant location coordinates pre-marked in the hotel map data; Based on the A* algorithm, the hotel map data is gridded and the hotel space is divided into grid cells of the same size. Each grid cell is marked as passable, including the interior of the room and the area where the wall is located as impassable, and the corridor area as passable. At the same time, each grid cell is assigned a corresponding coordinate index to facilitate subsequent algorithm calculation and node representation; Based on the robot's built-in sensors, it obtains real-time information about the surrounding environment and fixed obstacles marked in the map data, and marks the locations of obstacles in the map model; Create a priority queue to store the nodes to be processed. Initially, add the node corresponding to the current position of the service robot to the queue to obtain the priority queue: open_set; establish a dictionary to record the predecessor node of each node, which is initially empty, and obtain the predecessor node dictionary: came_from; record the actual cost from the current position of the robot, that is, the starting node to each node. Initially, the cost of all nodes is set to infinity, and only the cost of the starting node is set to 0, and the actual cost dictionary is obtained: g_score; record the estimated total cost from the starting node to each node. Initially, the cost of all nodes is set to infinity, and the cost of the starting node is set to the value of the heuristic function, and the estimated total cost dictionary is obtained: f_score; Loop operation: Take the node current with the lowest cost from the open_set priority queue; if the current node is the target node, that is, the node corresponding to the location of the restaurant, call the reconstruct_path function, use the came_from dictionary to trace back from the target node to the starting node, build a complete meal-picking path and return the path information; if current is not the target node, traverse all its neighbor nodes neighbor; for each neighbor node, calculate the actual cost tentative_g_score from current to neighbor; if tentative_g_score is less than the current g_score of neighbor, it means that a better path to reach the neighbor node has been found, and the values ​​of the corresponding neighbor node in came_from, g_score and f_score are updated; if the neighbor node is not in open_set, add it to the open_set queue for subsequent evaluation and processing; The operation continues in a loop until the target node is found or the open_set queue is empty. If the open_set queue is empty, it means that no feasible path can be found. At this time, the path planning failure is reported to the backend server.

[0012] Preferably, in the A* algorithm, the calculation formula for evaluating the total cost f(n) of each node is: ; Among them, f(n): represents the total estimated cost from the starting node to the target node through node n; g(n): is the actual cost from the starting node to the current node n; h(n): is the estimated cost from the current node n to the target node; Assume the current node coordinates are (x_n, y_n) and the target node coordinates are (x_goal, y_goal), then the Manhattan distance calculation formula is: ; Assume the current node coordinates are (x_n, y_n), the target node coordinates are (x_goal, y_goal), and the Euclidean distance calculation formula is: ; When traversing the neighboring nodes neighbor of the current node current, the actual cost tentative_g_score from current to neighbor is calculated, and the calculation formula is: tentative_g_score=g(current)+cost(current,neighbor); Where g(current) is the actual cost calculated from the starting node to the current node current; cost(current,neighbor) represents the cost of moving from the current node current to the neighbor node neighbor; Or consider the difficulty coefficient of passing through this path: a higher cost weight is given to corridors with more people and narrower corridors. For example, if the distance between the centers of adjacent grid cells is recorded as d in the map and there is no special difficulty coefficient, cost(current,neighbor) is directly d. If the corridor congestion is considered, the congestion coefficient is k (k>=1, the more severe the congestion, the larger the k value), then cost(current,neighbor) is expressed as d*k.

[0013] A method for listening to orders at an intelligent mobile room service workstation comprises the following steps: Step 1: The customer initiates a service request through the front-end application, verifies the validity and feasibility of the service request, checks whether the target room number is legal and within the hotel room number range, extracts keywords based on the service request, matches them with the preset service type keyword library, determines whether the service type is a type that the system can recognize, obtains the room number check result and service type, and determines the service module to be called based on the determined service type; Step 2: Conduct resource evaluation, evaluate the food delivery task resources and cleaning task resources separately, obtain the food delivery resource evaluation results and the cleaning resource evaluation results, and obtain the overall resource evaluation results by integrating the food delivery resource evaluation results and the cleaning resource evaluation results; plan the general process of task execution in combination with the internal layout of the hotel and the current location information of each resource, and obtain the task plan; generate task instructions based on the resource evaluation results and the task plan; send the generated task instructions to the selected service module, and update the order status to "allocated"; Step 3: After receiving the task instruction, the service module activates the room service task and feeds back the task start information to the service request classification module. It plans the meal pickup path and meal delivery path for the meal delivery task according to the service type, and plans the travel path for the cleaning task. The selected robot obtains the latest hotel internal map data from the hotel's central control system, and uses the path planning algorithm to calculate the optimal path from the current location to the target room. The selected robot reaches the target room based on path planning and navigation. Step 4: After the robot arrives at the target room, it notifies the guest of its arrival through voice prompts or screen display information, and interacts with the guest according to the service type to complete the corresponding room service task. After the service is completed, the robot asks the guest about satisfaction and records feedback information; After the selected robot completes the task, it returns to the charging station or standby area according to the return instruction of the central control system; The central control system updates the order status as completed or incomplete, and triggers a customer satisfaction survey, or triggers a task remediation plan to re-perform room service until the central control system updates the order status to completed.

[0014] The present invention provides an intelligent mobile room service workstation order listening system and method thereof, which has the following beneficial effects: 1. The intelligent mobile room service workstation listening system and method utilize built-in hotel map data and advanced path planning algorithms to plan the optimal path from the current location to the target location for the service robot, thereby reducing the time wasted by the robot during the movement, improving the speed of service execution, and also reducing energy consumption, thereby improving service efficiency and the requirements for automation and intelligent functions.

[0015] 2. The intelligent mobile room service workstation order listening system and method, and the meal pickup path technical solution are based on Wi-Fi signal strength fingerprint positioning, A* algorithm, sensor real-time perception and map fusion correction, etc., which realize accurate positioning and efficient path planning, can quickly determine the starting position and select the optimal path, and can flexibly respond to environmental changes and avoid obstacles in real time, ensure safe and efficient driving, reduce energy consumption and equipment loss, improve resource utilization efficiency, and ultimately improve service timeliness and customer satisfaction, which will help hotels improve their overall service quality and competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the framework of the present invention; Figure 2 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0017] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present invention are provided for the purpose of illustration and description, and are not intended to be exhaustive or to limit the present invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention, and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

[0018] like Figure 1 and Figure 2 As shown, the present invention provides a technical solution: an intelligent mobile room service workstation order listening system and method thereof, including a front-end application for receiving voice or text requests from customers, and supporting both voice and text input modes; The back-end server includes a speech recognition module, a natural language processing module and a service request classification module; the speech recognition module is used to convert the customer's speech request into a request text; the natural language processing module is used to understand the content of the request text or text request to obtain the service request; the service request classification module is used to classify the service request and call the corresponding service module for processing; Service robots: Service robots are used to receive task instructions from the backend server. Based on the task instructions, the selected robot performs room service tasks, including meal delivery and cleaning. After the service is completed, the service robot reports the task completion status and abnormal information to the central control system. The feedback module is used to send confirmation information to customers after the task is completed to ensure service satisfaction.

[0019] Preferably, the service request classification module is used to classify service requests and call corresponding service modules for processing. The implementation process is as follows: The service request classification module first obtains the service request parsed by the natural language processing module. The service request contains various customer demand descriptions, such as "deliver a breakfast to room 808" and "clean room 505"; and verifies the validity and feasibility of the service request, including: checking whether the room number of the target room is legal and within the hotel room number range, whether the service type is a recognizable type preset by the system, and obtaining the room number check result and service type respectively; based on the room number check result, according to the service type, determine the required resources and evaluate them to obtain the resource evaluation result; combined with the internal layout of the hotel and the current location information of each resource, plan the general process of task execution and obtain the task plan; based on the resource evaluation result and task type, Task planning generates task instructions; task instructions include detailed description of the task, resource identification for executing the task, execution order and time requirements of the task, task location and task standard details; detailed description of the task includes "dinner delivery to room 806: steak set meal, non-spicy, delivered as soon as possible" or "deep cleaning task for room 505", resource identification for executing the task includes the designated delivery robot number or cleaning robot number, execution order and time requirements of the task include completing the delivery within 30 minutes, the cleaning task is carried out within a certain time period and emergency rooms are handled first, and the task location includes the delivery robot's pick-up location, pick-up standard details, delivery location, delivery standard details, cleaning location, and cleaning standard details of the cleaning task.

[0020] Preferably, whether the service type is a recognizable type preset by the system, the process is as follows: Keyword extraction is performed based on the service request to obtain keywords. For example, for "deliver a breakfast to room 808", the keywords "deliver", "breakfast", and "room 808" are extracted; then the extracted keywords are matched with the preset service type keyword library to obtain the service type; if the keyword contains "deliver" and there are food-related words such as "breakfast", "lunch", "dinner", and "fruit", it is determined to be a meal delivery service type; if it contains the words "clean" and "cleaning", it is determined to be a cleaning service type; if there are words "repair" and "repair", it corresponds to the repair service type; determine the service module that needs to be called according to the service type; the service module includes a meal delivery robot, a cleaning robot, and a repair service scheduling robot; For food delivery services, the food delivery robot corresponding to the restaurant or food delivery is called, and the food delivery robot is responsible for interacting with the restaurant system, arranging meal preparation, meal pick-up route planning, and task setting for the food delivery robot; for cleaning services, the cleaning robot is called, which is responsible for cleaning task scheduling, cleaning tool preparation, and cleaning process control and execution; for maintenance services, the maintenance service scheduling robot is called to contact maintenance personnel, deploy maintenance tools, and arrange maintenance tasks.

[0021] Preferably, the required resources are determined and evaluated according to the service type to obtain the resource evaluation result, and the process is as follows: for the food delivery task, the current order pressure, food inventory and the state of the deployable food delivery robot of the restaurant are evaluated, and the food delivery robot state includes the current position of the food delivery robot, the current power of the food delivery robot and the current load of the food delivery robot, and the food delivery resource evaluation result is obtained; for the cleaning task, the scheduling of the cleaning robot and the availability of cleaning tools are considered to obtain the cleaning resource evaluation result; the food delivery resource evaluation result is combined with the cleaning resource evaluation result to obtain the resource evaluation result; It needs to be further explained that in the specific implementation process, for the meal delivery task: firstly, data collection is carried out. For the meal delivery task, a connection is established with the restaurant's order management system to obtain the number of orders currently being processed and to be processed, the estimated delivery time, analyze the complexity of the production of different dishes in the restaurant and the current preparation progress of each dish, and clarify the current order pressure by checking the task queue of each dish production position in the kitchen; perform pressure calculation and evaluation, and comprehensively evaluate the busyness of the restaurant based on the number of orders, the estimated delivery time and the complexity of the dishes; set evaluation indicators, if the current number of orders exceeds 80% of the restaurant's normal hourly processing volume, and there are more complex dishes to be prepared, the restaurant is judged to be in a high order pressure state; if the number of orders is between 40%-80% of the normal processing volume, it is medium pressure; if it is less than 40%, it is a low pressure state; obtain the restaurant's order pressure level; Conduct inventory inquiries, access the restaurant's food inventory management system, and check the remaining quantity of various ingredients; check the correspondence between the current inventory and the food demand of common dishes, and consider the ingredient combinations and usage standards of different dishes; by comparing the inventory with the meal requirements, determine which dishes can be prepared smoothly and which may be restricted due to insufficient ingredients; Conduct an inventory adequacy assessment to determine whether the ingredients are sufficient based on the inventory and expected order demand. If the inventory can meet all meal preparation needs expected within the next hour, the inventory is considered sufficient. If the inventory of some key ingredients can only meet the preparation of a small number of meals or is about to run out, the inventory is considered tight. Obtain an inventory adequacy assessment result. Location acquisition and analysis: The current location coordinates of all food delivery robots are acquired through the hotel's internal robot positioning system. Combined with hotel map data, the distance between each robot and the target room and restaurant is analyzed to preliminarily select a collection of service robots that are close and easy to quickly deploy to perform new tasks. Conduct power detection: read the power information from the real-time feedback data of the food delivery robots to understand the current remaining power percentage of each robot; based on the power consumption model of the robots, estimate whether each robot has enough power after completing a new task; completing a new task includes going to the restaurant to pick up food, delivering it to the target room, and a possible return trip; the power consumption model is based on previous operating data and the performance parameters of the robots; Load condition judgment: Check whether the food delivery robot is currently carrying other items (for example, some robots may be assisting in the transportation of other materials or have already picked up meals and are waiting for delivery) to determine its load status; for example, robot C has already picked up a meal and is on its way to another room, and is in a loaded state; while robot D is in an unloaded state and can take on new tasks at any time; Based on the load-bearing capacity of each robot, determine whether it can take on new meal delivery tasks; if the number and weight of meals in the new task exceeds the remaining load capacity of the robot, it is determined that the robot cannot be deployed; Comprehensive status evaluation and result generation: Taking into account the location, power, and load of the delivery robots, a list of available robots that can meet the requirements of the new delivery task is screened out, and the relevant status information of each robot is recorded as part of the resource evaluation results; For example, the final evaluation results show that Robot D and Robot E are in the right position, have sufficient power, and are unloaded, and can be deployed for new food delivery tasks, along with their specific location coordinates and details of the remaining power percentage; For cleaning tasks: first consider the scheduling of cleaning robots: Step 1: Get the scheduling plan: query the cleaning robot's scheduling from the hotel's cleaning task management system. The scheduling is made in advance based on the hotel's room occupancy rate, daily cleaning cycle, and the overall allocation of human and robot resources; analyze the correspondence between the current time and the scheduling plan, and determine which cleaning robots are idle and can be immediately deployed for new cleaning tasks, and which need to complete the current scheduled tasks before taking on new tasks; Step 2: Adjustment of urgent tasks: If the new cleaning task has a high degree of urgency, such as a guest's temporary request for emergency room cleaning, evaluate whether the existing schedule needs to be temporarily adjusted; check whether there are cleaning robots that are performing relatively non-urgent tasks, and suspend their current work to perform emergency cleaning tasks instead; based on whether the scheduling adjustment has been made and the final number of cleaning robots that can be deployed, record the corresponding scheduling adjustment information and available robot identification as part of the resource evaluation results, reflecting the availability of cleaning robot resources for new tasks; Then consider the availability of cleaning tools: first conduct a tool inventory query: connect to the hotel's cleaning tool inventory management system to check the inventory quantity and storage location of various cleaning tools; among them, cleaning tools include brooms, mops, cleaning agents for different purposes, and rags; check the cleaning tools and consumables required for the new cleaning task, and determine whether the cleaning tools are sufficient based on the inventory information to obtain the inventory quantity of the cleaning tools; tool status check: for existing cleaning tools, find out whether they are in good usable condition through regular maintenance records or on-site inspections, and obtain the status information of the cleaning tools; comprehensively consider the inventory quantity and status information of the cleaning tools to determine whether they can meet the needs of the new cleaning tasks and obtain the availability evaluation results; if the inventory of various required cleaning tools is sufficient and in good condition, the cleaning tool availability is determined to be good; if there is a shortage of key tools or a large number of tools are faulty and cannot be used, the cleaning tool availability is determined to be poor.

[0022] Preferably, the service module to be called is determined according to the service type, and the service parameters in the service request are passed to the selected service module, wherein the service parameters include the room number A and the type of food of the target room in the meal delivery service; the room number B and the cleaning requirements, such as deep cleaning or daily cleaning, of the target room in the cleaning service; and the room number C and the maintenance items of the target room in the maintenance service, so that the service module can accurately perform the task; then the service module activates the room service task and feeds back the task start information to the service request classification module; After the service module activates the room service task, it intelligently selects the most suitable service robot to perform the task based on the service request and the current location and load of the service robot, obtains the selected robot, and then assigns the task. After the task assignment is completed, the service request classification module sends the task instruction to the selected robot and updates the order status to "assigned"; After receiving the task instruction, the selected robot obtains the latest internal map data of the hotel from the hotel's central control system, and uses the path planning algorithm to calculate the optimal path from the current location to the target room; based on the optimal path, the selected robot uses built-in sensors to detect the surrounding environment in real time, avoid obstacles, and ensure safe driving; after the selected robot arrives at the target room, it notifies the guest of its arrival through voice prompts or screen display information, and interacts with the guest according to the type of service to complete the task; after the service is completed, the robot asks the guest about satisfaction and records feedback; after the selected robot completes the task, it returns to the charging station or standby area according to the return instruction of the central control system; the central control system updates the order status as completed or incomplete, and triggers a customer satisfaction survey, or triggers a task remediation plan to re-perform room service until the central control system updates the order status to completed.

[0023] Preferably, the service request classification module continuously monitors the execution status of the room service tasks. If an abnormality occurs, such as a food delivery robot malfunctioning or a cleaning robot failing to arrive on time, an abnormal alarm is generated and an abnormal dispatch instruction is generated. The abnormal dispatch instruction is sent to the logistics personnel, who manually handle the abnormal situation to ensure that the service can be completed smoothly or remedial measures are taken.

[0024] Preferably, the process of selecting a robot to perform room service tasks is as follows: For meal delivery tasks: determine the dish and drink information to be delivered according to the task instructions, and use the built-in hotel map data and path planning algorithm to plan the meal pickup path and meal delivery path from the current location to the restaurant. The path planning algorithm is the A* algorithm. The hotel map data contains the layout information of each floor of the hotel: the location of the rooms, the direction of the corridors, the distribution of stairs and elevators, and the specific coordinates of the restaurant, power distribution room, and storage area. For cleaning tasks: determine whether the tools and consumables required for cleaning are complete and can be used normally. If there are any missing or faulty tools, generate a cleaning tool missing report and report it to the backend server to obtain replenishment or repair arrangements. Consumables include brooms, mops, and detergents. Then, the path planning algorithm is also used to plan the travel path from the current position to the target room in combination with the target room, avoiding crowded areas and areas where other operations are being carried out. Then, the built-in sensors are used to sense the surrounding environment information in real time, and the surrounding environment information is fused and corrected with the hotel's internal map data to construct an environmental map, which is used to select the robot to avoid obstacles and follow the food delivery path or travel path to the target room. The built-in sensors include laser radar and cameras, which ensure that the selected robot reaches the target room safely and efficiently. Obstacles include static furniture and walls as well as dynamic personnel and other service equipment. It should be further explained that in the specific implementation process, the specific service operations are performed as follows: For the meal delivery task: after arriving at the restaurant, the identity information and meal pickup code are displayed on the screen, and the restaurant staff interacts to obtain the meal and place it on the meal delivery device; after reconfirming that the meal delivery information is correct, go to the target room, and when arriving at the door of the room, notify the guest through voice prompts or screen display information, where the voice prompt includes: Hello, your meal has been delivered; place the meal in a suitable location in the room according to the guest's instructions or the preset standard process, and provide services to assist the guest in opening the meal packaging; For cleaning tasks: after arriving at the target room, reconfirm the cleaning task requirements and key areas; perform floor cleaning in sequence according to the standard cleaning process, including sweeping, mopping, furniture wiping, and bathroom cleaning, to ensure that the room meets the cleaning standards; during the cleaning process, in order to protect the guests' property and privacy, report to the backend server when encountering special situations and wait for instructions, such as the special situation of finding valuables left behind by the guests; After completing the meal delivery or cleaning task, the service robot checks the task execution status and obtains the task completion status; for meal delivery tasks, it confirms that the meal has been delivered completely and the guest has no other needs; for cleaning tasks, it checks whether each area of ​​the room is clean and whether the cleaning tools have been put back in place; Organize the execution information during the task execution process, including the task start time, end time, actual execution path, and whether any abnormal situations are encountered, such as obstacles on the path causing temporary diversion, or special feedback from guests on the food during delivery; The task completion status and execution information are sent to the central control system through the wireless communication module, which is used by the central control system to update the order status, conduct service quality assessment and arrange subsequent tasks; if a communication failure occurs during the reporting process, the service robot will retry or record information according to the preset fault handling mechanism and wait for communication to be restored before sending again.

[0025] Preferably, the built-in hotel map data and path planning algorithm are used to plan the meal pickup path from the current location to the restaurant. The process is as follows: Based on the Wi-Fi signal strength fingerprint positioning, the current location is clarified, including the current floor of the hotel and the specific coordinate point nearby; Parse the specific location information of the restaurant from the task instructions sent by the backend server, or determine the target location based on the restaurant location coordinates pre-marked in the hotel map data; including knowing that the restaurant is located in a specific area on the third floor of the hotel, its coordinates are (x1, y1), and the robot's current location coordinates are (x0, y0); Based on the A* algorithm, the hotel map data is gridded and the hotel space is divided into grid cells of the same size. Each grid cell is marked as passable, including the interior of the room and the area where the wall is located as impassable, and the corridor area as passable. At the same time, each grid cell is assigned a corresponding coordinate index to facilitate subsequent algorithm calculation and node representation; Based on the robot's built-in sensors, it obtains real-time information about the surrounding environment and fixed obstacles marked in the map data, and marks the locations of obstacles in the map model; Create a priority queue to store the nodes to be processed. Initially, add the node corresponding to the current position of the service robot to the queue to obtain the priority queue: open_set; establish a dictionary to record the predecessor node of each node, which is initially empty, and obtain the predecessor node dictionary: came_from; record the actual cost from the current position of the robot, that is, the starting node to each node. Initially, the cost of all nodes is set to infinity, and only the cost of the starting node is set to 0, and the actual cost dictionary is obtained: g_score; record the estimated total cost from the starting node to each node. Initially, the cost of all nodes is set to infinity, and the cost of the starting node is set to the value of the heuristic function, and the estimated total cost dictionary is obtained: f_score; Loop operation: Take the node current with the lowest cost from the open_set priority queue. This cost is a combination of the actual cost from the starting node to the current node and the estimated cost to the target node, measured by f_score; if the current node is the target node, that is, the node corresponding to the location of the restaurant, call the reconstruct_path function, use the came_from dictionary to trace back from the target node to the starting node, build a complete meal-picking path and return the path information; if current is not the target node, traverse all its neighbor nodes neighbor; for each neighbor node, calculate the actual cost tentative_g_score from current to neighbor, which takes into account the distance between the two points and the difficulty of passing through the area; if tentative_g_score is less than the current g_score of neighbor, it means that a better path to reach the neighbor node has been found. At this time, update the values ​​of the corresponding neighbor node in came_from, g_score and f_score; if the neighbor node is not in open_set, add it to the open_set queue for subsequent evaluation and processing; The operation continues in a loop until the target node is found or the open_set queue is empty. If the open_set queue is empty, it means that no feasible path can be found. At this time, the path planning failure is reported to the backend server.

[0026] Preferably, in the A* algorithm, the calculation formula for evaluating the total cost f(n) of each node is: ; Among them, f(n): represents the total estimated cost from the starting node to the target node through node n. The algorithm determines which node to explore first by constantly comparing the f(n) values ​​of each node, and always prioritizes the node with the smallest f(n) value for expansion and exploration. g(n): is the actual cost from the starting node to the current node n. This cost is the cumulative cost of the path that has been traveled, for example, it is measured by the sum of the actual distances between nodes. Its calculation method will be gradually accumulated and updated as the robot moves. During initialization, the g (starting node) of the starting node is set to 0, and the other nodes are initialized to infinity. Among them, the actual distance is based on the specific map representation and application scenario, including Euclidean distance or Manhattan distance. h(n): is the estimated cost from the current node n to the target node (the node corresponding to the restaurant location); Assume the coordinates of the current node are (x_n, y_n), and the coordinates of the target node (restaurant location) are (x_goal, y_goal), then the Manhattan distance calculation formula is: ; It should be further explained that in the specific implementation process, in the grid map, only the sum of the number of squares that need to be moved from the current node to the target node is considered when moving in the horizontal and vertical directions. It is simple to calculate and can effectively guide the algorithm to search in the target direction. It is selected in indoor path planning scenarios: such as hotel maps are used after being gridded; Assume the current node coordinates are (x_n, y_n), the target node coordinates are (x_goal, y_goal), and the Euclidean distance calculation formula is: ; It is used in scenarios where high path planning accuracy is required and the map is not strictly gridded; When traversing the neighboring nodes neighbor of the current node current, the actual cost tentative_g_score from current to neighbor is calculated, and the calculation formula is: tentative_g_score=g(current)+cost(current,neighbor); Where g(current) is the actual cost calculated from the starting node to the current node current; cost(current,neighbor) represents the cost of moving from the current node current to the neighbor node neighbor. This cost is defined based on the actual situation, such as the actual physical distance between two points based on the precise coordinate distance information on the map; It should be further explained that, in the specific implementation process, or considering the difficulty coefficient of passing through this path: a higher cost weight is given to corridors with more people and narrower corridors; for example, if the distance between the centers of adjacent grid cells is recorded as d in the map, and there is no special difficulty coefficient, cost(current,neighbor) is directly d; if the corridor congestion is considered, the congestion coefficient is k (k>=1, the more severe the congestion, the larger the k value), then cost(current,neighbor) is expressed as d*k.

[0027] The meal pickup route technology solution is based on Wi-Fi signal strength fingerprint positioning, A* algorithm, real-time sensor perception and map fusion correction, etc., which realizes precise positioning and efficient route planning. It can quickly determine the starting position and select the optimal route. At the same time, it can flexibly respond to environmental changes and avoid obstacles in real time, ensuring safe and efficient driving, reducing energy consumption and equipment loss, and improving resource utilization efficiency. Ultimately, it improves service timeliness and customer satisfaction, which helps the hotel improve its overall service quality and competitiveness.

[0028] It should be further explained that, in the specific implementation process, a method for using the intelligent mobile room service workstation order listening system includes the following steps: Step 1: The customer initiates a service request through the front-end application, verifies the validity and feasibility of the service request, checks whether the target room number is legal and within the hotel room number range, extracts keywords based on the service request, matches them with the preset service type keyword library, determines whether the service type is a type that the system can recognize, obtains the room number check result and service type, and determines the service module to be called based on the determined service type; Step 2: Conduct resource evaluation, evaluate the food delivery task resources and cleaning task resources separately, obtain the food delivery resource evaluation results and the cleaning resource evaluation results, and obtain the overall resource evaluation results by integrating the food delivery resource evaluation results and the cleaning resource evaluation results; plan the general process of task execution in combination with the internal layout of the hotel and the current location information of each resource, and obtain the task plan; generate task instructions based on the resource evaluation results and the task plan; send the generated task instructions to the selected service module, and update the order status to "allocated"; Step 3: After receiving the task instruction, the service module activates the room service task and feeds back the task start information to the service request classification module. It plans the meal pickup path and meal delivery path for the meal delivery task according to the service type, and plans the travel path for the cleaning task. The selected robot obtains the latest hotel internal map data from the hotel's central control system, and uses the path planning algorithm to calculate the optimal path from the current location to the target room. The selected robot reaches the target room based on path planning and navigation. Step 4: After the robot arrives at the target room, it notifies the guest of its arrival through voice prompts or screen display information, and interacts with the guest according to the service type to complete the corresponding room service task. After the service is completed, the robot asks the guest about satisfaction and records feedback information; After the selected robot completes the task, it returns to the charging station or standby area according to the return instruction of the central control system; The central control system updates the order status as completed or incomplete, and triggers a customer satisfaction survey, or triggers a task remediation plan to re-perform room service until the central control system updates the order status to completed.

[0029] It should be further explained that, in the specific implementation process, service efficiency has been improved, automated processing has reduced manual intervention, the system receives customer requests through the front-end application, and the back-end server automatically performs a series of automated processes such as voice recognition, natural language processing, service request classification, resource evaluation, task planning and instruction generation, as well as service module execution and robot automatic navigation, which greatly reduces the tedious links of manual answering of calls, recording orders, and assigning tasks, significantly shortens service response time, and improves overall service efficiency; It can accurately determine the required resources according to the service type, comprehensively evaluate the resource status, intelligently select the most suitable service robot or personnel to perform the task, avoid idle or over-allocated resources, ensure that various tasks are executed efficiently, and achieve intelligent task allocation and resource scheduling; Using built-in hotel map data and advanced path planning algorithms, the service robot plans the optimal path from the current location to the target location, reducing the time wasted by the robot during the movement, improving the speed of service execution, and also reducing energy consumption, improving service efficiency and automation and intelligent functional requirements; Accurate resource assessment and intelligent scheduling mechanisms enable the rational allocation and full utilization of resources such as restaurants, cleaning tools, and robots, avoiding waste and idleness of resources, improving the overall utilization efficiency of hotel operating resources, and indirectly reducing operating costs; the automation system and robots have relatively high accuracy in performing tasks, which can reduce possible errors in manual services. At the same time, the standardized operation of robots helps reduce the loss of items caused by improper operation, thereby reducing related costs.

[0030] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field and related fields without creative work should fall within the scope of protection of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention are implemented according to the conventional means in the field unless otherwise specified and limited.

Claims

1. An intelligent mobile room service workstation order listening system, characterized in that: It includes the following components: Front-end application, used to receive voice or text requests from customers, and supports both voice and text input methods; The back-end server includes a speech recognition module, a natural language processing module and a service request classification module; the speech recognition module is used to convert the customer's speech request into a request text; the natural language processing module is used to understand the content of the request text or text request to obtain the service request; the service request classification module is used to classify the service request and call the corresponding service module for processing; Service robots: Service robots are used to receive task instructions from the backend server. Based on the task instructions, the selected robot performs room service tasks, including meal delivery and cleaning. After the service is completed, the service robot reports the task completion status and abnormal information to the central control system. The feedback module is used to send confirmation information to customers after the task is completed to ensure service satisfaction.

2. The intelligent mobile room service workstation order listening system according to claim 1 is characterized by: The service request classification module is used to classify service requests and call the corresponding service module for processing. The implementation process is as follows: The service request classification module first obtains the service request parsed by the natural language processing module and verifies the validity and feasibility of the service request, including: checking whether the room number of the target room is legal and within the hotel room number range, whether the service type is a recognizable type preset by the system, and obtaining the room number check result and service type respectively; based on the room number check result, according to the service type, the required resources are determined and evaluated to obtain the resource evaluation result; combined with the internal layout of the hotel and the current location information of each resource, the general process of task execution is planned to obtain the task plan; based on the resource evaluation results and task plan, task instructions are generated; task instructions include a detailed description of the task, resource identification for executing the task, the execution order and time requirements of the task, the task location and task standard details.

3. The intelligent mobile room service workstation order listening system according to claim 2 is characterized by: Whether the service type is a preset identifiable type of the system, the process is as follows: Extract keywords based on the service request to obtain keywords, and then match the extracted keywords with the preset service type keyword library to obtain the service type; determine the service module to be called according to the service type; the service module includes a food delivery robot, a cleaning robot, and a maintenance service scheduling robot; For food delivery services, the food delivery robot corresponding to the restaurant or food delivery is called, and the food delivery robot is responsible for interacting with the restaurant system to arrange meal preparation, meal pickup route planning, and task setting for the food delivery robot; For cleaning services, cleaning robots are called to be responsible for cleaning task scheduling, cleaning tool preparation, and cleaning process control and execution; for maintenance services, maintenance service scheduling robots are called to contact maintenance personnel, deploy maintenance tools, and arrange maintenance tasks.

4. The intelligent mobile room service workstation order listening system according to claim 3 is characterized by: According to the service type, the required resources are determined and evaluated to obtain the resource evaluation results. The process is as follows: for food delivery tasks, the restaurant's current order pressure, food inventory, and the status of the available food delivery robots are evaluated. The food delivery robot status includes the current location of the food delivery robot, the current power of the food delivery robot, and the current load of the food delivery robot to obtain the food delivery resource evaluation results; for cleaning tasks, the cleaning robot's scheduling and the availability of cleaning tools are considered; The cleaning resource assessment result is obtained; the meal delivery resource assessment result is combined with the cleaning resource assessment result to obtain the resource assessment result.

5. The intelligent mobile room service workstation order listening system according to claim 4 is characterized by: Determine the service module to be called according to the service type, and pass the service parameters in the service request to the selected service module. The service parameters include the room number A and the type of food for the meal delivery service; the room number B and the cleaning requirements for the target room for the cleaning service, and the room number C and the repair items for the target room; then the service module activates the room service task and feeds back the task start information to the service request classification module; After the service module activates the room service task, it intelligently selects the most suitable service robot to perform the task based on the service request and the current location and load of the service robot. After the selected robot is assigned, the service request classification module sends the task instruction to the selected robot and updates the order status to "assigned". After receiving the task instruction, the selected robot obtains the latest internal map data of the hotel from the hotel's central control system, and uses the path planning algorithm to calculate the optimal path from the current location to the target room; based on the optimal path, the selected robot uses built-in sensors to detect the surrounding environment in real time and avoid obstacles; after the selected robot arrives at the target room, it notifies the guest of its arrival through voice prompts or screen display information, and interacts with the guest according to the type of service to complete the task; after the service is completed, the robot asks the guest about satisfaction and records feedback; after the selected robot completes the task, it returns to the charging station or standby area according to the return instruction of the central control system; the central control system updates the order status as completed or incomplete, and triggers a customer satisfaction survey, or triggers a task remediation plan to re-perform room service until the central control system updates the order status as completed.

6. The intelligent mobile room service workstation order listening system according to claim 5, characterized in that: The service request classification module continuously monitors the execution status of room service tasks. If an abnormality occurs, such as a food delivery robot failure or a cleaning robot failing to arrive on time, an abnormal alarm is generated and an abnormal dispatch instruction is generated. The abnormal dispatch instruction is sent to the logistics personnel, who manually handle the abnormal situation.

7. The intelligent mobile room service workstation order receiving system according to claim 6, characterized in that: The process of selecting a robot to perform room service tasks is as follows: For meal delivery tasks: determine the dish and drink information to be delivered according to the task instructions, and use the built-in hotel map data and path planning algorithm to plan the meal pickup path and meal delivery path from the current location to the restaurant. The path planning algorithm is the A* algorithm. For cleaning tasks: determine whether the tools and consumables required for cleaning are complete and can be used normally. If they are missing or faulty, generate a cleaning tool missing report and report it to the backend server to obtain replenishment arrangements or repair arrangements; then use the path planning algorithm to plan the travel path from the current position to the target room in combination with the target room; Then, the built-in sensors are used to perceive the surrounding environment information in real time, and the surrounding environment information is integrated and corrected with the hotel's internal map data to construct an environmental map, which is used to select the robot to avoid obstacles and follow the delivery path or travel path to the target room; the built-in sensors include lidar and cameras.

8. The intelligent mobile room service workstation order receiving system according to claim 7, characterized in that: Using the built-in hotel map data and path planning algorithm, plan the route from the current location to the restaurant to pick up the food. The process is as follows: Based on the Wi-Fi signal strength fingerprint positioning, the current location is clarified, and the specific location information of the restaurant is parsed from the task instructions sent by the back-end server, or the target location is determined based on the restaurant location coordinates pre-marked in the hotel map data; Based on the A* algorithm, the hotel map data is gridded and the hotel space is divided into grid cells of the same size. Each grid cell is marked as passable or not, and a corresponding coordinate index is assigned to each grid cell. Based on the robot's built-in sensors, it obtains real-time information about the surrounding environment and fixed obstacles marked in the map data, and marks the locations of obstacles in the map model; Create a priority queue to store the nodes to be processed. Initially, add the node corresponding to the current position of the service robot to the queue to obtain the priority queue: open_set; create a dictionary to record the predecessor node of each node, which is initially empty, and obtain the predecessor node dictionary: came_from; record the actual cost from the current position of the robot, that is, the starting node to each node. Initially, the cost of all nodes is set to infinity, and only the cost of the starting node is set to 0, and obtain the actual cost dictionary: g_score; Record the estimated total cost from the starting node to each node. Initially, the cost of all nodes is set to infinity, and the cost of the starting node is set to the value of the heuristic function. Get the estimated total cost dictionary: f_score; Loop operation: take out the node current with the lowest cost from the open_set priority queue; if the current node is the target node, that is, the node corresponding to the location of the restaurant, call the reconstruct_path function, use the came_from dictionary to trace back from the target node to the starting node, build a complete meal-picking path and return the path information; if current is not the target node, traverse all its neighbor nodes neighbor; for each neighbor node, calculate the actual cost tentative_g_score from current to neighbor; if tentative_g_score is less than the current g_score of neighbor, update the values ​​of the corresponding neighbor node in came_from, g_score and f_score; if the neighbor node is not in open_set, add it to the open_set queue; continue the loop operation until the target node is found or the open_set queue is empty.

9. The intelligent mobile room service workstation order receiving system according to claim 8, characterized in that: In the A* algorithm, the total cost f(n) for evaluating each node is calculated as: ； Among them, f(n): represents the total estimated cost from the starting node to the target node through node n, g(n): is the actual cost from the starting node to the current node n; h(n): is the estimated cost from the current node n to the target node; Assume the current node coordinates are (x_n, y_n) and the target node coordinates are (x_goal, y_goal), then the Manhattan distance calculation formula is: ； Assume the current node coordinates are (x_n, y_n), the target node coordinates are (x_goal, y_goal), and the Euclidean distance calculation formula is: ； When traversing the neighboring nodes neighbor of the current node current, the actual cost tentative_g_score from current to neighbor is calculated, and the calculation formula is: tentative_g_score=g(current)+cost(current,neighbor); Where g(current) is the actual cost calculated from the starting node to the current node current; cost(current,neighbor) represents the cost of moving from the current node current to the neighbor node neighbor.

10. A method for listening to orders at an intelligent mobile room service workstation, characterized in that: The steps include: Step 1: The customer initiates a service request through the front-end application, verifies the validity and feasibility of the service request, checks whether the target room number is legal and within the hotel room number range, extracts keywords based on the service request, matches them with the preset service type keyword library, determines whether the service type is a type that the system can recognize, obtains the room number check result and service type, and determines the service module to be called based on the determined service type; Step 2: Conduct resource evaluation, evaluate the food delivery task resources and cleaning task resources separately, obtain the food delivery resource evaluation results and the cleaning resource evaluation results, and combine the food delivery resource evaluation results and the cleaning resource evaluation results to obtain the overall resource evaluation results; combine the internal layout of the hotel and the current location information of each resource to plan the general process of task execution and obtain the task plan; generate task instructions based on the resource evaluation results and task plan; send the generated task instructions to the selected service module, and update the order status to "allocated"; Step 3: After receiving the task instruction, the service module activates the room service task and feeds back the task start information to the service request classification module. It plans the meal pickup path and meal delivery path for the meal delivery task according to the service type, and plans the travel path for the cleaning task. The selected robot obtains the latest hotel internal map data from the hotel's central control system, and uses the path planning algorithm to calculate the optimal path from the current location to the target room. The selected robot reaches the target room based on path planning and navigation. Step 4: After the robot arrives at the target room, it notifies the guest of its arrival through voice prompts or screen display information, and interacts with the guest according to the service type to complete the corresponding room service task. After the service is completed, the robot asks the guest about satisfaction and records feedback information; After the selected robot completes the task, it returns to the charging station or standby area according to the return instruction of the central control system; The central control system updates the order status as completed or incomplete, and triggers a customer satisfaction survey, or triggers a task remediation plan to re-perform room service until the central control system updates the order status to completed.

Images