Pipeline system specialized for drone

Abstract

The following disclosure relates to a pipeline system specialized for a drone and an operating method thereof. The pipeline system may: receive an input of a user related to a target artificial neural network model; detect a list of drones connected to the pipeline system specialized for the drone; receive movement paths and sensor data of the drones on the basis of the list of the drones; analyze the movement paths and the sensor data of the drones; train the target artificial neural network model on the basis of the input of the user, the movement paths, and the sensor data; and select valid target drones from the list of the drones to transmit the trained target artificial neural network model to the valid target drones.

Description

Pipeline system specialized for drones

[0001] The disclosure below relates to a pipeline system specialized for drones and its operating method.

[0002] Conventional MLOps platforms provide convenient and automated methods for building and operating AI, targeting non-experts without specialized knowledge in AI and computer systems.

[0003] However, these platforms require users to 1) understand the type, meaning, and usability of their data, 2) plan a target AI service based on the data, and 3) select the type of AI model to be utilized in the service before using the platform.

[0004] This approach is still perceived as very difficult not only for those unfamiliar with AI, but also for those unfamiliar with data and IT, and increases the barrier to entry for utilizing MLOps platforms.

[0005] The embodiments aim to provide a new type of pipeline system that provides an environment for automatically learning an artificial neural network model based on a drone-specific pipeline system when users develop an artificial neural network model specialized for drones.

[0006] The problems to be solved by the present invention are not limited to the problems described above, and problems not mentioned can be clearly understood by a person having ordinary skill in the technical field to which the present invention pertains from this specification and the attached drawings.

[0007] An operation of a pipeline system specialized for a drone according to one embodiment may include a step of receiving a user's input related to a target artificial neural network model, a step of detecting a list of drones connected to the pipeline system specialized for the drone, a step of receiving movement paths and sensor data of drones based on the drone list, a step of analyzing the movement paths and the sensor data of the drones, a step of learning the target artificial neural network model based on the user's input, the movement paths and the sensor data, and a step of selecting valid target drones from the drone list and transmitting the target artificial neural network model, for which learning has been completed, to the valid target drones.

[0008] The step of analyzing the movement path and the sensor data may further include a step of analyzing the movement path and the sensor data according to the user's input and providing a visualization of the movement path and the sensor data of the drones.

[0009] The step of analyzing the movement path and the sensor data may further include a step of analyzing the movement path and the sensor data according to the user's input to provide at least one of automation of mission assignment of the drones, load balancing, and service auto-scaling.

[0010] The step of receiving input from the user may further include the step of providing an interface for receiving editing input from the user for nodes within the pipeline system.

[0011] The step of providing the above interface may include a step of providing the user with an interface for checking tasks considering the mobility of the drone for nodes in the pipeline, a step of providing an interface for modifying detailed settings of each of the nodes, and a step of providing an interface for deploying the target artificial neural network.

[0012] The step of detecting the above drone list may include a step of checking the status of drones connected to the system in the drone list, and controlling drones that are unable to fly among the drones to stop their flight.

[0013] The step of receiving the movement path and sensor data of the drones may include a step of labeling the sensor data using an automated labeling model included in the pipeline system based on the user's input.

[0014] A pipeline system specialized for a drone according to one embodiment may include an interface providing module for receiving a user's input related to a target artificial neural network model, a data classification module for detecting a list of drones connected to the pipeline system specialized for drone learning and receiving movement paths and sensor data of drones based on the list of drones, a data analysis module for analyzing the movement paths and the sensor data, and an artificial neural network model learning module for learning the target artificial neural network model based on the user's input, the movement path, and the sensor data, wherein the data analysis module selects valid target drones from the drone list, and the interface providing module may cause the artificial neural network learning module to transmit the target artificial neural network model, on which learning has been completed, to the valid target drones in response to the user's input.

[0015] According to one embodiment, a drone may include a memory including instructions, a sensor unit, a communication unit that communicates with a server connected to a pipeline system specialized for learning the drone, transmits movement paths and sensor data collected from the sensor unit, receives a target artificial neural network model, receives a user's task, and transmits the task result, and a processor, wherein the instructions, when executed by the processor, cause the drone to update the artificial neural network model included in the drone using the target artificial neural network model received from the communication unit, perform the user's task received from the communication unit, and retrain the artificial neural network model based on data collected from the sensor unit.

[0016] FIG. 1 is a flowchart illustrating an operation method of a pipeline system specialized for a drone according to one embodiment.

[0017] FIG. 2 is a block diagram of a pipeline system (200) specialized for a drone according to one embodiment.

[0018] FIG. 3 is a schematic diagram illustrating the operation of a drone and pipeline system according to one embodiment.

[0019] FIG. 4 is a diagram illustrating an example of a pipeline according to one embodiment.

[0020] FIG. 5 is a block diagram illustrating an electronic device according to one embodiment.

[0021] Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be modified and implemented in various forms. Therefore, the actual implementation is not limited to the specific embodiments disclosed, and the scope of this specification includes modifications, equivalents, or alternatives within the technical concepts described in the embodiments.

[0022] Although terms such as "first" or "second" may be used to describe various components, these terms should be interpreted solely to distinguish one component from another. For example, a first component may be referred to as a second component, and similarly, a second component may also be referred to as a first component.

[0023] When it is said that a component is "connected" to another component, it should be understood that it may be directly connected or connected to that other component, but there may also be other components in between.

[0024] Singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, the terms "comprises" or "has" should be understood to indicate the presence of a described feature, number, step, operation, component, part, or combination thereof, but not to exclude the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0025] In this document, each of the phrases "A or B", "at least one of A and B", "at least one of A or B", "A, B, or C", "at least one of A, B, and C", and "at least one of A, B, or C" may include any one of the items listed together in that phrase, or all possible combinations thereof.

[0026] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art. Terms defined in commonly used dictionaries should be interpreted to have a meaning consistent with their meaning in the context of the relevant technology, and will not be interpreted in an idealized or overly formal sense unless explicitly defined herein.

[0027] The embodiments can be implemented in various forms of products, such as personal computers, laptop computers, tablet computers, smartphones, televisions, smart home appliances, intelligent vehicles, kiosks, and wearable devices. Hereinafter, the embodiments will be described in detail with reference to the attached drawings. In the description with reference to the attached drawings, identical components are assigned the same reference numerals regardless of the drawing numbers, and redundant descriptions thereof will be omitted.

[0028]

[0029] FIG. 1 is a flowchart illustrating an operation method of a pipeline system specialized for a drone according to one embodiment.

[0030] The operations of FIG. 1 may be performed in the order and manner illustrated, but the order of some operations may be changed or some operations may be omitted without departing from the spirit and scope of the illustrated embodiment. Multiple operations illustrated in FIG. 1 may be performed in parallel or simultaneously.

[0031] For convenience of explanation, steps (110 to 150) are described as being performed using a drone-specific pipeline system (200, hereinafter referred to as pipeline system (200)) illustrated in FIG. 2. However, these steps (110 to 150) may be used via any other suitable electronic device and within any suitable system.

[0032]

[0033] FIG. 2 is a block diagram of a pipeline system (200) specialized for a drone according to one embodiment.

[0034] Referring to FIG. 2, a pipeline system (200) specialized for a drone according to one embodiment may include an interface provision module (210), a data classification module (220), a data analysis module (230), and an artificial neural network learning module (240). The term "module" used in this document may mean, for example, a unit including one or a combination of two or more of hardware, software, or firmware. "Module" may be used interchangeably with terms such as unit, logic, logical block, component, or circuit. A "module" may be a minimum unit of an integrally formed component or a part thereof. A "module" may also be a minimum unit performing one or more functions or a part thereof. A "module" may be implemented mechanically or electronically. For example, a "module" may include at least one of an application-specific integrated circuit (ASIC) chip, field-programmable gate arrays (FPGAs), or programmable-logic device, known or to be developed in the future, that performs certain operations.

[0035] FIG. 3 is a schematic diagram illustrating the operation of a drone and pipeline system according to one embodiment.

[0036] The description with reference to FIG. 1 and FIG. 2 can be equally applied to FIG. 3, and overlapping content can be omitted.

[0037] Referring to FIG. 3, the pipeline system (200) can receive drone status and drone sensor data from drones (300). Furthermore, a user can directly control the drones (300) or transmit missions / commands, etc. to the drones (300) via the pipeline system (200). The pipeline system (200) can classify data and label the data via the data classification module (220).

[0038] Although not shown in the drawing, a drone (300) according to one embodiment may include a memory, a sensor unit, a communication unit, and a processor including instructions.

[0039] The communication unit can communicate with a server connected to the pipeline system (200), transmit movement path and sensor data collected from the sensor unit, receive a target artificial neural network model, receive a user's task, and transmit the task result.

[0040] The communication unit may include, but is not limited to, a short-range wireless communication unit, a Bluetooth communication unit, a BLE (Bluetooth Low Energy) communication unit, a near field communication unit, a WLAN (Wi-Fi) communication unit, a Zigbee communication unit, an infrared (IrDA, infrared Data Association) communication unit, a WFD (Wi-Fi Direct) communication unit, an UWB (ultra wideband) communication unit, an Ant+ communication unit, etc.

[0041] The wireless communication unit may include, but is not limited to, a cellular network communication unit, an Internet communication unit, a computer network (e.g., a local area network or wide area network) communication unit, etc. The wireless communication unit may also use subscriber information (e.g., an International Mobile Subscriber Identity (IMSI)) to identify and authenticate electronic devices within the communication network.

[0042] The drone can execute instructions via a processor (e.g., NPU), update the artificial neural network model included in the drone using the target artificial neural network model received from the drone's communication unit, perform the user's task received from the communication unit, and retrain the artificial neural network model on its own based on data collected from the sensor unit.

[0043] The data analysis module (230) can visualize and display data based on data received from the data classification module (220) and user input, and perform tasks such as automation of mission deployment, load balancing, and service auto-scaling. The data analysis module (230) can perform tasks such as automation of drone deployment in conjunction with an artificial neural network learning module (240).

[0044] The artificial neural network learning module (240) can create / update a target artificial neural network model based on the user's input input into the pipeline system (200) and data received from the data classification model. For example, if the user manipulates the pipeline system to learn an artificial neural network model for servicing a "drone taxi," the artificial neural network learning module (240) can create an artificial neural network model suitable for a "drone taxi" or update an existing artificial neural network model.

[0045] The artificial neural network learning module (240) can automatically transmit the generated or updated artificial neural network model to the drone for update.

[0046] The drone-specific pipeline system (200) can operate in conjunction with MLOps. MLOps, a term combining machine learning and operations, refers to maintaining, managing, and monitoring artificial neural network models to ensure their continuous and stable deployment in a production environment. MLOps integrates the development and operation of artificial neural network models, enabling automated maintenance, management, and operation of ML systems. MLOps encompasses not only the development of artificial neural network models, but also the stages of data collection and analysis, as well as the learning and deployment process—in other words, the entire AI life cycle.

[0047] Conventional MLOps platforms can provide convenient, automated AI deployment and operation methods for non-experts without specialized knowledge in AI and computer systems. However, because these platforms aim to provide a pipeline for AI learning / inference and service operation utilizing fixed-location server computers, their use in mobile drone-based AI learning / inference and service operation can be challenging.

[0048] A user refers to a subject who receives services from an MLOps system through a user terminal, and may be referred to as a client, customer, etc. In one embodiment, a user terminal is a device that receives a predetermined command from a user and executes a corresponding operation, and may be a digital device that includes an audio output function, a wired / wireless communication function, or other functions. In one embodiment, a terminal may be a concept that includes all digital devices equipped with a memory means and a microprocessor equipped with a computing capability, such as a tablet PC, a smartphone, a personal computer (e.g., a laptop computer, etc.), a smart TV, a mobile phone, a navigation system, a web pad, a PDA, a workstation, etc.

[0049] According to one embodiment, a user terminal may refer to any user device capable of installing and executing an application related to the MLOps system. In this case, the user terminal can perform all operations of the service, including configuring service screens, entering data, transmitting and receiving data, and storing data, under the control of the application. The application is implemented for use in both PC and mobile environments, and may be implemented as an independently operating program or configured as an in-app within a specific application, enabling operation within said specific application.

[0050] Below, the pipeline system (200) is described in detail with reference to FIGS. 1 to 3.

[0051] In step (110), an interface providing module (210) according to one embodiment may provide an artificial neural network model generation pipeline service to a user. The interface providing module (210) may receive user input related to a target artificial neural network model.

[0052] According to one embodiment, the target artificial neural network model may refer to an artificial neural network model that the user wishes to build. The user may manipulate a pipeline composed of drone-related terms rather than specialized terms related to artificial intelligence. The interface provision module (210) may receive input regarding the user's pipeline manipulation and provide interfaces for generating the target artificial neural network model.

[0053] For example, a user may be unfamiliar with neural network types such as CNNs, RNNs, and GANs. In this case, instead of directly inputting an RNN or LSTM, the user can edit a pipeline provided in the form of blocks and nodes connecting them to create a target artificial neural network model.

[0054] In step (120), the data classification module (220) according to one embodiment can detect a list of drones connected to the pipeline system (200). The data classification module (220) can detect a list of drones controllable in the pipeline system (200) based on a user's input.

[0055] Drones (or Unmanned Aerial Vehicles (UAVs)) can use a variety of sensors to gather information and monitor their environment. The sensors mounted on a drone can vary depending on its purpose and application. For example, a drone may be equipped with GPS, an IMU, a barometer, a magnetometer, a LIDAR, a camera, an RGB sensor, a thermal imaging sensor, and a radar sensor. Furthermore, depending on the drone's purpose, it may be equipped with a variety of sensors.

[0056] The data classification module (220) can check the status of drones (300) connected to the system from the drone list, and control drones (300) that are unable to fly among the drones (300) to stop their flight.

[0057] Drones can lose control due to a variety of factors, from technical failures to external interference. For example, a drone can lose control if communication with its operator is interrupted due to signal interference, signal weakening, or exceeding the communication range. Control can also be lost due to GPS signal issues, battery failure, hardware defects, software defects, environmental factors, and other factors. In this case, the data classification module (220) can exclude drones from the list of controllable drones if no data is received or if poor data is received from the list of drones the user wishes to control. The data classification module (220) can then request data from the controllable drones (300) to receive movement path and sensor data.

[0058] In step (130), the data classification module (220) according to one embodiment may receive movement paths and sensor data of drones (300) based on a drone list. The data classification module (220) may label the sensor data using an automated labeling model included in the pipeline system (200) based on a user's input.

[0059] The data classification module (220) can categorize and classify the received movement path and sensor data. Categorizing the data received from the drone sensor may mean organizing and labeling information obtained from various sensors of the drone. The specific categorization strategy may vary depending on the type of sensor equipped on the drone and the mission objective. The data classification module (220) can identify the type of sensor equipped on the drone and determine the importance of the data based on the drone's mission objective. Thereafter, the data classification module (220) can classify the data into time-based categories (e.g., takeoff / landing data, flight data), space-based categories (e.g., geographical data, altitude data), and mission-based categories (e.g., road defect data, temperature / humidity data).

[0060] The data classification module (220) can label data to enable machine learning algorithms to be used for data analysis. For example, when collecting images for object detection, the images can be labeled with categories such as "person," "vehicle," and "building."

[0061] The process of automatically labeling data, often referred to as "annotation" or "labeling," can be crucial for training machine learning models to recognize patterns or objects in sensor data.

[0062] For example, supervised learning with manual annotation can train a machine learning model on a labeled dataset where each data point is associated with a specific label. The labeling process is typically manual, but can be performed semi-automatically using human annotators. Annotators can review sensor data and assign appropriate labels to objects or features of interest. In addition to the described embodiments, object detection and recognition, semantic segmentation, and other tasks can also be used.

[0063] As another example, self-supervised learning can be used as an automatic labeling method. Self-supervised learning is a type of representation learning that aims to obtain good representations from unlabeled data. The model trains by automatically identifying target elements within the input without labels. Therefore, self-supervised learning is also called a pretext task. Self-supervised learning can be performed in at least one of the following ways: intra-sample prediction, which predicts one part of a data sample based on another part, and inter-sample prediction, which predicts relationships between data samples within a batch.

[0064] The data classification module (220) may be equipped with a data storage to facilitate easy information retrieval when data retrieval is required. The data classification module (220) may store and manage categorized data using a structured directory system or database.

[0065] In step (140), the data analysis module (230) according to one embodiment can analyze the movement path and sensor data of the drones (300). The data analysis module (230) can analyze the movement path and sensor data according to a user's input and provide a visualization of the movement path and sensor data of the drones (300).

[0066] The data analysis module (230) can use graphs and charts to quantify and visualize altitude, speed, or sensor data over time in order to monitor trends and patterns related to the movement of the drone.

[0067] The data analysis module (230) can visualize geographic distribution using maps and a Geographic Information System (GIS). The data analysis module (230) can be linked to a GIS platform to provide various information, such as flight paths, points of interest, or environmental conditions.

[0068] The data analysis module (230) can provide a heat map related to the drone's mission. The data analysis module (230) can visualize data using the heat map to indicate data intensity or concentration in a specific area.

[0069] The data visualization method provided by the data analysis module (230) can be selected based on the type of data being processed and the specific objectives of the drone's mission. Data visualization can help interpret data collected from the drone and gain meaningful insights. The visualization method provided by the data analysis module (230) is not limited to the described embodiments and may include various methods, such as 3D modeling.

[0070] In step (150), an artificial neural network learning module (240) according to one embodiment can learn a target artificial neural network model based on the user's input, movement path, and sensor data.

[0071] Equipping drones with artificial neural network models can perform tasks more efficiently and improve drone performance compared to direct human control. For example, drones equipped with artificial neural network models can perform the following tasks:

[0072] Drones can recognize and classify objects in images or videos captured by their cameras. Therefore, drones can identify people, vehicles, buildings, and other objects during surveillance, search, and rescue operations.

[0073] Drones can learn from their flight environment and make decisions based on real-time data, enabling them to navigate autonomously. This could be useful for tasks such as mapping, exploration, or inspection, which require drones to navigate complex and dynamic environments.

[0074] Drones can analyze data collected from sensors such as cameras and multispectral imaging devices to provide information on crop health, pest detection, and yield estimation. Therefore, drones can optimize resource allocation and monitor crop health, contributing to precision agriculture.

[0075] Drones can be trained over time to track specific targets or objects. Therefore, drones can be used for wildlife surveillance, tracking moving vehicles, or search and rescue operations.

[0076] Drones can analyze weather data collected by sensors to predict or monitor changing weather conditions. Therefore, drones can be used for weather research and monitoring, especially in difficult-to-access or dangerous areas.

[0077] The tasks that can be performed by a drone equipped with an artificial neural network model are not limited to the described embodiments.

[0078] According to one embodiment, a data analysis module (230) may analyze movement paths and sensor data based on user input, thereby providing at least one of automated mission deployment of drones (300), load balancing, and service auto-scaling. The data analysis module (230) may be linked to an artificial neural network learning module (240) to enable automated drone deployment based on the analyzed data of a target artificial neural network model.

[0079] Mission deployment automation is a method for optimizing drone deployment for specific missions. The data analysis module (230) analyzes historical data, weather conditions, and other relevant factors to provide information on when and where to deploy drones to maximize efficiency and mission success. For example, it can determine optimal routes, timing, and resource allocation, providing information for drone squadrons performing surveillance, search and rescue missions, and more.

[0080] Load balancing is a method of analyzing real-time data from multiple drones (300) to assess their current workload, battery level, and operational status. The data analysis module (230) can efficiently balance and allocate the drones' workloads based on the data. For example, in scenarios where multiple drones are operating simultaneously, such as a squadron deployed on a large-scale mission, the workload can be evenly distributed to each drone or drone squadron to prevent individual drones from being overburdened or left unused.

[0081] Service auto-scaling can predict demand for drone services based on past usage patterns and mission requirements, adjusting the number of deployed drones or adjusting drone movements to meet demand. For example, in situations like package delivery or surveillance, drone demand can be predicted over a given period and the number of active drones can be automatically adjusted to handle the workload.

[0082] The function of the data analysis module (230) is not limited to the described embodiment, and the data analysis module (230) can perform dynamic path change and adaptation, fault detection and recovery, etc. in conjunction with the artificial neural network learning module (240).

[0083] In step (160), the artificial neural network learning module (240) according to one embodiment can select valid target drones (300) from the drone list and transmit the target artificial neural network model for which learning has been completed to the valid target drones (300). The valid target drone may mean a drone that can automatically update or accept updates, excluding drones with unstable connection due to hardware / software defects with the pipeline system (200) and drones that have turned off the update function.

[0084] The newly trained target artificial neural network model is automatically updated on standby drones, ensuring they remain up-to-date. Drones can enable or disable automatic updates, allowing for selective updates based on the nature of the mission.

[0085]

[0086] FIG. 4 is a diagram illustrating an example of a pipeline according to one embodiment.

[0087] The description referring to FIGS. 1 to 3 can be equally applied to FIG. 4, and overlapping content can be omitted.

[0088] Referring to FIG. 4, a pipeline system (200) specialized for a drone according to one embodiment can build an artificial neural network into a pipeline-based UI and provide it to a user.

[0089] According to one embodiment, the pipeline UI may include a resource preparation UI, a data cleansing UI, a model training UI, and a model deployment UI. However, Figure 4 is merely an example illustrating a pipeline-based UI and is not necessarily limited thereto.

[0090] The pipeline system (200) can perform easy artificial intelligence learning. For example, the pipeline system (200) can provide a GUI environment that allows data-customized artificial intelligence model selection and learning with just a click.

[0091] The pipeline system (200) can reduce model building time and costs. For example, the pipeline system (200) can provide high-performance AI models for language / visual / emotional analysis, thereby reducing AI building time and costs.

[0092] The pipeline system (200) can operate automated artificial intelligence. For example, the pipeline system (200) can support an automatic relearning function for service operation data, enabling advanced, service-tailored artificial intelligence operation.

[0093] The pipeline system (200) can provide efficient model learning. For example, the pipeline system (200) can provide distributed node processing technology capable of distributed learning and an integrated memory function for high-capacity model learning.

[0094] The pipeline system (200) can provide efficient learning schedule management. For example, the pipeline system (200) can provide automatic sequential learning of various parameter settings and rapid learning through accelerated communication between GPUs.

[0095] The pipeline system (200) can provide a convenient development environment. For example, the pipeline system (200) can support integration with existing development environments and tuning of custom code.

[0096] The pipeline system (200) can provide large-scale resource monitoring. For example, the pipeline system (200) can monitor the operating status of GPU resources and manage network interfaces between GPU servers.

[0097] The pipeline system (200) can provide purpose-specific GPU management. For example, the pipeline system (200) can create workspaces to allocate GPU resources and efficiently manage the allocated GPU resources.

[0098] The pipeline system (200) can provide rapid fault detection. For example, the pipeline system (200) can support rapid fault detection and accurate cause analysis through real-time status monitoring.

[0099] The pipeline system (200) can perform the aforementioned operations by receiving user input through the interface provision module (210).

[0100] The interface providing module (210) may provide an interface for receiving editing input from a user for nodes within the pipeline system (200). The interface providing module (210) may provide an interface for the user to check tasks considering the mobility of drones for nodes within the pipeline. The interface providing module (210) may additionally provide nodes related to "drone mobility" in the resource preparation (401) interface and the data refinement (402) interface. Nodes related to "drone mobility" may include nodes that can input data and conditions regarding motion sensors, GPS data, obstacle detection, trajectory prediction, wind and weather conditions, energy consumption, communication delay, mobile network connectivity, and dynamic environment mapping.

[0101] The interface provision module (210) may provide an interface for modifying detailed settings of each node. For example, if a user wishes to create or update a "weather observation" model by specializing a node related to "wind and weather conditions," the user may manipulate detailed settings and weight settings for weather sensors, wind speed and direction, precipitation detection, visibility estimation, temperature consideration, weather forecast reception, and weather-based route planning in the "wind and weather conditions" node.

[0102] The interface provision module (210) can receive user input, train and test an artificial neural network model for added nodes, and provide a process of fine-tuning the model to the model training interface (403).

[0103] The interface provision module (210) can receive user input and provide a model deployment interface (404) for deployment of the generated target artificial neural network.

[0104]

[0105] FIG. 5 is a block diagram illustrating an electronic device according to one embodiment.

[0106] One or more blocks and combinations of blocks of FIG. 5 may be implemented by a special-purpose hardware-based computer performing a specific function, or by a combination of special-purpose hardware and computer instructions. The descriptions made with reference to FIGS. 1 to 4 may be equally applicable to FIG. 5 . For example, an electronic device (500) according to one embodiment may include a pipeline system (200).

[0107] As shown in FIG. 5, the electronic device (500) may include a memory (510) and a processor (520). The electronic device (500) may further include a communication module, and the communication module may include a transmitter and a receiver.

[0108] An electronic device (500) according to one embodiment may include a memory (510) and a processor (520) connected to the memory (510) via a system bus or other suitable circuitry.

[0109] The electronic device (500) may store program code in memory (510). In one embodiment, the memory (510) may include one or more physical memory devices, such as local memory or one or more bulk storage devices. In this case, the local memory may include random access memory (RAM) or other volatile memory devices commonly used while actually executing the program code. The bulk storage device may be implemented as a hard disk drive (HDD), a solid state drive (SSD), or other non-volatile memory device.

[0110] As the executable program code stored in the memory (510) is executed by the electronic device (500), the processor (520) may perform various operations described in the present disclosure. For example, the memory (510) may store program code for causing the processor (520) to perform one or more operations described in FIGS. 1 to 4.

[0111] Depending on the specific type of device being implemented, the electronic device (500) may include fewer components than those illustrated or additional components not illustrated in FIG. 5. Additionally, one or more of the components may be incorporated into, or otherwise form part of, another component.

[0112] A processor (520) according to one embodiment is a hardware configuration that performs overall control functions for controlling the operations of an electronic device (500). For example, the processor (520) may control the electronic device (500) overall by executing programs stored in a memory (510) within the electronic device (500). The processor (520) may be implemented as a central processing unit (CPU), a graphics processing unit (GPU), an application processor (AP), a neural processing unit (NPU), or the like, provided within the electronic device (500), but is not limited thereto.

[0113] The processor (520) receives a user's input related to a target artificial neural network model, detects a list of drones connected to a pipeline system (200) specialized for drones, receives movement paths and sensor data of drones (300) based on the drone list, analyzes the movement paths and sensor data of drones (300), learns the target artificial neural network model based on the user's input, movement paths and sensor data, selects valid target drones (300) from the drone list, and transmits the target artificial neural network model, for which learning has been completed, to the valid target drones (300).

[0114]

[0115] The embodiments described above may be implemented using hardware components, software components, and / or a combination of hardware components and software components. For example, the devices, methods, and components described in the embodiments may be implemented using a general-purpose computer or a special-purpose computer, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing instructions and responding to them. The processing device may execute an operating system (OS) and software applications running on the operating system. Furthermore, the processing device may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing device is sometimes described as being used alone; however, one of ordinary skill in the art will recognize that the processing device may include multiple processing elements and / or multiple types of processing elements. For example, a processing unit may include multiple processors, or a processor and a controller. Other processing configurations, such as parallel processors, are also possible.

[0116] Software may include a computer program, code, instructions, or a combination of one or more of these, which may configure a processing device to perform a desired operation or may, independently or collectively, command the processing device. The software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual equipment, computer storage medium or device, or transmitted signal wave, for interpretation by the processing device or for providing instructions or data to the processing device. The software may also be distributed over networked computer systems and stored or executed in a distributed manner. The software and data may be stored on a computer-readable recording medium.

[0117] The method according to the embodiment may be implemented in the form of program commands that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program commands, data files, data structures, etc., alone or in combination, and the program commands recorded on the medium may be those specially designed and configured for the embodiment or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute program commands such as ROMs, RAMs, and flash memories. Examples of program commands include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc.

[0118] The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

[0119] Although the embodiments described above have been described with limited drawings, those skilled in the art will appreciate that various technical modifications and variations can be applied based on the described embodiments. For example, appropriate results can still be achieved even if the described techniques are performed in a different order than described, and / or components of the described systems, structures, devices, circuits, etc. are combined or combined in a different manner than described, or are replaced or substituted with other components or equivalents.

[0120] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims described below.

Claims

1. In the operation method of a pipeline system specialized for drones, A step of receiving user input related to a target artificial neural network model; A step of detecting a list of drones connected to a pipeline system specialized for the above drones; A step of receiving movement paths and sensor data of drones based on the above drone list; A step of analyzing the movement path and sensor data of the above drones; A step of learning the target artificial neural network model based on the user's input, the movement path, and the sensor data; and A step of selecting valid target drones from the above drone list and transmitting the target artificial neural network model, for which learning has been completed, to the valid target drones. A method of operating a pipeline system specialized for drones, comprising:

2. In paragraph 1, The step of analyzing the above movement path and the above sensor data is A step of analyzing the movement path and the sensor data according to the input of the user, and providing the movement path and the sensor data of the drones in a visualized form. A method of operating a pipeline system specialized for drones, further comprising:

3. In paragraph 1, The step of analyzing the above movement path and the above sensor data is A step of analyzing the movement path and the sensor data according to the input of the user, thereby providing at least one of automation of mission assignment of the drones, load balancing, and service auto-scaling. A method of operating a pipeline system specialized for drones, further comprising:

4. In paragraph 1, The step of receiving the user's input is A step of providing an interface for receiving editing input for nodes within the pipeline system from the user. A method of operating a pipeline system specialized for drones, further comprising:

5. In paragraph 4, The steps for providing the above interface are A step of providing the user with an interface for checking tasks considering the mobility of the drone in the nodes within the pipeline; A step for providing an interface that can modify detailed settings of each of the above nodes; and A step of providing an interface for deploying the above target artificial neural network. A method of operating a pipeline system specialized for drones, comprising:

6. In paragraph 1, The steps for detecting the above drone list are: Step of checking the status of drones connected to the system from the above drone list and controlling drones that cannot fly among the above drones to stop their flight A method of operating a pipeline system specialized for drones, comprising:

7. In paragraph 1, The step of receiving the movement path and sensor data of the above drones is A step of labeling the sensor data using an automated labeling model included in the pipeline system based on the user's input. A method of operating a pipeline system specialized for drones, comprising:

8. A computer program stored on a computer-readable recording medium to execute the method of claim 1 by being combined with hardware.

9. In a pipeline system specialized for drones, An interface providing module for receiving user input related to a target artificial neural network model; A data classification module that detects a list of drones connected to a pipeline system specialized for the drones and receives movement paths and sensor data of the drones based on the list of drones; A data analysis module that analyzes the above movement path and the sensor data; and An artificial neural network learning module that learns the target artificial neural network model based on the user's input, the movement path, and the sensor data. Including, The above data analysis module Select valid target drones from the above drone list, The above interface providing module A pipeline system specialized for drones, which causes the artificial neural network learning module to transmit the target artificial neural network model, for which learning has been completed, to the valid target drones in response to the user's input.

10. In drones, Memory containing instructions; Sensor section; A communication unit that communicates with a server connected to a pipeline system specialized for the drone, transmits movement path and sensor data collected from the sensor unit, receives a target artificial neural network model, receives a user's task, and transmits the task result; and processor; Including, The above instructions, when executed by the processor, cause the drone to: Using the target artificial neural network model received from the communication unit, the artificial neural network model included in the drone is updated, Perform the tasks of the user received from the above communication department, A drone that retrains the artificial neural network model based on data collected from the sensor unit.

Images