KR-20260067666-A - SYSTEM FOR MONITORING THE PERFORMANCE OF A ROBOT'S MISSION AND METHOD USING THEREOF

KR20260067666AKR 20260067666 AKR20260067666 AKR 20260067666AKR-20260067666-A

Abstract

The present invention relates to a method for monitoring the performance of a robot's mission, and a method for monitoring the performance of a robot's mission to determine and control its state, comprising the steps of: measuring and collecting state data generated during mission performance through a state sensor of the robot; converting the collected state data into time-series data and storing it; analyzing the time-series data in an artificial intelligence-based analysis unit to extract feature information of the state data; and determining the state of the robot's mission performance based on the feature information and performing real-time control and feedback.

Inventors

정동경

Assignees

주식회사 현대케피코

Dates

Publication Date: 20260513
Application Date: 20241106

Claims (18)

As a method for monitoring the performance of a robot's mission to determine its state and control it, (a) A step of measuring and collecting state data generated during mission execution through the robot's state sensor; (b) a step of converting the collected state data into time series data and storing it; (c) a step of analyzing the above time series data in an artificial intelligence-based analysis unit to extract feature information of the above state data; and (d) A step of determining the mission execution status of the robot based on the above feature information and performing real-time control and feedback A method for monitoring the mission performance of a robot including
In paragraph 1, The state sensor used in step (a) above includes one or more of an encoder, force and torque sensors and position and attitude sensors attached to the robot. Method for monitoring the mission performance of a human robot.
In paragraph 1, In the above step (b), the state data is converted into time-series data by the data processing and storage unit, and the converted data is stored in a matrix structure. Method for monitoring the mission performance of a human robot.
In paragraph 1, In step (c) above, the AI-based analysis unit analyzes the time series data using a convolutional neural network (CNN) with a multi-convolutional layer structure and uses a kernel structure of multiple rows, 1 column, 3 columns, and 2 columns. Method for monitoring the mission performance of a human robot.
In paragraph 1, The above step (d) involves the robot control unit determining the mission execution status of the robot based on the feature information, and performing real-time control and feedback if a disturbance or abnormal state is detected. Method for monitoring the mission performance of a human robot.
In paragraph 5, The above robot control unit enters backup mode to resolve the problem situation in the event that the robot fails to perform its mission. Method for monitoring the mission performance of a human robot.
In paragraph 1, A step in which a learning control unit uses mission execution data to train and update a model of an AI-based analysis unit, provides the updated model to the AI-based analysis unit, the AI-based analysis unit transmits the updated model to a robot control unit, and the robot control unit determines the real-time mission execution status and reflects it in controlling the robot. A method for monitoring a robot's mission performance that further includes
In Paragraph 7, The above learning control unit and robot control unit form a dual-loop structure, wherein the dual-loop structure comprises an internal control loop configured by the robot control unit, which controls the robot based on state data generated during real-time mission execution and determines the mission execution status, and an external learning and monitoring loop configured by the learning control unit, which learns and analyzes mission execution data collected from the internal control loop to update the model of the artificial intelligence-based analysis unit and performs learning and state evaluation regarding mission execution. Method for monitoring the mission performance of a human robot.
In paragraph 8, The above external learning and monitoring loop provides the model of the AI-based analysis unit, updated by the learning control unit, to the robot control unit, and the robot control unit determines the real-time mission execution status based on this and controls the robot. Method for monitoring the mission performance of a human robot.
A state sensor for measuring and collecting state data; A data processing and storage unit for converting state data collected from the above state sensor into time-series data and storing it; An artificial intelligence-based analysis unit for analyzing the above time series data to extract feature information of the above state data; and A robot control unit for determining the mission execution status of the robot based on the above feature information and performing real-time control and feedback A robot mission performance monitoring system including
In Paragraph 10, The above state sensor includes one or more of an encoder, force and torque sensors, and position and attitude sensors attached to the robot. Monitoring system for the mission performance of a human robot.
In Paragraph 10, The above data processing and storage unit converts the collected state data into time-series data, and the converted data is stored in a matrix structure. Monitoring system for the mission performance of a human robot.
In Paragraph 10, The above artificial intelligence-based analysis unit analyzes the above time series data using a convolutional neural network (CNN) with a multiple convolutional layer structure and uses a kernel structure of multiple rows, 1 column, 3 columns, and 2 columns. Monitoring system for the mission performance of a human robot.
In Paragraph 10, The robot control unit above determines the mission execution status of the robot based on the feature information above, and performs real-time control and feedback when it detects disturbances or abnormal conditions. Monitoring system for the mission performance of a human robot.
In Paragraph 14, The above robot control unit enters backup mode to resolve the problem situation in the event that the robot fails to perform its mission. Monitoring system for the mission performance of a human robot.
In Paragraph 10, It further includes a learning control unit that uses mission execution data to train and update a model of an AI-based analysis unit and provides the updated model to the AI-based analysis unit, wherein the AI-based analysis unit transmits the updated model to a robot control unit, and the robot control unit determines the real-time mission execution status based thereon and reflects it in controlling the robot. Monitoring system for the mission performance of a human robot.
In Paragraph 16, The above learning control unit and robot control unit form a dual-loop structure, wherein the dual-loop structure comprises an internal control loop configured by the robot control unit, which controls the robot based on state data generated during real-time mission execution and determines the mission execution status, and an external learning and monitoring loop configured by the learning control unit, which learns and analyzes mission execution data collected from the internal control loop to update the model of the artificial intelligence-based analysis unit and performs learning and state evaluation regarding mission execution. Monitoring system for the mission performance of a human robot.
In Paragraph 17, The above external learning and monitoring loop provides the model of the AI-based analysis unit, updated by the learning control unit, to the robot control unit, and the robot control unit determines the real-time mission execution status based on this and controls the robot. Monitoring system for the mission performance of a human robot.

Description

System for monitoring the performance of a robot's mission and method thereof The present invention relates to a robot mission performance monitoring system and a method thereof. Robots are used in various industrial processes for process automation and service provision. When performing tasks, these robots generate diverse movements through multiple joints and combine them to carry out process operations or services. Generally, a robot (or robot arm) performs a task by following a specific target position and posture. During this process, the robot's end-effector operates by making contact with the work object or changing its position, and correctly monitoring and controlling these movements is a key factor in the success of the task. Conventional robot arm control methods focus primarily on accurately following the trajectory of the end effector, and lack specific monitoring methods to determine whether the mission has been successfully performed. In addition, while a robot arm can perform accurate movements along a target trajectory in an environment free from disturbances, in real-world environments, robot arms are often affected by external forces or environmental factors, and such disturbances are a significant factor hindering mission performance. Furthermore, existing control and monitoring methods are insufficient for accurately determining mission results by simultaneously considering the complex interactions between factors such as position, attitude, force, and torque during the robot arm's task execution. In particular, when the robot arm comes into contact with a work object (e.g., peg-in-hole operations), there are limitations in accurately identifying changes in force and attitude caused by contact and reflecting them in the task execution. In addition, existing technologies mostly analyze the position or force at a specific moment rather than analyzing data on the movement of a robot arm on a time-series basis. However, since the motion of a robot arm is a continuous movement over time expressed as time-series data, there is a need to analyze it comprehensively. FIG. 1 illustrates a method for monitoring the mission performance of a robot according to an embodiment of the present invention. FIG. 2 illustrates a schematic diagram of a robot mission performance monitoring system according to an embodiment of the present invention. Figure 3 shows a graph of changes over time in state data collected from a robot according to an embodiment of the present invention. Figure 4 is a schematic diagram in which state data measured according to an embodiment of the present invention is arranged by time and feature and expressed in the form of a matrix. FIG. 5 is a block diagram showing a computer system for implementing a method according to an embodiment of the present invention. The aforementioned objectives of the present invention, as well as other objectives, advantages, and features, and the methods for achieving them, will become clear from the embodiments described in detail below together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but can be implemented in various different forms, and the following embodiments are provided merely to easily inform those skilled in the art of the purpose, structure, and effects of the invention, and the scope of the rights of the present invention is defined by the description in the claims. Meanwhile, the terms used in this specification are for describing the embodiments and are not intended to limit the invention. In this specification, the singular form includes the plural form unless specifically stated otherwise in the text. As used in this specification, "comprises" and/or "comprising" do not exclude the presence or addition of one or more other components, steps, actions, and/or elements to the mentioned components, steps, actions, and/or elements. According to an embodiment of the present invention, the method and system for monitoring the mission performance of a robot (10), analyzing state data collected therefrom to determine and control the mission performance status of the robot (10) are structured, and real-time feedback and control are performed through artificial intelligence-based analysis. FIG. 1 illustrates a method for monitoring the mission performance of a robot according to an embodiment of the present invention. As described above, the method for monitoring a robot's mission performance according to an embodiment of the present invention includes the steps of: measuring and collecting state data generated during mission performance through a state sensor of the robot (S710); converting the collected state data into time-series data and storing it (S720); analyzing the time-series data in an artificial intelligence-based analysis unit (300) to extract feature information of the state data (S730); and determining the mission performance status of the robot based on the f