KR-20260065751-A - Autonomous industrial plant operating system and method based on layered digital twin and edge agentic AI

Abstract

The present invention relates to an autonomous industrial plant system and method based on a hierarchical digital twin and edge agentic AI, comprising an industrial facility including a physical actuator and a sensor unit; a local digital twin positioned adjacent to the industrial facility to provide a virtual simulation environment reflecting physical laws; an edge agentic AI that utilizes the local digital twin as a virtual sandbox to pre-verify control scenarios and directly applies verified optimal control values to the actuator; and a main digital twin that performs global operational management through a satellite communication network. The present invention realizes a combination of effects, including ultra-low latency autonomous control even in a high-latency satellite communication environment, runtime pre-verification of AI control commands, delta-based satellite communication data reduction, and uninterrupted operation through an autonomous survival mode in the event of communication disconnection.

Inventors

• 안범주

Assignees

• 안범주

Dates

Publication Date

20260511

Application Date

20260324

Claims (1)

1. In autonomous industrial plant systems, Industrial equipment deployed on sea or land, comprising a physical driving unit and a sensor unit for measuring state; A local digital twin positioned adjacent to the above industrial facility and providing a virtual simulation environment reflecting physical laws based on data from the sensor unit; Edge Agentic AI deployed at the site of the above industrial facility to pre-verify control scenarios using the local digital twin as a virtual simulator and directly apply the verified optimal control value to the actuator (Direct Actuation); and Main Digital Twin located at the ground control center, which performs global operational management by synchronizing data with the aforementioned Edge Agentic AI via a satellite communication network An autonomous operating system including

Description

Autonomous industrial plant operating system and method based on layered digital twin and edge agentic AI Autonomous industrial plant operating system and method based on layered digital twin and edge agentic AI The present invention relates to remote control and autonomous operation technology for industrial facilities, and more specifically, to a hierarchical digital twin system that ensures the safety of facilities and performs optimal control autonomously through edge intelligence even in low-bandwidth, high-latency communication environments such as satellite communication environments, and an autonomous operation method using the same. Industrial facilities (100) located in remote areas, such as offshore gas production platforms or chemical process plants in remote areas, often rely on satellite communication networks (500) due to their geographical characteristics. In the case of geostationary (GEO) satellites, the one-way latency alone is about 240 to 280 ms, and the data transmission cost is tens of times higher than that of ground optical communication networks. Such high latency and limited bandwidth structurally make real-time direct control from a ground control center (600) impossible. Conventional remote control systems adopted a structure in which control commands were transmitted from the control center in an open-loop manner and the drive unit (110) at the site manually executed them. However, this method lacked a mechanism to safely stop or maintain the drive unit (110) at the site in the event of a satellite communication outage, and thus contained a serious safety vulnerability in which a communication outage immediately led to a state of loss of control. In addition, existing digital twin technology is limited to the so-called 'digital shadow' level, which is mainly limited to offline simulation during the design verification stage or the generation of alarms after detecting anomalies during operation. Consequently, there is no verification mechanism for the artificial intelligence (AI) to directly issue commands to the plant drive unit (110) where complex physical laws are applied. In other words, even though judgment errors in the AI model can lead directly to actual equipment damage or safety accidents, a pre-simulation structure that verifies the corresponding scenario in real-time at runtime for every control command has not been disclosed in the prior art. In addition, conventional technology adopted a method of transmitting all raw sensor data from the field directly to the control center via satellite communication, which resulted in excessive bandwidth usage and unnecessary waste of satellite communication costs. The present invention aims to solve the problems of the prior art as described above and proposes a hierarchical digital twin structure based on edge intelligence. FIG. 1 is a block diagram showing the overall system architecture of the present invention, illustrating the three-layer connection structure of edge-satellite-ground and the relationships of key components (100, 200, 300, 400, 500, 600, 700). FIG. 2 is a flowchart showing an internal logic process in which an edge agentic AI (300) utilizes a local digital twin (200) as a virtual simulator, showing the operation flow of a state recognition module (310), a control scenario generation unit (320), a pre-simulation module (330), a safety verification unit (340), and a direct authorization unit (350). FIG. 3 is a timing diagram showing the sequence of control transfer by the autonomous survival mode management unit (370) when communication of the satellite communication network (500) is interrupted, showing the process of monitoring the communication status, determining whether a threshold is exceeded, activating the autonomous survival mode, and restoring it. FIG. 4 is a data flow diagram showing the process in which a data optimization module (360) extracts only the error (Delta) between the raw data of the sensor unit (120) and the predicted value of the local digital twin (200) and transmits it to the main digital twin (400) through a satellite communication network (500). Preferred embodiments of the present invention will be described in detail below with reference to the attached drawings. Prior to this, terms and words used in this specification and claims should not be interpreted as being limited to their ordinary or dictionary meanings. Instead, based on the principle that the inventor can appropriately define the concepts of terms to best describe their invention, they should be interpreted in a meaning and concept consistent with the technical spirit of the present invention. Therefore, the embodiments described in this specification and the configurations illustrated in the drawings are merely some of the most preferred embodiments of the present invention and do not represent all of the technical spirit of the present invention. It should be understood that various equivalents and modifications c