System and method for automatic interpretation and dissemination of doctrine-based operation order

Abstract

The present invention relates to a doctrine-based automatic interpretation and dissemination system for operational orders. More specifically, in a command and control system according to the present disclosure that receives operational orders from a superior unit and automatically interprets and transmits them as operational orders from the perspective of a subordinate unit, the system comprises: a communication unit that receives an operational order serialized in a first C-BML (Coalition Battle Management Language) format from a superior unit; a command interpretation unit that deserializes the operational order in the first C-BML format to extract the superior unit's operational control line, friendly and enemy unit deployment information, and the superior unit's mission; a terrain and threat analysis unit that extracts a unique Area of ​​Operations (AO) of the subordinate unit based on the extracted operational control line and pre-established offline terrain data, and analyzes enemy infiltration approach routes and firepower threats within the Area of ​​Operations; and a task derivation and mission generation unit that recalculates the threat analysis results and the superior unit's mission from the perspective of the subordinate unit to infer explicit tasks, derives implicit task candidates using an artificial intelligence (AI) model, and generates mission sentences based on core tasks confirmed by user approval. It may include an ontology verification unit that automatically verifies whether the generated mission statement and the operational plan of the subordinate unit violate the doctrine based on an ontology knowledge graph reflecting military doctrine; and a document serialization unit that controls the allocation of the verified operational plan of the subordinate unit to each subordinate unit of the subordinate unit, serializes it into a second C-BML format, and transmits it through the communication unit. Other embodiments are also possible.

Description

Technology Field

[0001] The present disclosure relates to a system and method for automatic interpretation and dissemination of operational orders. More specifically, the present disclosure relates to a doctrine-based system and method for automatic interpretation and dissemination of operational orders. Background Technology

[0002] In modern military operations, comprehensive operational orders (OPORDs) from higher-level units must be reinterpreted in accordance with the unique areas of responsibility (AOs) and military doctrines of subordinate units; however, the conventional manual method relies on the subjectivity of staff officers, which can frequently lead to inconsistencies in interpretation and omissions of tasks.

[0003] Furthermore, there is a limitation in that operational delivery is delayed because the process of deriving implicit tasks from vast amounts of information and drafting mission statements based on the five Ws (who, what, where, when, why, and how), as well as the process of dividing confirmed plans into subordinate unit units and converting them into standard exchange formats such as C-BML, are not automated. Moreover, in a battlefield environment where communication networks are frequently interrupted, command and control functions may be paralyzed, and there is a risk that hallucinations violating doctrine may occur if artificial intelligence (AI) is simply introduced for rapid decision-making.

[0004] Therefore, there is an increasing need to develop a highly reliable command and control system that structures the automatic reinterpretation of operational orders and task analysis through collaboration between artificial intelligence and commanders, blocks logical errors through ontology-based doctrine verification, and supports chain propagation based on standard formats even in the event of communication breakdown. Prior art literature

[0005] Patent Document: Korean Published Patent Application 10-2565906 The problem to be solved

[0006] The embodiments disclosed in this disclosure aim to provide a system and method that can resolve the problems of inter-unit interpretation discrepancies or omission of essential tasks that occurred due to reliance on manual work by staff, and support the commander's rapid and accurate situational awareness by automatically recalculating and interpreting comprehensive operational orders received from a superior unit into axes and enemy threats from the perspective of the subordinate unit based on the subordinate unit's unique area of ​​responsibility (AO) and offline terrain data.

[0007] In addition, the embodiments disclosed in this disclosure aim to provide a system and method that can improve the consistency of operational planning and drastically reduce the time required for document preparation by structuring the three types of task analysis procedures—explicit tasks, implicit tasks, and core tasks—according to doctrine into data, confirming core tasks by combining the commander's approval (Human-in-the-Loop) with implicit task candidates derived through an artificial intelligence (AI) model, and automatically synthesizing a Mission Statement and Information Requirements (CCIR / PIR) based on the 5W principles (Who, What, When, Where, Why, How) based on this.

[0008] In addition, the embodiments disclosed in this disclosure aim to provide a system and method capable of deriving tactical measures that are 100% reliable in the actual battlefield by utilizing an ontology knowledge graph reflecting military doctrine, qualification query (CQ), and SHACL rules to strictly and automatically verify at the data level whether operational plans proposed by artificial intelligence or written by a user (such as position placement, concealment, and visual observation conditions) violate doctrine, thereby fundamentally blocking the hallucination phenomenon of generative artificial intelligence.

[0009] In addition, the embodiments disclosed in this disclosure aim to provide a system and method capable of ensuring perfect interoperability between heterogeneous systems and achieving delay-free transmission of operational orders from the highest level to the lowest level by automatically assigning verified operational plans of subordinate units to each subordinate unit, serializing and digitally signing them in C-BML, a global standard exchange format, utilizing an offline queue in the event of communication interruption due to the battlefield environment to restore and transmit without interruption according to urgency priority, and allowing the receiving unit to recursively repeat the procedure through the same interface. means of solving the problem

[0010] A command and control system according to the present disclosure for achieving the aforementioned technical objectives, which receives an operational order from a superior unit and automatically interprets and transmits it as an operational order from the perspective of a subordinate unit, comprising: a communication unit that receives an operational order serialized in a first C-BML (Coalition Battle Management Language) format from a superior unit; a command interpretation unit that deserializes the operational order in the first C-BML format to extract the operational control line of the superior unit, information on the deployment of friendly and enemy units, and the mission of the superior unit; a terrain and threat analysis unit that extracts a unique Area of ​​Operations (AO) of the subordinate unit based on the extracted operational control line and pre-established offline terrain data, and analyzes enemy infiltration approach routes and firepower threats within the Area of ​​Operations; and a task derivation and mission generation unit that recalculates the threat analysis results and the superior unit mission from the perspective of the subordinate unit to infer explicit tasks, derives implicit task candidates using an artificial intelligence (AI) model, and generates mission sentences based on core tasks confirmed by user approval. It may include an ontology verification unit that automatically verifies whether the generated mission statement and the operational plan of the subordinate unit violate the doctrine based on an ontology knowledge graph reflecting military doctrine; and a document serialization unit that controls the verification of the operational plan of the subordinate unit to be assigned to each subordinate unit of the subordinate unit, serialized into a second C-BML format, and transmitted through the communication unit.

[0011] In addition, the terrain and threat analysis unit can automatically synthesize a polygon of the responsibility area dedicated to the subordinate unit by cutting out the operational control line information of the superior unit to match the shape of the responsibility area of ​​the subordinate unit through a polygon clipping technique.

[0012] In addition, the terrain and threat analysis unit may calculate a range threat area based on the enemy artillery ammunition type (Ammunition Policy) extracted from the first C-BML format by referring to a pre-established weapon catalog, automatically filter and exclude enemy units that do not reach the responsibility area of ​​the subordinate unit from the threat targets, and calculate and provide a weighted ratio of friendly and enemy combat power based on the deployment information of friendly and enemy units within the responsibility area.

[0013] In addition, the terrain and threat analysis unit can derive approach routes and entry points by performing majority voting based on grid data and a hybrid search technique based on road network data to extract the enemy infiltration approach route that can enter the polygon of the responsibility area.

[0014] In addition, the task derivation and mission generation unit can recalculate the main and auxiliary attack axes based on the superior unit standards specified in the first C-BML format into axes based on the derived enemy infiltration approach route and the responsibility area polygon, and use the established ontology and unit organization knowledge graph to automatically deploy enemy unit information included in the superior unit operation order to the level of lower enemy units that the subordinate unit must directly respond to, thereby concretizing it as an analysis target.

[0015] In addition, the task derivation and mission generation unit can combine the recalculated axis and specific enemy unit information with the specified tasks of the superior unit and the subordinate unit's area of ​​responsibility to automatically infer the independent specified tasks of the subordinate unit that are not specified in the superior unit's operational order.

[0016] Additionally, it further includes a multimodal input unit that receives only the user's speech-to-text (STT) and drawing gestures without keyboard typing and converts them into normalized events; and the task derivation and mission generation unit can classify the intent of the voice text received through the multimodal input unit and automatically route an inference request to either an on-device model in an offline environment or a high-performance model in a server network.

[0017] In addition, the task derivation and mission generation unit inputs the inferred explicit tasks, military doctrines, and missions into the artificial intelligence model to generate the implicit task candidates additionally required for the execution of operations, presents them to a user screen linked with the multimodal input unit, and can finally select the core tasks from among the specified tasks and the confirmed items through the user's confirmation or rejection input regarding the implicit task candidate group.

[0018] In addition, the task derivation and mission generation unit can keep the mission sentence blank when the core task is not selected, and automatically fill and create the mission sentence by combining the core task and mission information in a 5W1H template in response to a trigger in which the core task is selected.

[0019] In addition, the task derivation and mission generation unit can automatically infer and generate commander information requirements and priority information requirements that the subordinate unit must verify through reconnaissance assets based on the mission sentence and threat analysis results written above.

[0020] In addition, the above-mentioned ontology verification unit is equipped with an ontology schema and SHACL query rules extended based on the MIM (Multilateral Interoperability Programme Information Model) standard, and can block illusion errors of the artificial intelligence model at the data level by comparing the automatically generated mission statement, the information requirements, and the position deployment data with the SHACL rules.

[0021] In addition, the ontology verification unit may model the tactical requirements specified in the military doctrine in the form of a qualifying query (CQ) and include them in the ontology schema, and automatically verify through the query and shackle rules whether the placement of friendly positions on the position placement data satisfies the daytime visual observation conditions on the offline terrain data or the concealment conditions from enemy threats, and if the conditions are not met, it may generate an alarm to the user and reject the mission statement and position placement data.

[0022] In addition, the document serialization unit can divide the verified operational plan into units of subordinate units based on the subordinate unit standards, individually assign the responsibility area and assigned firepower to each subordinate unit, and manage the coordinates of targets and control means from a single source to automatically synchronize without data conflicts.

[0023] In addition, when the above document serialization unit serializes the above-configured sub-unit operational orders into the above-mentioned second C-BML format, it may selectively support conversion and expansion into heterogeneous standard exchange formats including USMTF (United States Message Text Format) by utilizing a built-in format adapter structure.

[0024] In addition, the above system is linked with the defense network PKI (Public Key Infrastructure), and the document serialization unit can add an encrypted digital signature of the commander of the subordinate unit to the serialized second C-BML format data payload.

[0025] In addition, the communication unit may load and wait for the second C-BML format operation command with the electronic signature attached in an offline queue when the server communication network is disconnected due to the battlefield environment, and automatically transmit the data loaded in the offline queue according to a preset urgency priority in the order of fragment command, operation command, and artificial intelligence inference request as soon as a network connection restoration signal is detected.

[0026] In addition, the system of the subordinate unit that receives the operation command of the second C-BML format through the communication unit may be configured to treat the received operation command of the second C-BML format as an operation command of the superior unit and recursively repeat the operation of the command interpretation unit to the communication unit so that terrain-based interpretation, task creation, and division and transmission procedures are sequentially propagated from the highest unit to the lowest unit through the same interface.

[0027] A method performed by a command and control system according to the present disclosure for achieving the aforementioned technical task, which receives an operational order from a superior unit and automatically interprets and transmits it as an operational order from the perspective of a subordinate unit, comprising: a communication unit receiving an operational order serialized in a first C-BML (Coalition Battle Management Language) format from a superior unit; a command interpretation unit deserializing the operational order in the first C-BML format to extract the operational control line of the superior unit, information on the deployment of friendly and enemy units, and the mission of the superior unit; a terrain and threat analysis unit extracting the unique Area of ​​Operations (AO) of the subordinate unit based on the extracted operational control line and pre-established offline terrain data, and analyzing enemy infiltration approach routes and firepower threats within the Area of ​​Operations; and a task derivation and mission generation unit recalculating the threat analysis results and the superior unit mission from the perspective of the subordinate unit to infer explicit tasks, deriving implicit task candidates using an artificial intelligence (AI) model, and then generating mission sentences based on core tasks confirmed by user approval. The method may include: a step in which an ontology verification unit automatically verifies whether the generated mission statement and the operational plan of the subordinate unit violate the doctrine based on an ontology knowledge graph reflecting military doctrine; and a step in which a document serialization unit controls the verification of the operational plan of the subordinate unit to be assigned to each subordinate unit of the subordinate unit, serializes it into a second C-BML format, and transmits it through the communication unit.

[0028] In addition, the terrain and threat analysis unit may include a step of automatically synthesizing a polygon of responsibility area dedicated to the subordinate unit by cutting out the operational control line information of the superior unit to match the shape of the subordinate unit's responsibility area through a polygon clipping technique.

[0029] In addition, the terrain and threat analysis unit may include the step of calculating a range threat area based on the enemy artillery ammunition type (Ammunition Policy) extracted from the first C-BML format by referring to a pre-established weapon catalog, automatically filtering out enemy units that do not reach the responsibility area of ​​the subordinate unit from the threat targets, and calculating and providing a weighted ratio of friendly and enemy combat power based on information on the deployment of friendly and enemy units within the responsibility area.

[0030] In addition to this, a computer-readable recording medium for recording a computer program for implementing the present disclosure may be further provided. Effects of the invention

[0031] According to the aforementioned means for solving the problem of the present disclosure, a comprehensive operational order received from a superior unit is automatically recalculated based on the area of ​​responsibility (AO) of a subordinate unit and pre-established offline terrain data, and by precisely analyzing enemy infiltration approach routes and firepower threats using a hybrid search technique based on a grid network and a road network, the time required for terrain analysis and order interpretation, which previously relied entirely on manual work by staff officers, can be innovatively reduced, and the effect of preventing discrepancies in interpretation between units or the omission of critical threat analysis can be provided.

[0032] According to the aforementioned means for solving the problem of the present disclosure, the three types of task analysis procedures—explicit tasks, implicit tasks, and core tasks prescribed in military doctrine—are systematically structured, and by combining the commander's final approval (Human-in-the-Loop) with implicit task candidates proposed by artificial intelligence (AI) and automatically synthesizing mission sentences and information requirements (CCIR / PIR) based on the confirmed core tasks based on the 5W template, the document writing fatigue of commanders and staff can be significantly reduced, and the consistency and speed of establishing operational plans can be maximized.

[0033] According to the aforementioned means for solving the problem of the present disclosure, by providing an ontology knowledge graph that precisely reflects military doctrine and qualification query (CQ) and SHACL verification rules, and by strictly and automatically verifying at the data level whether mission statements and position deployment data proposed by artificial intelligence or drawn by a user satisfy tactical requirements such as visual observation or concealment, it is possible to fundamentally block fatal hallucination phenomena or tactical contradictions that may occur when introducing generative artificial intelligence, thereby providing the effect of establishing an operational plan that is 100% reliable in the actual battlefield.

[0034] According to the aforementioned means for solving the problem of the present disclosure, a multimodal input method that recognizes only the user's speech-to-text (STT) and drawing gestures without keyboard typing is supported, and by analyzing the user's voice intent and automatically routing inference to an on-device (Tier 2) AI model in an offline environment where communication is disconnected, and to a high-performance server (Tier 3) model when connected to a network, it is possible to maximize usability in urgent battlefield situations and provide the effect of ensuring uninterrupted operational command even in communication blind spots.

[0035] According to the aforementioned means for solving the problem of the present disclosure, by automatically assigning responsibility areas and assigned firepower to each subordinate unit unit based on verified operational plans of subordinate units, synchronizing coordinates of targets, etc., with a single source (SSOT), and then serializing and digitally signing them in C-BML, a global standard exchange format, and disseminating them, it is possible not only to ensure perfect interoperability with other systems and allied forces, but also to provide the effect of achieving sequential propagation of operational orders from the highest level to the lowest level without delay or interruption through automatic transmission according to the priority of the offline queue upon communication restoration and a recursive iterative execution mechanism of subordinate units.

[0036] The effects of the present disclosure are not limited to those mentioned above, and other unmentioned effects will be clearly understood by a person skilled in the art from the description below. Brief explanation of the drawing

[0037] FIG. 1 is a block diagram showing a simplified structure of a doctrine-based operational command automatic interpretation and dissemination system according to an example of the present invention. FIG. 2 is a flowchart illustrating a doctrine-based operational order automatic interpretation and propagation method according to an example of the present invention. FIG. 3 is a flowchart illustrating a terrain-based responsibility zone synthesis and enemy threat analysis method according to an example of the present invention. FIG. 4 is a flowchart illustrating an AI-based task derivation and multimodal mission sentence generation method according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating an ontology and doctrine-based operational plan verification method according to an embodiment of the present invention. FIG. 6 is a flowchart illustrating a subordinate command partitioning assignment and serialization method according to an embodiment of the present invention. FIG. 7 is a flowchart illustrating an offline queue management and chain propagation method according to an embodiment of the present invention. Specific details for implementing the invention

[0038] Throughout this disclosure, the same reference numerals denote the same components. This disclosure does not describe all elements of the embodiments, and general content in the art to which this disclosure pertains or content that overlaps between embodiments is omitted. The terms part, module, component, and block as used in the specification may be implemented in software or hardware, and depending on the embodiments, a plurality of parts, modules, components, and blocks may be implemented as a single component, or a single part, module, component, or block may include a plurality of components.

[0039] Throughout the specification, when a part is described as being "connected" to another part, this includes not only cases where they are directly connected but also cases where they are indirectly connected, and indirect connections include connections made via a wireless communication network.

[0040] Furthermore, when it is stated that a part "includes" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0041] Throughout the specification, when it is stated that a component is located "on" another component, this includes not only cases where a component is in contact with another component, but also cases where another component exists between the two components.

[0042] The terms first, second, etc. are used to distinguish one component from another, and the components are not limited by the aforementioned terms.

[0043] Singular expressions include plural expressions unless there is an obvious exception in the context.

[0044] In each step, identification codes are used for convenience of explanation and do not describe the order of the steps; the steps may be performed differently from the specified order unless a specific order is clearly indicated in the context.

[0045] The operating principles and embodiments of the present disclosure will be described below with reference to the attached drawings.

[0046] In this specification, the device according to the present disclosure includes all various devices capable of performing computational processing and providing results to a user. For example, the device according to the present disclosure may include all of a computer, a server device, and a portable terminal, or may be in the form of any one of them.

[0047] Here, the computer may include, for example, a notebook, desktop, laptop, tablet PC, slate PC, etc. equipped with a web browser.

[0048] The above server device is a server that processes information by communicating with an external device, and may include an application server, a computing server, a database server, a file server, a game server, a mail server, a proxy server, and a web server.

[0049] The above portable terminal may include, for example, all types of handheld-based wireless communication devices such as PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), WiBro (Wireless Broadband Internet) terminals, smartphones, etc., as well as wearable devices such as watches, rings, bracelets, anklets, necklaces, glasses, contact lenses, or head-mounted devices (HMDs).

[0050] FIG. 1 is a block diagram showing a simplified structure of a doctrine-based operational command automatic interpretation and dissemination system according to an example of the present invention.

[0051] Referring to FIG. 1, a doctrine-based operational order automatic interpretation and propagation system (10) according to one embodiment of the present invention can automatically reinterpret an operational order issued from a superior unit in terms of the responsibility area and organization of a subordinate unit, and after undergoing task analysis based on military doctrine and the generation of mission sentences based on artificial intelligence, logically verify it and propagate it sequentially to the lowest subordinate unit without delay. For example, the doctrine-based operational order automatic interpretation and propagation system (10) may include a communication unit (100), a command interpretation unit (200), a terrain and threat analysis unit (300), a task derivation and mission generation unit (400), an ontology verification unit (500), a document serialization unit (600), and a multimodal input unit (700).

[0052] In one embodiment, the communication unit (100) can perform the role of a bidirectional interface that receives a first operational command payload serialized in the C-BML (Coalition Battle Management Language) standard format from a higher-level unit system (e.g., AI-CDSS) and transmits a finally confirmed second operational command to a lower-level unit. The communication unit (100) maintains the security and integrity of data transmission and reception by applying a military standard encryption system (e.g., ARIA-256), and can monitor the offline state of a disconnected server communication network in real time, taking into account the specific characteristics of the battlefield environment. When a communication disconnection is detected, the communication unit (100) can load digitally signed operational commands or artificial intelligence inference request data awaiting transmission into an internal offline queue and keep them waiting. Subsequently, when a network connection recovery signal is detected, the communication unit (100) automatically transmits data loaded in the offline queue to the corresponding subordinate unit system according to the pre-set urgency priority in the order of fragment command (FRAGO), operation command (OPORD), and artificial intelligence inference request, thereby ensuring uninterrupted command and control even in communication blind spots.

[0053] In one embodiment, the command interpretation unit (200) deserializes and parses the first C-BML format operational command received through the communication unit (100), and can precisely extract operational control lines set by the superior unit, information on the deployment of friendly and enemy units, restrictions, and superior unit mission data. The command interpretation unit (200) can convert the extracted information into an internal system data model and transmit it so that other components, such as the terrain and threat analysis unit (300) and the task derivation and mission creation unit (400), can utilize it immediately. In addition, if parsing fails due to data structure corruption or compatibility issues, the original text is displayed on the screen as a fallback to assist the commander in continuously being aware of the situation.

[0054] In one embodiment, the terrain and threat analysis unit (300) can automatically synthesize a unique area of ​​responsibility (AO) polygon unique to a subordinate unit by precisely cutting it out using a polygon clipping technique based on the operational control line coordinate information extracted from the command interpretation unit (200) and a pre-established offline terrain data bundle. The terrain and threat analysis unit (300) can individually calculate the range threat area according to the enemy artillery ammunition type (Ammunition Policy) extracted by referring to an internal weapon catalog, and can automatically filter and exclude enemy units whose calculated area does not reach the subordinate unit's area of ​​responsibility polygon from the threat targets to reduce the commander's cognitive load. Furthermore, the terrain and threat analysis unit (300) can derive an entry point by performing a hybrid search technique that combines tactical majority voting based on offline grid network data and path search based on road network data to extract an enemy infiltration approach route that can enter the area of ​​responsibility, and can calculate and provide a Combat Power Ratio with weights based on the organization and firepower information of friendly and enemy units.

[0055] In one embodiment, the task derivation and mission generation unit (400) can recalculate the main and auxiliary attack axes in the superior unit's command into axes from the perspective of the subordinate unit (e.g., main attack, auxiliary attack, fixation) based on the polygons of enemy infiltration approach routes and responsibility areas derived from the terrain and threat analysis unit (300). The task derivation and mission generation unit (400) can utilize the established ontology and unit organization knowledge graph to expand and concretize enemy unit information roughly included in the superior unit's command to the level of lower enemy units, and based on this, can automatically infer the independent explicit tasks of the subordinate unit. Subsequently, the task derivation and mission generation unit (400) inputs the inferred explicit tasks and military doctrine into an artificial intelligence model to derive a group of implicit task candidates required for the execution of operations, and can finalize the core tasks through user confirmation or rejection input (Human-in-the-Loop). In response to a trigger that confirms a core task, the task derivation and mission generation unit (400) automatically fills in and creates a mission sentence by combining the core task and mission information in a 5W template, and can automatically generate and provide Commander's Required Information (CCIR) and Priority Required Information (PIR) that must be verified through reconnaissance assets based on the created mission sentence and threat analysis results.

[0056] In one embodiment, the ontology verification unit (500) can strictly and automatically verify whether the previously generated mission statement and operation plan violate the doctrine at the data level by utilizing an ontology schema and SHACL query rules based on the MIM (Multilateral Interoperability Programme Information Model) standard that reflects military doctrine. In particular, the ontology verification unit (500) models the tactical requirements specified in the military doctrine in the form of a Competency Question (CQ) to precisely compare whether the friendly position deployed by the user satisfies the Line of Sight (LOS) conditions on the terrain data or the conditions for concealment from enemy firepower threats. If the verification unit (500) identifies a contradiction caused by a hallucination error of the generative artificial intelligence or fails to meet the tactical conditions as a result of the comparison, it can generate a warning alarm on the screen and immediately reject the relevant measures and mission statement, thereby inducing the user to make corrections and ensuring the integrity of the operation plan.

[0057] In one embodiment, the document serialization unit (600) can divide the operational plan of a subordinate unit, which has been finally confirmed through the ontology verification unit (500), into subordinate unit units (e.g., company) to form operational orders for individual subordinate units. During the division and allocation process, the document serialization unit (600) can individually assign the responsibility area and assigned firepower of the subordinate unit, and perform automatic synchronization by managing them in an integrated manner under the Single Source of Truth (SSOT) principle so that the coordinate data of targets and control means do not conflict with each other. Once the decomposition and configuration are complete, the document serialization unit (600) can serialize the operational orders for subordinate units into a second C-BML format using a built-in format adapter structure and, if necessary, support extended conversion to other heterogeneous standard exchange formats such as USMTF. Finally, the document serialization unit (600) can control the transmission of an operation order to a subordinate unit through the communication unit (100) after adding an encrypted digital signature of the subordinate unit commander to the serialized data payload through a digital signature module linked to the defense network PKI (Public Key Infrastructure).

[0058] In one embodiment, the multimodal input unit (700) can exclude complex physical input methods such as keyboard typing to help the commander make rapid decisions in a real battlefield environment, and can receive intuitive gesture inputs such as speech-to-text (STT), sketches on the display, and drag-and-drop, and immediately convert them into normalized tactical events. The multimodal input unit (700) can maximize the efficiency of interaction by analyzing the intent of text collected through the voice module, processing simple rule-based tasks such as symbol placement on its own, and automatically routing advanced inference instructions, such as troop deployment judgments or operational analysis requests, to an offline on-device (Tier 2) lightweight model or a high-performance (Tier 3) artificial intelligence model on a server network depending on the network communication status, by linking with the task derivation and mission generation unit (400).

[0059] FIG. 2 is a flowchart illustrating a doctrine-based operational command automatic interpretation and propagation method according to an example of the present invention.

[0060] Referring to FIG. 2, in one embodiment, at operation 21, in the command reception step, the communication unit (100) can receive a first operational command (OPORD) payload serialized in the C-BML (Coalition Battle Management Language) standard format from a higher-level unit system. For example, the communication unit (100) can collect operational command data in the first C-BML format that complies with global coalition standards such as SISO-STD-011-2014 by linking with a command and control system (e.g., AI-CDSS) of a higher-level unit, such as a regiment or brigade, through a communication network. In this process, the communication unit (100) can apply a defense standard encryption system (e.g., ARIA-256-GCM) module to unlock the security of the received payload and primarily verify the integrity of the data, and by detecting the connection status of the communication network in real time considering the special environment of a battlefield, it can update and maintain a network status flag for the offline queue processing mechanism described later.

[0061] In one embodiment, in operation 22, during the deserialization and extraction step, the command interpreter (200) can deserialize and parse the received first operation command and extract the operational control line of the superior unit, information on the deployment of friendly and enemy units, and mission data of the superior unit. For example, the command interpreter (200) can analyze the payload structure of the received first C-BML format and perform deserialization to extract various operational control line data, including the Forward Edge of Battle Area (FEBA), Battle Boundary, and Coordination Line of Fire Support (CFL) set by the superior unit, as well as information on the current deployment coordinates and organization of enemy and friendly forces, and mission data and constraints describing the objectives of the superior unit, in the form of individual objects that can be utilized by the internal system. If deserialization parsing fails due to data structure damage caused by a communication failure or format compatibility issues, the command interpretation unit (200) can assist the commander in recognizing the operational situation without omission and responding manually by displaying the entire received original text in a fallback form on the user interface to prevent system errors.

[0062] In one embodiment, in operation 23, during the zone of responsibility extraction and threat analysis step, the terrain and threat analysis unit (300) can extract a unique zone of responsibility (AO) polygon of a subordinate unit based on pre-established offline terrain data and the extracted operational control line, and tactically analyze enemy infiltration approach routes and fire threat areas within the zone of responsibility. For example, the terrain and threat analysis unit (300) can automatically synthesize a zone of responsibility polygon dedicated to the subordinate unit by cutting out the operational control line information of the superior unit to match the shape of the subordinate unit's zone of responsibility using a polygon clipping technique. Additionally, by referring to a pre-established weapon catalog, the range threat area according to the type of enemy artillery ammunition (Ammunition Policy, e.g., basic round, propellant round, etc.) extracted from the first C-BML format can be precisely calculated, and enemy units located at a distance where the calculated range area does not reach the subordinate unit's zone of responsibility can be automatically filtered out and excluded from the threat targets, thereby reducing the cognitive load of the commander. Furthermore, in order to extract the enemy infiltration approach route that can enter the polygon of the aforementioned area of ​​responsibility, a hybrid search technique is performed that combines tactical majority voting based on grid data reflecting elevation and vegetation information with path search based on road network data to derive actual approach routes and entry points, and a Combat Power Ratio reflecting the weights of firepower and support units based on the deployment information of friendly and enemy units within the said area of ​​responsibility can be calculated and visually provided.

[0063] In one embodiment, in operation 24, at the core task derivation and mission sentence synthesis step, the task derivation and mission generation unit (400) can infer explicit tasks by recalculating the threat analysis results and superior unit missions from the perspective of subordinate units, and after confirming core tasks by receiving approval from the commander (user) for implicit task candidates derived using an AI model, can automatically synthesize mission sentences based on the 5W principles. For example, the task derivation and mission generation unit (400) can utilize a multimodal input unit that receives only the user's speech utterance (STT) and drawing gestures without keyboard typing and converts them into normalized events, classify the intent of the received voice text, and automatically route the inference request to either an on-device model in an offline environment or a high-performance model in a server network depending on the network status. Simultaneously, the main and secondary attack axes based on the superior unit's standards specified in the first C-BML format are recalculated into axes (main attack, secondary attack, fixation, etc.) from the perspective of the subordinate unit, based on the derived enemy infiltration approach route and the subordinate unit's area of ​​responsibility polygon. Furthermore, using the established ontology and unit organization knowledge graph, enemy unit information (e.g., infantry regiment) comprehensively described in the superior unit's operational order can be automatically expanded and concretized to the level of lower-level enemy units (e.g., tank battalion, infantry company) that the subordinate unit must directly respond to. By combining the recalculated axes and the concretized enemy unit information with the superior unit's specified tasks, the subordinate unit's independent specified tasks are automatically inferred. These are then input into the artificial intelligence model along with military doctrine and missions to generate a pool of implicit task candidates additionally required for operational execution and present them on the display.The core task is finally selected from the confirmed items and the aforementioned explicit tasks through a Human-in-the-Loop method in which the user performs confirmation or rejection input regarding the presented implicit task candidate group. When no core task is selected, the mission sentence is kept blank, and in response to a trigger that selects a core task, the mission sentence can be automatically filled in and written by combining the core task and mission information into a 5W template. Furthermore, based on the written mission sentence and threat analysis results, the Commander's Requirement for Information (CCIR) and Priority Requirement for Information (PIR) that subordinate units must verify through reconnaissance assets such as drones can be automatically inferred and generated.

[0064] In one embodiment, in operation 25, at the doctrine and ontology verification step, the ontology verification unit (500) can automatically verify whether the synthesized mission statement and established operation plan violate the doctrine based on an ontology knowledge graph and SHACL rules that reflect the military doctrine. For example, the ontology verification unit (500) may have an ontology schema and SHACL query rules extended based on the MIM (Multilateral Interoperability Programme Information Model) standard, and may model the tactical requirements specified in the military doctrine in the form of a Competency Question (CQ) and include them in the ontology schema. By utilizing this, the ontology verification unit (500) can strictly and automatically verify at the data level whether the placement of friendly positions on the position placement data satisfies the daytime visual observation (LOS) conditions on the offline terrain data or the concealment conditions from enemy threats by comparing the automatically generated mission sentence, the information requirements, and the position placement data drawn by the commander on the display with the eligibility query and shackle rule. If the verification results fail to meet tactical conditions or if contradictions are found due to hallucination errors in the generative artificial intelligence model, the ontology verification unit (500) can immediately issue a warning alarm to the user and reject the corresponding plan containing the mission sentence and position placement data to ensure the integrity of the operation.

[0065] In one embodiment, in operation 26, during the sub-unit assignment and serialization step, the document serialization unit (600) can decompose the sub-unit operation plan for which doctrine verification has been completed into each sub-unit (e.g., company) unit, assign responsibility areas and firepower, and then serialize and assemble it into a second C-BML format. For example, the document serialization unit (600) can divide the verified operation plan into units of sub-units based on the sub-unit standard (e.g., sub-companies based on the battalion standard), and individually assign the boundary lines of the responsibility area to be handled by each sub-unit and the assigned firepower assets to constitute operation orders for each sub-unit. In this process, to prevent operational confusion, the document serialization unit (600) can process the coordinates of targets and control means by metadataizing them into a Single Source of Truth (SSOT) based on unit location, thereby ensuring automatic synchronization without data conflicts between each sub-unit. The operational orders assigned to each sub-unit are serialized into the above-mentioned second C-BML format, and at this time, the document serialization unit (600) can achieve interoperability with various Allied Forces systems by selectively supporting conversion and expansion into heterogeneous standard exchange formats including USMTF (United States Message Text Format) according to the data transmission and reception target system using a built-in format adapter structure.

[0066] In one embodiment, in operation 27, at the electronic signature and transmission step, the doctrine-based operational command automatic interpretation and dissemination system (10) may add the commander's electronic signature to the serialized second C-BML operational command and transmit it to the corresponding subordinate unit system through the communication unit (100). For example, the system (10) may be equipped with an electronic signature module linked to a defense network PKI (Public Key Infrastructure) server, and may grant legal validity and reliability by generating and adding an encrypted electronic signature proving the commander of the subordinate unit to the data payload of the second C-BML format serialized through the document serialization unit (600). The communication unit (100) evaluates the communication environment before transmitting the command, and if it is identified that the server communication network is in an offline state with the network disconnected due to the battlefield environment, such as terrain obstacles or enemy electronic warfare jamming, it may load the operation command in the second C-BML format with the electronic signature attached into an offline queue within the system and keep it waiting to preserve it without loss. Subsequently, as soon as a signal for network connection restoration is detected, the communication unit (100) may automatically transmit the data loaded in the offline queue in sequence according to a pre-set urgency priority order, in the order of Fragment Command (FRAGO), Operation Command (OPORD), and general artificial intelligence inference request, taking into account the urgency of the operation. Consequently, the system of the subordinate unit that receives the second C-BML format operation command through the communication unit (100) can configure the system so that terrain-based analysis, AI-based task generation, and divisional delivery procedures are transmitted sequentially without interruption or error through the same interface standard from the highest-level brigade unit to the lowest-level sub-unit by treating the received second C-BML format operation command as an operation command of the upper unit from its own perspective and recursively repeating the operation of the command interpretation unit (200) or the transmission operation of the communication unit (100).

[0067] FIG. 3 is a flowchart illustrating a terrain-based responsibility zone synthesis and enemy threat analysis method according to an example of the present invention.

[0068] Referring to FIG. 3, in one embodiment, at operation 31, during the upper control line loading step, the terrain and threat analysis unit (300) can load coordinate data of the upper unit's operational control line (FEBA, etc.) extracted in operation 22. For example, the terrain and threat analysis unit (300) can call into offline memory an array of spatial coordinates of key linear elements that define an operational area, such as the forward boundary line (FEBA), combat boundary line (Boundary), fire support coordination line (CFL), and control line (Phase Line), from the operational command data structure (ParsedOPORD) of the upper unit (e.g., brigade) parsed by the command analysis unit (200). Since the data loaded at this time is comprehensive geographic information created from a macroscopic perspective of the upper command, the terrain and threat analysis unit (300) can perform coordinate system synchronization and rendering scale matching between the upper command control line data and internal offline terrain data bundles (e.g., elevation data, raster terrain features, etc.) as a preliminary preparation task to process it to fit the actual operational radius of the subordinate unit.

[0069] In one embodiment, in operation 32, during the AO polygon synthesis step, the terrain and threat analysis unit (300) can automatically synthesize a subordinate unit's exclusive area of ​​responsibility (AO) polygon by using a polygon clipping technique to cut out only the area corresponding to the left and right lateral and forward limits of the subordinate unit from the superior unit's operational control line information. For example, the terrain and threat analysis unit (300) can mathematically calculate the unique area of ​​responsibility limits specifically assigned to the subordinate unit (e.g., Charlie Battalion) by applying a curve-preserving polygon clipping technique, such as the Greiner-Hormann algorithm, to multiple control line data displayed on the superior unit's vast operational map. By doing this, instead of loading and analyzing the entire massive operational map of the superior unit, only the area inside the closed curve where the left and right lateral boundary lines and the front and rear boundary lines of the subordinate unit intersect is precisely cut out to create an area of ​​responsibility polygon (AOPolygon) dedicated to the subordinate unit, thereby drastically reducing the scope of spatial analysis operations to be performed thereafter and promoting the speed of data processing.

[0070] In one embodiment, in operation 33, during the weapon catalog linkage threat analysis step, the terrain and threat analysis unit (300) may calculate a range threat area based on the enemy artillery ammunition type information included in the superior unit command by referring to a pre-established weapon catalog, and may automatically filter and exclude enemy units from the threat analysis target if the calculated area does not reach the polygon of the responsibility area dedicated to the subordinate unit. For example, the terrain and threat analysis unit (300) may compare and map the specifications (caliber, weapon category, etc.) of the enemy artillery unit extracted from the first C-BML format by the command interpretation unit (200) with the pre-established KPA / ROK JSON weapon catalog. Based on the comparison results, the range threat area can be precisely calculated when the enemy artillery uses a standard round and when it uses a special round (Ammunition Policy), such as an extended round. In the calculation process, enemy units determined to have a maximum range that does not reach within the polygon boundary of the subordinate unit's area of ​​responsibility synthesized in Operation 32, even if the type of ammunition is changed to a 120mm mortar, are considered as meaningless threats that cannot directly strike the subordinate unit and can be automatically filtered out and excluded from the analysis target and display target. Through this intelligent filtering based on ammunition type and range, the doctrine-based operational command automatic interpretation and dissemination system (10) can significantly reduce the cognitive load so that the commander can focus only on the key threats facing him (e.g., a self-propelled artillery battalion with a long range).

[0071] In one embodiment, in operation 34, at the step of calculating the combat power ratio, the terrain and threat analysis unit (300) can calculate a weighted Combat Power Ratio based on the organization and firepower information of friendly and enemy units deployed within the polygon of the responsibility area and provide it to the commander screen. For example, the terrain and threat analysis unit (300) can go beyond simply comparing the number of ground troops on a one-to-one basis and analyze the detailed organization knowledge graph of friendly and enemy units identified within the polygon of the responsibility area dedicated to the subordinate unit to sum the basic combat power index possessed by each unit node. In addition, a substantial and comprehensive Combat Power Ratio can be calculated by converting the power of firepower support assets, such as artillery power, mechanized units, and air support assigned to the unit, into doctrinal weights and multiplying them. The calculated result value (e.g., friendly forces vs. enemy forces 1 to 1.7) can be evaluated to determine whether it is within the doctrinal limit of the defensive operation (e.g., 1 to 3), and can be visually provided through a user interface linked with a multimodal input unit (700) so that the commander can intuitively determine whether the current operational environment is advantageous or disadvantageous to friendly forces.

[0072] In one embodiment, in operation 35, during the hybrid approach route search step, the terrain and threat analysis unit (300) can derive an approach route and an entry point by performing a hybrid search technique that combines a tactical majority decision based on offline grid data and a path search based on road graph data to extract an enemy infiltration approach route that can enter the polygon of the responsibility area. For example, when analyzing the path through which an attack arrow of a superior unit penetrates the responsibility area of ​​a subordinate unit, the terrain and threat analysis unit (300) can determine whether the area is a terrain where tanks or vehicles can maneuver using a majority method (Grid BFS) by using mobility grid data that reflects terrain elevation and vegetation information and is searchable with a space complexity of O(1). At the same time, by performing a forward BFS on road network data (Road Graph) that includes highways or national roads (e.g., Jungbu Inland Expressway, Yeongnam-daero, etc.), it is possible to combine and search for actual enemy maneuver routes that include detours in mountainous terrain. Through this hybrid search technique, the terrain and threat analysis unit (300) can classify the types of enemy infiltration approach routes that can enter the polygon of the responsibility area into categories such as penetrates and enters and ends, and derive multiple expected entry points (line segment intersections) through which the enemy can actually cross into our subordinate unit's responsibility area and transmit them to the task derivation and mission creation unit (400).

[0073] Referring to FIG. 4, in one embodiment, in operation 41, at the multimodal input receiving step, the multimodal input unit (700) can receive the user's speech utterance (STT) and drawing gesture input on the map and convert them into normalized tactical events. For example, in the urgent and special environment of a battlefield, it may be physically restricted for a commander or staff officer to input complex text by keyboard typing, so the multimodal input unit (700) may be configured to receive only the user's speech utterance (STT) and drawing gesture on the display without keyboard typing. The multimodal input unit (700) can recognize strokes sketched by the user on an operational map using touch or a stylus pen by linking with a built-in symbol tree module and classify them into linear elements such as arrows, boundary lines, and cooperation lines, or detect drag and drop events in which specific military symbols are dragged and dropped onto the map. At the same time, the multimodal input unit (700) converts a user’s utterance, such as “add 2nd Company defense symbol,” into text through a voice recognition engine, and normalizes this complex multimodal input data into standardized events (e.g., symbol placement event, sketch result object, etc.) that the system can interpret and transmit them to other components.

[0074] In one embodiment, in operation 42, during the axis recalculation and enemy unit deployment phase, the task derivation and mission generation unit (400) can recalculate the main attack / supporting attack axes in the superior unit's command from the perspective of the subordinate unit based on the enemy infiltration approach route derived in operation 35, and can concretize the superior enemy unit by automatically deploying it into the subordinate enemy unit that the subordinate unit will respond to using an organizational knowledge graph. For example, in the operation order issued by the superior unit (e.g., brigade), only the macroscopic main attack and supporting attack axes from the superior unit's perspective are indicated, but within the responsibility area of ​​the subordinate unit (e.g., Charlie Battalion), the meaning of the axes may be tactically completely different. Accordingly, the task derivation and mission generation unit (400) can analyze the main and auxiliary attack axes based on the superior unit standards specified in the operational order of the first C-BML format based on the extracted enemy infiltration approach route and the polygon of the responsibility area dedicated to the subordinate unit, and recalculate them into new tactical axes (e.g., main attack, auxiliary attack, fixation, etc.) from the perspective of the subordinate unit. In addition, since enemy unit information is often listed only at a general level of higher echelons, such as '1st Infantry Regiment', in the superior unit operational order document, the task derivation and mission generation unit (400) can use the established military ontology and unit organization knowledge graph to automatically deploy the comprehensive enemy unit information included in the operational order of the first C-BML format to the level of lower enemy units (e.g., subordinate tank battalion, mechanized infantry company, etc.) that the subordinate unit must directly confront and respond to, thereby subdividing the target of the operational analysis into specific and actual threat units.

[0075] In one embodiment, in operation 43, during the automatic specification task inference step, the task derivation and mission generation unit (400) can combine the recalculated axis and detailed enemy unit information with the subordinate unit's area of ​​responsibility to automatically infer a unique specification task of the subordinate unit that is not specified in the superior unit's operational order. For example, since the superior unit does not kindly issue specific detailed tasks to be performed by the subordinate unit, the task derivation and mission generation unit (400) can logically combine the recalculated axis and the detailed information at the subordinate enemy unit level with the superior unit's specification task (e.g., "repel enemy attack toward Daegu") specified in the operational order of the first C-BML format and the location data of the subordinate unit's area of ​​responsibility. Through this, the task derivation and mission generation unit (400) can automatically infer and generate unique and specific tasks for the subordinate unit that are not specified in the operational orders of the superior unit, such as "repel the attack capabilities of the enemy 14th Tank Battalion infiltrating through the eastern approach route on the right side of the FEBA" or "clear out the remaining special forces units within the area of ​​responsibility," thereby drastically reducing the manual work required by the staff.

[0076] In one embodiment, in operation 44, during the AI ​​inference routing and implicit task generation step, the task derivation and mission generation unit (400) can automatically route the inferred explicit task and military doctrine information to an appropriate model among an on-device (Tier 2) or server (Tier 3) AI model to request inference, and generate a group of implicit task candidates additionally required for operation execution through the AI ​​model and present them to the user screen. For example, the task derivation and mission generation unit (400) can classify the intent of voice text received through the multimodal input unit (700) in real time, and in the case of simple symbol mapping or basic operation modification instructions where security is important, it can automatically route the inference request to an on-device model (Tier 2, lightweight LLM) that operates independently in an offline environment. On the other hand, in cases where advanced computation is required, such as complex wargaming computation or the establishment of an overall operation plan, an intelligent distributed processing architecture can be implemented that routes to a high-performance model (Tier 3) of a server network only when communication is connected. Afterward, the task derivation and mission generation unit (400) inputs the specified tasks of the subordinate unit, the military doctrine, and the mission of the superior unit that were automatically inferred, into the prompt of the selected artificial intelligence model, and generates a plurality of implicit tasks (e.g., "reinforcing the main supply route," "constructing a temporary pontoon bridge for crossing," etc.) that must be performed incidentally to complete the current operation, and then presents them as visual cards on a user screen (tablet display, etc.) linked with the multimodal input unit (700).

[0077] In one embodiment, in operation 45, at the Human-in-the-Loop core task confirmation stage, the doctrine-based operational command automatic interpretation and dissemination system (10) receives a user's confirmation or rejection input (Human-in-the-Loop) regarding the presented implicit task candidate group, confirms the implicit task, and can finally select a core task from among the confirmed explicit and implicit tasks. For example, to exclude the risk that the implicit task candidates proposed by the generative AI model lack military realism or are detached from the current battlefield situation, the system (10) may be designed not to confirm the operation by relying entirely on the judgment of the AI. That is, after viewing the implicit task candidate group presented on the screen, the commander or operations staff may perform an input to accept (confirm) or reject (reject) a specific item through a display touch or voice command. The task derivation and mission generation unit (400) can establish a highly reliable decision-making system based on human control (Human-in-the-Loop) by combining the list of implicit task items that are finally confirmed after passing the user's confirmation or rejection input and the list of explicit tasks that were previously derived, and by selecting the core tasks (e.g., 1 to 2) that have the most decisive influence on the success or failure of the operation from among these by receiving the commander's final selection.

[0078] In one embodiment, in operation 46, during the conditional 5W mission statement synthesis step, the task derivation and mission generation unit (400) checks whether the core task is selected, and if the core task is not selected, keeps the mission statement in an empty state, and in response to a trigger in which the core task is selected, can automatically fill and create a mission statement by combining the core task and mission information in a 5W template. For example, since a mission statement must clearly specify who, what, when, where, and why must be done according to military doctrine, the task derivation and mission generation unit (400) can keep the mission statement text box in an empty state to prevent violations of specifications caused by free text input in a waiting state where the core task is not selected. Subsequently, in response to a trigger signal in which the core task is firmly selected by the commander through operation 45, the task derivation and mission generation unit (400) can inject data objects into each slot of a pre-set 5W template. That is, by combining mission information such as the subject (Charlie Battalion), time (immediately upon receipt), place (key control point within the area of ​​responsibility), target and action (confirmed core task content), and purpose (achievement of higher intentions through higher unit missions), and automatically filling in and creating the above-mentioned mission sentence in a logical and seamless manner, the time spent by the staff drafting orders from scratch can be reduced, and the formality of the document can be perfectly guaranteed.

[0079] FIG. 5 is a flowchart illustrating an ontology and doctrine-based operational plan verification method according to one embodiment of the present invention.

[0080] Referring to FIG. 5, in one embodiment, in the CCIR / PIR automatic generation step at operation 51, the task derivation and mission generation unit (400) can automatically infer and generate Commander Information Requirements (CCIR) and Priority Information Requirements (PIR) that subordinate units must verify through reconnaissance assets based on the mission sentence and threat analysis results created in operation 46. For example, the task derivation and mission generation unit (400) can synthesize terrain and threat data such as the enemy's main attack direction or expected infiltration approach route (e.g., Route 903), and the expected location of enemy artillery units (e.g., Nakdong River bank) analyzed earlier, and automatically generate Priority Information Requirements (PIR) in the form of specific text regarding where to deploy reconnaissance assets, such as drones possessed by subordinate units, and what to verify. At the same time, a comprehensive set of Commander's Required Information (CCIR) can be constructed by deriving both Frontline Allied Information (FFIR) and Essential Security Information (EEFI) that a commander must be informed of for operational control, such as the readiness of friendly defensive positions or company-level casualties during engagement. Through this, the commander can intuitively verify the information requirements proposed by AI on the screen and, if necessary, add or modify additional requirements via speech-to-text (STT) to secure an informational advantage in operations.

[0081] In one embodiment, in operation 52, during the ontology schema and shackle rule loading step, the ontology verification unit (500) can load military doctrine ontology schemas and shackle query rules extended based on the MIM (Multilateral Interoperability Programme Information Model) standard. For example, the ontology verification unit (500) can build its own command and control (C2) ontology schema (TBox) and load it into memory by specializing and extending key tactical entities, such as detailed mission tasks or fire support tasks, which are not specifically defined in the existing MIM standard but are based on the MIM standard model established for smooth data exchange between allied forces such as NATO, to suit the purpose of the system. In addition, the ontology verification unit (500) can complete a foundational system capable of logically evaluating instance data flowing in at subsequent stages by loading verification rules written in SHACL, a W3C standard, and SPARQL query templates linked thereto that explore the knowledge graph into the runtime verification environment to constrain the structural consistency and attribute values ​​of the data.

[0082] In one embodiment, in operation 53, at the AI ​​product logic verification step, the ontology verification unit (500) can block hallucination errors of the artificial intelligence model at the data level by comparing the automatically generated mission sentence, the commander information requirements and priority information requirements, and the input position deployment data with the shackle query rules. For example, the ontology verification unit (500) can project operational elements derived by the artificial intelligence model or user (e.g., unit identifiers, time, target coordinates, and various symbol data drawn by drag and drop on a map within the mission sentence) into the form of an ABox, which is instance data within the ontology knowledge graph. By passing these structured instance data through a pre-loaded set of SHACL verification rules, the ontology verification unit (500) can precisely determine from the logical structure of the data itself whether assets that must be essential to a specific task are missing, or whether there is a contradiction in the temporal / spatial sequence. Through this data-level constraint comparison, the ontology verification unit (500) can fundamentally filter and block the hallucinatory phenomenon in which a generative artificial intelligence based on a massive language model (LLM) plausibly suggests fictional facts or impossible measures that are completely unsuitable for a military context.

[0083] In one embodiment, in operation 54, at the step of verifying tactical conditions based on a Competency Question (CQ), the ontology verification unit (500) models the tactical requirements specified in the military doctrine in the form of a Competency Question (CQ) and includes them in the ontology schema, and can automatically verify through the Competency Question and the Shackle Query rule whether the placement of friendly positions included in the position placement data satisfies the daytime visual observation conditions on the offline terrain data or the concealment conditions from enemy threats. For example, the ontology verification unit (500) may quantify and logically organize the essential principles of defense and attack operations described in the military manual into about 45 or more Competency Questions (CQs). The ontology verification unit (500) can calculate terrain elevation data based on clear daytime weather to clearly control the positions of subordinate units at the battalion headquarters command post and query whether the line of sight (LOS) is secured, or query whether friendly artillery positions are deployed in terrain with certain concealment, such as behind a ridge, so as not to be exposed to enemy counter-artillery radar or direct fire, in conjunction with terrain bundles and threat analysis results. By automatically performing these eligibility queries through SPARQL calculations and SHACL rules, it is possible to precisely evaluate whether the measures established by the commander or AI are thoroughly in accordance with doctrine in terms of terrain and tactics.

[0084] In one embodiment, in operation 55, at the rejection and alarm stage, the ontology verification unit (500) may generate an alarm on the commander screen and reject the corresponding mission statement and position deployment data to induce modification if the verification result of operation 53 or operation 54 fails to meet the conditions. For example, the ontology verification unit (500) may be configured to derive one of four health statuses—Pass, Warning, Fail, or Unknown—for the operation plan data subject to verification. If a failure to meet conditions is identified, such as when a position of a specific company is completely exposed to enemy firepower and receives a Fail judgment in the concealment CQ verification, or when it is confirmed to be deployed outside the operational control line range, the ontology verification unit (500) may immediately display a red warning alarm or a pop-up message on the user interface. At the same time, the ontology verification unit (500) blocks state transitions and rejects drafts so that the corresponding position deployment data or defective mission sentences that are contrary to doctrine cannot be transmitted to subordinate units, thereby strongly inducing the commander to readjust the position of the position to a safe rear area or correct the content of the task to conform to military doctrine, thereby minimizing the risk of operational failure on the battlefield.

[0085] FIG. 6 is a flowchart illustrating a subordinate command partitioning assignment and serialization method according to one embodiment of the present invention.

[0086] Referring to FIG. 6, in one embodiment, in operation 61, during the operation plan assignment and decomposition step, the document serialization unit (600) can divide the verified integrated operation plan of the subordinate unit into subordinate unit units (e.g., company) based on the subordinate unit standards, and individually assign the responsibility area and assigned firepower information to each subordinate unit to configure operation orders for each subordinate unit. For example, the document serialization unit (600) can perform a procedure to independently decompose data for each subordinate subordinate unit (company unit) that will actually execute the integrated operation plan of the entire subordinate unit (battalion unit) whose doctrinal integrity has been secured through the ontology verification unit (500). At this time, based on the overall operation control line and defense data drawn on the map by the commander through the multimodal input unit (700), the spatial boundary line of the unique responsibility area (AO) that each subordinate unit is to be responsible for can be geometrically divided and assigned. In addition, the document serialization unit (600) can configure customized operational command data so that when a specific subordinate unit views the command through a receiving system, it can clearly identify only the specific mission sentence, explicit and implicit tasks, control line, and assigned target assigned to it, and immediately perform operations by individually distributing (assigning) firepower and support forces, such as artillery, armor, and reconnaissance assets assigned to the entire battalion, which are assigned by the superior unit, according to the weight of the main mission and expected intensity of engagement of each subordinate unit.

[0087] In one embodiment, in operation 62, during the SSOT coordinate synchronization step, the document serialization unit (600) can automatically synchronize the coordinates of targets and control means without data conflicts by managing them as Single Source of Truth (SSOT) data based on unit locations during the allocation process. For example, to fundamentally prevent data fragmentation or operational coordinate inconsistency problems that may inevitably occur during the process of dividing and replicating a macroscopic operational plan into operational orders for multiple sub-units, the document serialization unit (600) can metadataize and integrate the spatial coordinate data of all tactical objects, such as target locations, fire support coordination lines (CFL), and battle boundary lines, based on a Single Source of Truth (SSOT) architecture. That is, the coordinate data is not individually hardcoded or redundantly stored in each document file in which it is divided, but can be structured in the form of a pointer that references a central single object (e.g., a global object based on unit locations having a unique identifier). This ensures that when a commander of a higher-level unit fine-tunes the location of a specific target or modifies the operational control line, coordinate synchronization occurs immediately and automatically within the operational command data of all related subordinate units without data consistency conflicts, thereby enabling a flexible response to real-time changes in the battlefield situation.

[0088] In one embodiment, in operation 63, at the format adapter-based serialization step, the document serialization unit (600) serializes the assigned and configured sub-unit operational orders into a second C-BML format using an embedded format adapter structure, and, if necessary, performs conversion and expansion into a heterogeneous standard exchange format such as USMTF. For example, the data set of each sub-unit operational order that has been assigned and configured can be converted into a payload in a second C-BML (Coalition Battle Management Language) format (e.g., JSON or XML format) that complies with the SISO-STD-011-2014 standard, which is the official data exchange standard between allied forces such as NATO. At this time, the document serialization unit (600) does not stop at simply serializing into a fixed C-BML format, but can selectively support data structure conversion and expansion into various heterogeneous standard exchange formats, including USMTF (United States Message Text Format), in accordance with the specifications of the data transmission and reception target system by utilizing a flexible format adapter architecture embedded within the system. Through this, the generated operational orders can be seamlessly linked with the old command and control systems of the U.S. military or other allied forces without separate manual conversion work, thereby maximizing interoperability.

[0089] In one embodiment, in operation 64, during the digital signature addition step, the digital signature unit of the doctrine-based operational order automatic interpretation and dissemination system (10) is linked with the National Defense Network PKI, so that the encrypted digital signature of the commander of the corresponding subordinate unit can be added to the serialized second C-BML format data payload. For example, in order for the generated second C-BML format operational order to fully possess military and legal validity and be formally issued to subordinate units, the digital signature module (M4-4) within the system (10) can be securely linked with a National Defense Network PKI (Public Key Infrastructure) server authentication system in the form of a closed network where security is strictly maintained. The system (10) can generate and add a digital signature value that cryptographically proves the identity of the commander of the corresponding subordinate unit (e.g., battalion commander) to the entire serialized second C-BML data payload by utilizing a secure hash function such as SHA-256 and an RSA asymmetric key algorithm. In this process, the highest level military standard encryption method, such as ARIA-256-GCM, may be applied, and as soon as the signature is completed, the system (10) guarantees that the document has not been tampered with during the transmission process and transmits it to the communication unit (100) to automatically trigger a subsequent offline queue loading waiting procedure or a network transmission procedure to a subordinate unit system (transition to S6 transmission state).

[0090] FIG. 7 is a flowchart illustrating an offline queue management and chain propagation method according to one embodiment of the present invention.

[0091] Referring to FIG. 7, in one embodiment, in operation 71, at the communication status check step, the communication unit (100) can check the network connection status before transmission to the subordinate unit system. For example, the communication unit (100) can evaluate the availability of communication with other systems (e.g., AI-CDSS) by receiving network broadcasts of the operating system in real time or performing periodic polling. In a battlefield environment, the server communication network between command posts may be frequently disconnected due to radio shadowing in mountainous terrain, adverse weather conditions, or enemy electronic warfare and communication jamming attacks. Therefore, the communication unit (100) can determine subsequent operations by precisely classifying the current communication connection status into status codes such as 'connected', 'disconnected', or 'reconnecting' just before a data transmission command is issued.

[0092] In one embodiment, in operation 72, during the offline queue loading step, the communication unit (100) may load the operation command in the second C-BML format with an electronic signature added in operation 64 into an offline queue within the communication module and keep it waiting when the network is in an offline state where it is disconnected. For example, if the communication unit (100) confirms that the server communication network is in an offline state where it is disconnected due to the battlefield environment, it may safely load the payload of the operation command in the second C-BML format with the electronic signature added into an internal non-volatile storage-based offline queue to prevent data loss and ensure the continuity of operations. At this time, the communication unit (100) may display a status notification such as "1 item waiting to be transmitted" on the user interface so that the commander can intuitively recognize that the operation command has not yet been issued and is in a waiting state, and may support the safe queuing of data in the background without any separate retry operation by the user. In addition, this offline queue can hold not only documents in C-BML format but also complex AI inference request data, such as Wargaming, which needs to be routed to high-performance AI models (Tier 3) on the server network.

[0093] In one embodiment, in operation 73, during the urgency-based transmission step, when a network connection recovery signal is detected, the communication unit (100) can automatically transmit data loaded in the offline queue to the subordinate unit system according to a preset urgency priority (e.g., FRAGO > OPORD > AI inference request). For example, as soon as the network connection recovery signal is detected through background monitoring, the communication unit (100) can process dequeue and transmission by scheduling according to a preset urgency priority that reflects the urgency of the operation, instead of transmitting the data loaded in the offline queue in a simple first-in, first-out (FIFO) manner. Specifically, the transmission of a Fragment Order (FRAGO) C-BML payload, which requires immediate action due to rapid changes in the battlefield situation, is assigned as 1st priority (Urgent), the transmission of a Regular Operation Order (OPORD) C-BML is assigned as 2nd priority (Normal), and general artificial intelligence inference requests to assist in establishing operation plans are assigned as 3rd priority (Normal), thereby ensuring that important operational instructions reach subordinate units first. When the transmission of all items loaded in the queue is completed and the queue is emptied, the communication unit (100) can clearly provide feedback to the commander on the success of the operation order delivery by generating a pop-up event on the display that says "Subordinate unit propagation completed."

[0094] In one embodiment, in operation 74, during the sub-unit recursive execution step, a sub-unit system linked to the doctrine-based operational command automatic interpretation and dissemination system (10) can receive an operational command in the second C-BML format through the communication unit (100), treat it as an operational command of the upper unit, and recursively repeat the terrain-based interpretation, AI task generation, and divisional delivery operations of operations 21 to 27. For example, a sub-unit system (e.g., company) having the same architecture as the upper unit system that received the operational command in the second C-BML format through the communication unit (100) can treat the received operational command in the second C-BML format as the first operational command of the upper unit (e.g., battalion) from its own perspective and recursively repeat the entire pipeline operation from the command interpretation unit to the communication unit. In other words, since transmission and reception operate based on the same C-BML interface that is perfectly compatible with each other, the transmission results of one echelon (e.g., battalion) naturally interlock with the reception input of its subordinate unit (e.g., company). Through this, without the need to build a separate format conversion module or the staff to manually re-enter data, the operational intent initially issued from the highest-level unit (e.g., brigade) is configured to sequentially propagate terrain-based interpretation, task creation, and divisional delivery procedures down to the lowest-level unit (e.g., platoon) through the same interface. This configuration ensures perfect operational consistency between units and dramatically improves the speed of command decision delivery, even in large-scale multinational combined operations or complex inter-echelon operational environments.

[0095] Meanwhile, the disclosed embodiments may be implemented in the form of a recording medium that stores instructions executable by a computer. The instructions may be stored in the form of program code and, when executed by a processor, may generate a program module to perform the operation of the disclosed embodiments. The recording medium may be implemented as a computer-readable recording medium.

[0096] Computer-readable recording media include all types of recording media that store instructions that can be decoded by a computer. Examples include ROM (Read Only Memory), RAM (Random Access Memory), magnetic tape, magnetic disk, flash memory, optical data storage devices, etc.

[0097] As described above, the disclosed embodiments have been explained with reference to the attached drawings. Those skilled in the art will understand that the present disclosure may be practiced in forms different from the disclosed embodiments without changing the technical spirit or essential features of the present disclosure. The disclosed embodiments are illustrative and should not be interpreted restrictively. Explanation of the symbols

[0098] 10: Doctrine-based Operational Command Automatic Interpretation and Dissemination System 100: Communications Department 200: Command interpreter 300: Terrain and Threat Analysis Division 400: Task Derivation and Mission Creation Unit 500: Ontology Validation Unit 600: Document Serializer 700: Multimodal Input

Claims

Claim 1 A command and control system that receives an operational order from a superior unit and automatically interprets and transmits it as an operational order from the perspective of a subordinate unit, comprising: a communication unit that receives an operational order serialized in a first C-BML (Coalition Battle Management Language) format from a superior unit; and a command interpretation unit that deserializes the operational order in the first C-BML format to extract the operational control line of the superior unit, information on the deployment of friendly and enemy units, and the mission of the superior unit. A doctrine-based operational command automatic interpretation and dissemination system characterized by comprising: a terrain and threat analysis unit that extracts a unique Area of ​​Operations (AO) of the subordinate unit based on the extracted operational control line and pre-established offline terrain data, and analyzes enemy infiltration approach routes and firepower threats within the Area of ​​Operations; a task derivation and mission generation unit that infers explicit tasks by recalculating the threat analysis results and superior unit missions from the perspective of the subordinate unit, derives implicit task candidates using an artificial intelligence (AI) model, and generates mission sentences based on core tasks confirmed by user approval; an ontology verification unit that automatically verifies whether the generated mission sentences and the subordinate unit's operational plans violate doctrine based on an ontology knowledge graph reflecting military doctrine; and a document serialization unit that assigns the verified operational plans of the subordinate unit to each subordinate unit of the subordinate unit, serializes them into a second C-BML format, and controls transmission through the communication unit. Claim 2 A doctrine-based operational command automatic interpretation and dissemination system, wherein, in paragraph 1, the terrain and threat analysis unit automatically synthesizes a polygon of responsibility area dedicated to the subordinate unit by cutting out the operational control line information of the superior unit to match the shape of the subordinate unit's responsibility area through a polygon clipping technique. Claim 3 In paragraph 2, the terrain and threat analysis unit calculates a range threat area based on the enemy artillery ammunition type (Ammunition Policy) extracted from the first C-BML format by referring to a pre-established weapon catalog, automatically filters out and excludes enemy units that do not reach the responsibility area of ​​the subordinate unit from the threat targets, and calculates and provides a weighted ratio of friendly and enemy combat power based on information on the deployment of friendly and enemy units within the responsibility area, characterized by a doctrine-based operational order automatic interpretation and dissemination system. Claim 4 In paragraph 3, the doctrine-based operational command automatic interpretation and dissemination system is characterized in that the terrain and threat analysis unit derives approach routes and entry points by performing majority voting based on grid data and a hybrid search technique based on road network data to extract the enemy infiltration approach routes capable of entering the polygon of the responsibility area. Claim 5 In paragraph 4, the task derivation and mission generation unit recalculates the main and auxiliary attack axes based on the superior unit standards specified in the first C-BML format into axes based on the subordinate unit perspective, based on the derived enemy infiltration approach route and the responsibility area polygon, and automatically deploys the enemy unit information included in the superior unit operational order to the level of the subordinate unit that the subordinate unit must directly respond to, using the previously constructed ontology and unit organization knowledge graph, thereby concretizing it as an analysis target; a doctrine-based operational order automatic interpretation and dissemination system. Claim 6 In paragraph 5, the task derivation and mission generation unit combines the recalculated axis and specific enemy unit information with the specified task of the superior unit and the subordinate unit's area of ​​responsibility to automatically infer the independent specified task of the subordinate unit that is not specified in the superior unit's operational order, thereby forming a doctrine-based operational order automatic interpretation and dissemination system. Claim 7 A doctrine-based operational command automatic interpretation and dissemination system according to claim 6, further comprising a multimodal input unit that receives only the user's speech-to-text (STT) and drawing gestures without keyboard typing and converts them into normalized events, wherein the task derivation and mission generation unit classifies the intent of the voice text received through the multimodal input unit and automatically routes an inference request to either an on-device model in an offline environment or a high-performance model in a server network. Claim 8 In claim 7, the task derivation and mission generation unit inputs the inferred explicit task, military doctrine, and mission into the artificial intelligence model to generate a group of implicit task candidates including the implicit task candidates additionally required for the execution of operations, presents this group to a user screen linked with the multimodal input unit, and finally selects the core task from among the specified tasks and the confirmed items through the user's confirmation or rejection input regarding the group of implicit task candidates. This characterizes a doctrine-based operational command automatic interpretation and dissemination system. Claim 9 In claim 8, the task derivation and mission generation unit maintains the mission sentence in an empty state when the core task is not selected, and automatically fills and creates the mission sentence by combining the core task and mission information in a 5W1H template in response to a trigger in which the core task is selected, thereby forming a doctrine-based operational command automatic interpretation and dissemination system. Claim 10 A method performed by a command and control system that receives an operational order from a superior unit and automatically interprets and transmits it as an operational order from the perspective of a subordinate unit, comprising: a step in which a communications unit receives an operational order serialized in a first C-BML (Coalition Battle Management Language) format from a superior unit; and a step in which a command interpretation unit deserializes the operational order in the first C-BML format to extract the operational control line of the superior unit, information on the deployment of friendly and enemy units, and the mission of the superior unit. A method for automatically interpreting and disseminating doctrine-based operational orders, characterized by comprising: a terrain and threat analysis unit that extracts a unique Area of ​​Operations (AO) of the subordinate unit based on the extracted operational control line and pre-established offline terrain data, and analyzes enemy infiltration approach routes and firepower threats within the Area of ​​Operations; a task derivation and mission generation unit that infers explicit tasks by recalculating the threat analysis results and superior unit missions from the perspective of the subordinate unit, derives implicit task candidates using an artificial intelligence (AI) model, and then generates mission sentences based on core tasks confirmed by user approval; an ontology verification unit that automatically verifies whether the generated mission sentences and the subordinate unit's operational plans violate doctrine based on an ontology knowledge graph reflecting military doctrine; and a document serialization unit that assigns the verified operational plans of the subordinate unit to each subordinate unit of the subordinate unit, serializes them into a second C-BML format, and controls transmission through the communication unit.

Images