Task intention scheduling execution method, device and equipment for dynamic scene

Abstract

The application relates to a dynamic scene-oriented task intention scheduling execution method, device and equipment. The method comprises the following steps: constructing an intention block library, the intention block library stores a plurality of intention blocks, the intention blocks are pre-defined function types and are bound with three-dimensional scheduling attributes; receiving a target task of an industrial robot, calling a task model matched with the target task; determining a scene type in which the industrial robot is located according to sensor collection information, selecting a corresponding scheduling strategy according to the scene type, screening an intention block to be executed in this round according to the scheduling strategy and three-dimensional scheduling attributes of each node intention block in a directed acyclic graph; performing conflict resolution on the intention block to be executed, generating an execution queue according to the conflict resolution result, and calling corresponding hardware and software resources to execute corresponding functions according to the three-dimensional scheduling attributes of the intention block in the execution queue. The method can enable the industrial robot to efficiently, safely and autonomously complete the target task in a dynamic industrial scene.

Description

Technical Field

[0001] This application relates to the field of industrial robot technology, and in particular to a method, apparatus and device for scheduling and executing task intents in dynamic scenarios. Background Technology

[0002] With the continuous development of industrial robot technology, industrial robot control technology based on preset programs or fixed processes has emerged. The core feature of this technology is that it enables industrial robots to execute basic industrial tasks according to pre-programmed instruction sequences. In production scenarios with simple tasks and structured environments, standardized work processes can be constructed, effectively replacing manual labor in repetitive and mechanical operations. This provided technical support for the initial improvement of industrial production efficiency in the early stages. Consequently, in the current field of industrial robots, most equipment still commonly adopts this preset program or fixed process-based control method to carry out daily task execution.

[0003] However, current control methods based on preset programs or fixed processes reveal numerous problems such as low efficiency, low safety, and waste of resources when dealing with complex and ever-changing modern industrial scenarios. First, low task reusability makes it difficult to abstract general operating logic into reusable modules. Developing new task programs requires repeated investment of significant manpower and time, resulting not only in low development efficiency but also in the need for individual adjustments to customized programs during later maintenance, leading to a serious waste of human and time resources. Second, poor scenario adaptability prevents dynamic adjustments to execution based on real-time environmental changes (such as workpiece position shifts), task urgency, or the robot's own cognitive state (such as changes in equipment load). First, the robot's response to complex or unexpected situations is slow and its operation is improper, directly affecting production efficiency. Second, the human-machine interaction efficiency is low. In human-machine collaboration or remote monitoring scenarios, the robot cannot intuitively and efficiently convey its intentions and decision-making processes to the operators, increasing the difficulty of understanding and communication costs for the operators, and indirectly reducing collaboration efficiency. Third, the robot's conflict handling capability is weak. When there are multiple concurrent instructions (such as receiving assembly and handling instructions at the same time) or abnormal events (such as equipment failure warnings), the lack of effective priority judgment and conflict resolution mechanisms may lead to safety risks such as equipment collisions and task interruptions, threatening production safety and task continuity. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, and device for scheduling and executing task intents in dynamic scenarios to address the aforementioned technical problems.

[0005] A task intent scheduling and execution method for dynamic scenarios, the method comprising: An intent block library is constructed, which stores multiple atomic and reusable intent blocks. Each intent block predefines a function type and binds a three-dimensional scheduling attribute. The function type corresponds to the basic operation or advanced behavior of the industrial robot. The three-dimensional scheduling attribute includes capability level, task abstraction level, and main input / output modality. The system receives the target task of the industrial robot and invokes a task model that matches the target task. The task model includes a directed acyclic graph where nodes are intent blocks, and the connection relationships between nodes represent the execution dependencies of each intent block. The scenario type of the industrial robot is determined based on the real-time sensor data, and a corresponding scheduling strategy is selected based on the scenario type. Based on the scheduling strategy and the three-dimensional scheduling attributes of the intent blocks of each node in the directed acyclic graph, the intent blocks to be executed in this round are selected. The intent blocks to be executed are conflict-resolved, and an execution queue is generated based on the conflict resolution results. The industrial robot calls the corresponding hardware and software resources to execute the corresponding functions according to the three-dimensional scheduling attributes of the intent blocks in the execution queue.

[0006] A task intent scheduling and execution device for dynamic scenarios, the device comprising: The intent block definition module is used to build an intent block library, which stores multiple atomic and reusable intent blocks. Each intent block predefines a function type and binds three-dimensional scheduling attributes. The function type corresponds to the basic operation or advanced behavior of the industrial robot. The three-dimensional scheduling attributes include capability level, task abstraction level, and main input / output modality. The task model invocation module is used to receive the target task of the industrial robot and invoke the task model that matches the target task; the task model includes a directed acyclic graph with nodes as intent blocks, and the connection relationship between nodes represents the execution dependency relationship of each intent block; The task intent scheduling module is used to determine the scene type of the industrial robot based on the real-time sensor data, select the corresponding scheduling strategy based on the scene type, and filter the intent blocks to be executed in the current round based on the scheduling strategy and the three-dimensional scheduling attributes of the intent blocks of each node in the directed acyclic graph. The task intent execution module is used to resolve conflicts in intent blocks to be executed and generate an execution queue based on the conflict resolution results. The industrial robot calls the corresponding hardware and software resources to execute the corresponding functions according to the three-dimensional scheduling attributes of the intent blocks in the execution queue.

[0007] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps: An intent block library is constructed, which stores multiple atomic and reusable intent blocks. Each intent block predefines a function type and binds a three-dimensional scheduling attribute. The function type corresponds to the basic operation or advanced behavior of the industrial robot. The three-dimensional scheduling attribute includes capability level, task abstraction level, and main input / output modality. The system receives the target task of the industrial robot and invokes a task model that matches the target task. The task model includes a directed acyclic graph where nodes are intent blocks, and the connection relationships between nodes represent the execution dependencies of each intent block. The scenario type of the industrial robot is determined based on the real-time sensor data, and a corresponding scheduling strategy is selected based on the scenario type. Based on the scheduling strategy and the three-dimensional scheduling attributes of the intent blocks of each node in the directed acyclic graph, the intent blocks to be executed in this round are selected. The intent blocks to be executed are conflict-resolved, and an execution queue is generated based on the conflict resolution results. The industrial robot calls the corresponding hardware and software resources to execute the corresponding functions according to the three-dimensional scheduling attributes of the intent blocks in the execution queue.

[0008] The aforementioned task intent scheduling and execution method, apparatus, and equipment for dynamic scenarios, by constructing an intent block library that stores atomic and reusable intent blocks, enables highly modular reuse of general industrial robot operation logic, significantly shortening the development cycle and cost of new tasks, and effectively improving task reusability and development efficiency. After receiving the target task, it calls a directed acyclic graph containing intent blocks, clarifies the execution logic based on the dependencies between nodes, determines the scene type and matches the scheduling strategy based on real-time sensor information, and filters the intent blocks to be executed based on the strategy and three-dimensional attributes. The robot can break free from the limitations of fixed execution processes, dynamically adapt to different scene requirements, and significantly enhance scene adaptability and robustness. By resolving conflicts in the intent blocks to be executed to generate an execution queue, and calling corresponding hardware and software resources according to the three-dimensional scheduling attributes of the intent blocks, it can avoid system chaos caused by execution conflicts, ensure stable system operation, and achieve precise resource matching to support efficient task execution. This invention enables industrial robots to efficiently, safely, and autonomously complete target tasks in dynamic industrial scenarios. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating a task intent scheduling and execution method for dynamic scenarios in one embodiment. Figure 2 This is a flowchart illustrating a task intent scheduling and execution method for dynamic scenarios in a specific embodiment. Figure 3 This is a schematic diagram of the DAG construction process in one embodiment; Figure 4 This is a schematic diagram of the dynamic scheduling strategy selection process in one embodiment; Figure 5 This is a structural block diagram of a task intent scheduling and execution system for dynamic scenarios in one embodiment; Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0011] In one embodiment, such as Figure 1 As shown, a task intent scheduling and execution method for dynamic scenarios is provided, including the following steps: Step 102: Construct an intent block library. The intent block library stores multiple atomic and reusable intent blocks. Each intent block has a predefined function type and is bound to a three-dimensional scheduling attribute.

[0012] An intent block is a unit that decomposes the intention of an industrial robot to perform a task into the smallest, independent, and reusable functional units. Each intent block encapsulates specific operational logic and possesses atomicity (indivisible) and three-dimensional attribute binding. The three-dimensional scheduling attributes include capability level (SAT level), task abstraction level, and main input / output modality. The function type is a predefined core capability label for the intent block, corresponding to the industrial robot's basic operations or advanced behaviors. Basic operations include fundamental actions such as gripping and movement, while advanced behaviors correspond to complex logic (such as path planning).

[0013] It is understandable that by using atomic encapsulation and reusable design, common operational logic can be abstracted into standardized modules. There is no need to write code from scratch for each new task. Instead, existing intent blocks can be called or combined, which greatly shortens the development cycle. The binding of three-dimensional scheduling attributes provides precise scheduling tags for subsequent dynamic scene adaptation and resource calling.

[0014] Step 104: Receive the target task of the industrial robot and invoke the task model that matches the target task. The task model includes a directed acyclic graph where nodes are intent blocks, and the connections between nodes represent the execution dependencies of each intent block.

[0015] The task model is a predefined execution scheme for the target task. In essence, it is a structured execution logic composed of intent blocks. This embodiment uses a directed acyclic graph (DAG). The nodes of the DAG correspond to intent blocks, and the directed edges represent the execution dependencies between intent blocks (such as moving above the workpiece requires execution before grabbing the workpiece). There are no circular dependencies in the graph, ensuring that the execution logic is consistent.

[0016] By invoking the task model based on intent block DAG, complex tasks are decomposed into traceable and manageable combinations of intent blocks. The task structure is visualized through DAG, which facilitates debugging and modification. At the same time, the clear definition of dependencies ensures the orderly execution of tasks, providing a clear logical framework for subsequent dynamic scheduling and indirectly improving the stability of the system.

[0017] Step 106: Determine the scene type of the industrial robot based on the real-time sensor data, select the corresponding scheduling strategy based on the scene type, and filter out the intent blocks to be executed in this round based on the scheduling strategy and the three-dimensional scheduling attributes of the intent blocks of each node in the directed acyclic graph.

[0018] Sensor-collected information refers to environmental data (such as obstacle distance and workpiece position) and robot status data acquired by the robot through devices such as LiDAR and cameras. Scene type is a classification of the robot's environment and task status based on real-time information, including high-safety scenes (such as close-range human-robot interaction and high-speed movement), low-cognitive-load scenes (such as needing to explain decisions to the operator), and routine scenes. The scheduling strategy is a set of intent block selection rules tailored to different scene types, determining how to select the intent block to be executed from the DAG.

[0019] It is understandable that step 106 enables the robot to no longer rely on a fixed execution process, but to dynamically adjust the block selection logic according to environmental changes (such as the sudden appearance of obstacles). At the same time, it combines the three-dimensional scheduling attributes of the intent block for filtering to ensure that the selected intent block is accurately matched with the scene requirements, thereby improving the robustness of the system and the efficiency of human-computer interaction.

[0020] Step 108: Conflict resolution is performed on the intent blocks to be executed. An execution queue is generated based on the conflict resolution results. The industrial robot calls the corresponding hardware and software resources to execute the corresponding functions based on the three-dimensional scheduling attributes of the intent blocks in the execution queue.

[0021] Conflict resolution is a mechanism that handles conflicts when multiple pending intent blocks are logically mutually exclusive (e.g., emergency stop versus continued operation) or resource-competitive (e.g., simultaneously calling the same sensor), by using priority rules. The execution queue is a list of intent blocks arranged in execution order after conflict resolution, representing the sequence of instructions actually executed by the robot.

[0022] In the aforementioned task intent scheduling and execution method for dynamic scenarios, by constructing an intent block library that stores atomic and reusable intent blocks, the general industrial robot operation logic can achieve highly modular reuse, significantly shortening the development cycle and cost of new tasks, and effectively improving task reusability and development efficiency. After receiving the target task, the method calls a directed acyclic graph containing intent blocks, clarifies the execution logic based on the dependencies between nodes, determines the scenario type and matches the scheduling strategy based on real-time sensor information, and filters the intent blocks to be executed based on the strategy and three-dimensional attributes. The robot can break free from the limitations of fixed execution processes, dynamically adapt to different scenario requirements, and significantly enhance scenario adaptability and robustness. By resolving conflicts of the intent blocks to be executed to generate an execution queue, and calling corresponding software and hardware resources according to the three-dimensional scheduling attributes of the intent blocks, the method can avoid system chaos caused by execution conflicts, ensure stable system operation, achieve precise resource matching, support efficient task execution, and provide structured support for the robot to autonomously complete the target task, promoting its more orderly autonomous execution.

[0023] In one embodiment, the method further includes: after the intent blocks in the queue to be executed are completed, determining whether all node intent blocks in the directed acyclic graph have been executed; if not, filtering the next round of intent blocks to be executed, until all node intent blocks in the directed acyclic graph have been executed, and determining that the target task is completed.

[0024] In one embodiment, the method further includes: if there is no task model matching the target task, selecting or dynamically generating intent blocks that match the requirements of the target task from the intent block library, binding three-dimensional scheduling attributes to the selected or generated intent blocks, constructing a directed acyclic graph based on the execution dependencies of each intent block, and using the directed acyclic graph as a new task model corresponding to the target task.

[0025] In one embodiment, the three-dimensional scheduling attributes include a capability level comprising a perception layer, an action layer, and a reasoning and prediction layer; a task abstraction level comprising a target layer, a planning layer, and an operation layer; and main input / output modalities comprising visual, auditory, and tactile senses.

[0026] In this embodiment, the design of the three-dimensional scheduling attribute makes subsequent dynamic scheduling (such as filtering redundant modal blocks in high-security scenarios and calling inference layer blocks in low-awareness scenarios) no longer rely on fuzzy experience, but based on explicit attributes to achieve accurate block selection, thereby realizing scenario adaptation and efficient human-computer interaction.

[0027] Specifically, during the atomization and encapsulation of intent blocks, the intent of the industrial robot to perform a task is decomposed into a series of minimal, independent, and reusable intent blocks. Each intent block is bound to three dimensions of attributes: SAT Level: Represents the capability level of the intent block, including the perception layer (S1: basic sensor input), the action layer (A1: performing specific actions), and the reasoning / prediction layer (T1: advanced logical reasoning, decision-making, or prediction).

[0028] Task abstraction layer: Represents the level of abstraction of intent blocks within a task, including strategic, tactical, and operational layers.

[0029] Main Input / Output Modality: Represents the input / output modality that the intent block primarily relies on, including visual, auditory, and tactile.

[0030] For example, an intent block for detecting obstacles ahead might have the following attributes: SAT level = perception (S1), task abstraction level = operation, and main modality = vision. An intent block for planning obstacle avoidance paths, on the other hand, might have the following attributes: SAT level = reasoning (T1), task abstraction level = tactics, and main modality = vision.

[0031] In one embodiment, determining the scene type of the industrial robot based on real-time sensor data includes: integrating environmental information and robot body state information collected by sensors into a scene feature vector; matching the scene feature vector with a preset scene rule base to determine whether the scene type is a high-safety scene, a low-cognitive-load scene, or a normal scene.

[0032] In this embodiment, the scheduling engine monitors the robot's environment, the urgency of the task, and other relevant status information in real time. Based on the monitored scene information, it dynamically schedules and executes intent blocks in the DAG, achieving dynamic scene-adaptive scheduling during the scheduling process. High-safety scenario adaptation: In scenarios requiring extremely high safety (such as when a robotic arm performs high-speed, precision movements), the scheduling engine will prioritize and activate intent blocks with redundant modalities or high reliability attributes. For example, when a potential hazard is detected, it will not only rely on visual recognition but also simultaneously activate vibration signals fed back by tactile sensors, forming redundant perception and feedback of visual light + tactile vibration to improve safety margins.

[0033] Low cognitive load scenario adaptation: When the robot needs to explain its decision-making logic to the operator or a higher-level system, or when the scenario complexity allows, the scheduling engine will enable intent blocks at a higher SAT level (such as reasoning / prediction at the T1 level) and can explain the decision-making process through auditory modalities (such as voice output), thereby reducing the operator's cognitive load.

[0034] Typical scenario adaptation: If the match is a typical scenario, the scheduling engine will strictly schedule intent blocks according to the topological order of the task DAG.

[0035] In one embodiment, selecting a corresponding scheduling strategy based on the scenario type includes: if the scenario type is a high-security scenario, the scheduling strategy is to prioritize screening intent blocks with redundant input / output modalities or high reliability characteristics; if the scenario type is a low cognitive load scenario, the scheduling strategy is to enable intent blocks at the inference prediction layer capability level and prepare multimodal interpretation information; if the scenario type is a regular scenario, the scheduling strategy is to screen intent blocks according to the topological order of the directed acyclic graph.

[0036] In this embodiment, redundant input / output modalities refer to the simultaneous activation of multiple perception / execution channels in critical scenarios. Cognitive load refers to the amount of mental effort required for an operator to understand the robot's decision-making process. This embodiment directly maps scenario types to scheduling strategies: in high-safety scenarios, redundant modalities or high-reliability intent blocks are prioritized to improve safety margins at the strategy level and reduce safety risks in scenarios such as high-speed, precision motion; in low-cognitive-load scenarios, inference and prediction layer intent blocks are proactively activated and multimodal interpretations are prepared to address the pain point of low human-computer interaction efficiency and reduce the mental effort (cognitive load) required for operators to understand decisions; in regular scenarios, execution according to the DAG order ensures the efficiency of regular tasks, achieving a balance between safety, interaction, and efficiency in different scenarios, making the scheduling strategy more targeted.

[0037] In one embodiment, the intention blocks to be executed in the current round are selected based on the scheduling strategy and the three-dimensional scheduling attributes of each node intention block in the directed acyclic graph. This includes: selecting node intention blocks from the directed acyclic graph where all preceding dependent intention blocks have been executed, as candidate intention blocks; and performing attribute matching on the candidate intention blocks according to the requirements of the current scheduling strategy for the three-dimensional scheduling attributes, and determining the candidate intention blocks that pass the attribute matching as the intention blocks to be executed in the current round.

[0038] In one embodiment, conflict resolution of intent blocks to be executed includes: presetting the security weight priority of each intent block; when mutual exclusion or resource contention is detected in the execution logic of an intent block, retaining the high-priority intent block and suppressing or delaying the low-priority intent block.

[0039] In this embodiment, the security weight priority of each intent block is preset. When mutual exclusion or resource contention is detected in the execution logic of the intent block: if it is an irreconcilable security conflict, the high-priority intent block is retained and the low-priority intent block is suppressed; if it is a conflict that can be adjusted in time, the high-priority intent block is retained and the low-priority intent block is delayed until the high-priority intent block is completed and then restored.

[0040] It is understandable that by pre-setting priorities (such as security intent blocks taking precedence over regular task blocks), urgent or critical instructions are ensured to be executed first, reducing security risks. At the same time, the corresponding resources are called based on the three-dimensional scheduling attributes of intent blocks (such as visual modal intent blocks calling cameras), ensuring precise matching of hardware and software resources during execution and improving the accuracy of task execution. The structured queue of tasks to be executed further ensures the orderliness of task execution, ultimately achieving a balance between system security and efficiency.

[0041] In one specific embodiment, such as Figure 2 As shown, a flowchart illustrating the task intent scheduling and execution process for dynamic scenarios is provided. It includes: (S101) When the system receives a new industrial task instruction, such as "move workpiece P001 from warehouse A to assembly station B", the task planner first parses the instruction.

[0042] (S102) The scheduling engine first determines whether the task is new. For example, it checks whether a corresponding task model already exists in the system by comparing the task ID or the hash value of the task content. If it already exists, it jumps directly to step S106.

[0043] (S103) If it is a new task, the "Intent Block Wrapper" is invoked to select or dynamically generate the minimum intent block unit required to constitute the task from the "Intent Block Library". The intent block library pre-stores modular intent blocks for various basic operations (such as gripping, moving, and visual detection) and advanced behaviors (such as path planning and anomaly detection).

[0044] (S104) Each selected intent block is bound to its three-dimensional attributes through the "Intent Block Attribute Definition" module. For example: "Detect workpiece P001" intent block: SAT level = perception (S1), task abstraction level = operation, main modality = vision.

[0045] "Planning an obstacle avoidance path" intent block: SAT level = reasoning (T1), task abstraction level = tactics, main modality = vision.

[0046] (S105) The logical dependencies of all intent blocks are input into the "DAG building module" to generate the "Task DAG" for this task. This DAG represents the execution order and dependencies of the intent blocks. For detailed steps of this construction process, see [link to documentation]. Figure 3 The details are as follows: (S105a) Define and input dependencies: Each intent block comes with a list of dependencies. For example, the dependency list for the "Grab Workpiece" intent block is ["Move over workpiece", "Open gripper"]. The DAG building module receives all intent blocks and their dependency lists as input.

[0047] (S105b) Creating Nodes and Directed Edges: The module first creates each intent block as a node in a graph. Then, it adds directed edges between the nodes based on a list of dependencies, for example, from the "Move Over Workpiece" node to the "Grab Workpiece" node.

[0048] (S105c) Cyclic Dependency Validation: After adding all edges, the module executes a topological sorting algorithm (such as Kahn's algorithm or a depth-first search-based algorithm) to check for cyclic dependencies in the graph. If a cycle is detected (e.g., A depends on B, and B depends on A), the system throws an exception indicating a task logic error, thus ensuring that the generated DAG is a directed acyclic graph. After successful validation, the generated DAG is fixed in the form of an adjacency list or adjacency matrix data structure.

[0049] (S106) If the task already exists, load its pre-built task DAG directly from storage.

[0050] (S107) The "Scene Monitor" module in the scheduling engine enters the continuous monitoring loop. It continuously collects information about the robot's surrounding environment (such as obstacles and personnel positions) through various environmental perception sensors (such as LiDAR and cameras) and combines this information with the robot's own status (such as the current speed of the robotic arm end effector and the remaining time of the task).

[0051] (S108) Based on this real-time information and pre-defined scenario rules, the scheduling engine assesses the current scenario type. For detailed steps of this assessment and decision-making process, please refer to [link to relevant documentation]. Figure 4 The details are as follows: (S108a) Scene feature vectorization: The scene monitor integrates the collected multi-source information, such as the distance to the nearest obstacle, the confidence level of the appearance of people, the speed of the robotic arm, etc., into a real-time scene feature vector V_scene.

[0052] (S108b) Scene Rule Base Matching: The scheduling engine matches V_scene against its internally maintained scene rule base. This rule base defines the mapping relationship between different scene types and feature vectors, for example: Rule 1 (High-Safety Scenario): IF (Distance to nearest obstacle < 0.5 m OR Person confidence > 0.8) AND (Robotic arm speed > 0.5 m / s) THEN Scenario type = "High-Safety Scenario".

[0053] Rule 2 (Low Cognitive Load / Explanation Requirement Scenario): IF (Number of task execution failures > 2 OR Operator issues "Explanation" instruction) THEN Scenario Type = "Low Cognitive Load / Explanation Requirement Scenario".

[0054] Rule 3 (Regular Scenario): ELSE Scenario Type = "Regular Scenario".

[0055] (S109-S112) The scheduling engine selects and executes the corresponding scheduling policy from the policy library based on the matched scenario type: (S110) High security strategy: If the match is "high security scenario", the scheduling engine will prioritize scheduling intent blocks with redundant modalities (such as enabling visual and tactile obstacle avoidance at the same time) or high reliability (such as reducing speed and increasing checkpoints).

[0056] (S111) Explanation Priority Strategy: If the match is "low cognitive load / explanation requirement scenario", the scheduling engine will enable a higher-order T1 layer intent block (such as "decision logic analysis") and prepare to output explanation information through voice or other means.

[0057] (S112) Standard order strategy: If the match is "normal scenario", the scheduling engine will strictly schedule intent blocks according to the topological order of the task DAG.

[0058] (S113-S114) After the scheduling engine determines the intent blocks that need to be executed in this round, the "conflict resolver" checks whether there are potential conflicts between these intent blocks and with new external instructions (such as the operator pressing the emergency stop button).

[0059] (S115) Once a conflict is detected, such as when the "emergency stop" intent block and the "continue to execute task" intent block are triggered at the same time, the conflict resolution mechanism will sort the conflicting intent blocks according to the preset "safety weight priority" and dynamically suppress or delay the execution of the low-priority intent block by inserting a high-priority instruction into the "execution queue" of the scheduling engine or removing / suspending a low-priority instruction, so as to ensure that the high-priority intent block (such as emergency stop) can be executed first, thereby resolving the conflict.

[0060] (S116) Intent blocks that have not experienced a conflict or whose conflicts have been successfully resolved are identified as the final execution unit for this round and are added by the system to a first-in-first-out (FIFO) "queue to be executed" to await the call of the robot body.

[0061] (S117) The robot's execution module retrieves instructions from the "execution queue" and coordinates the invocation of corresponding hardware and software resources according to the three-dimensional attributes (SAT level, task abstraction layer, main modality) bound to the intent block. For example, if the main modality of the intent block is "vision", it will drive the vision sensor and image processing unit; if it is the "action" layer, it will drive the servo motor to perform physical actions.

[0062] (S118) During execution, the system continuously monitors the execution status of the intent block. Once a completion signal is received from the underlying hardware or a preset completion condition is met (e.g., the sensor detects that the gripper has closed in place), the current intent block is determined to be completed.

[0063] (S119) After each intent block is completed, the scheduling engine checks whether all nodes (intent blocks) in the task DAG have been marked as "executed".

[0064] (S120) If all intent blocks have been executed, the system determines that the entire industrial task is completed and the process ends. If there are still unexecuted intent blocks in the task DAG, the process will return to step S107 and enter the next round of "scene monitoring and scheduling" loop to deal with possible changes in the environment and task status, and continue to execute subsequent tasks.

[0065] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0066] In one embodiment, such as Figure 5 As shown, a task intent scheduling and execution system for dynamic scenarios is provided, including an intent block management and generation module, a task and scheduling control module, and a robot body module. Among them: The intent block management and generation module is responsible for building atomic and reusable intent blocks. It includes an intent block library, an intent block wrapper, and an intent block attribute definition unit. The intent block library is used to store basic intent blocks, the intent block wrapper is used to encapsulate intent blocks, and the intent block attribute definition unit is used to pre-bind three-dimensional scheduling attributes to intent blocks and generate intent block instances. The task and scheduling control module is responsible for receiving target tasks and dynamically scheduling intent blocks based on real-time scenarios. It includes a task planner, a DAG construction module, a scene monitor, a scheduling engine, and a conflict resolver. The task planner receives task instructions and provides task logic. The DAG construction module generates a directed acyclic graph of tasks containing the dependencies of intent blocks. The scene monitor receives real-time data from environmental perception sensors to determine the scene type. The scheduling engine combines the scene type and the task DAG to filter the intent blocks to be executed. The conflict resolver resolves conflicts in the intent blocks to be executed, thereby generating an execution queue. The robot body module is responsible for performing functions and realizing human-machine interaction. It includes a communication module, a perception module, and an execution module. The communication module is used to process interpretation instructions and feedback instructions, and can also output multimodal interpretation information to the operator or external system. The perception module is used to execute perception instructions and provide feedback on perception results. The execution module is used to output physical actions.

[0067] In one embodiment, a task intent scheduling and execution device for dynamic scenarios is provided, comprising: The intent block definition module is used to build an intent block library. The intent block library stores multiple atomic and reusable intent blocks. Each intent block predefines a function type and binds three-dimensional scheduling attributes. The function type corresponds to the basic operation or advanced behavior of the industrial robot. The three-dimensional scheduling attributes include capability level, task abstraction level, and main input / output modality. The task model invocation module is used to receive the target task of the industrial robot and invoke the task model that matches the target task. The task model includes a directed acyclic graph with nodes as intent blocks, and the connection relationship between nodes represents the execution dependency relationship of each intent block. The task intent scheduling module is used to determine the scene type of the industrial robot based on the real-time sensor data, select the corresponding scheduling strategy based on the scene type, and filter out the intent blocks to be executed in this round based on the scheduling strategy and the three-dimensional scheduling attributes of the intent blocks of each node in the directed acyclic graph. The task intent execution module is used to resolve conflicts in intent blocks to be executed. Based on the conflict resolution results, an execution queue is generated. The industrial robot calls the corresponding hardware and software resources to execute the corresponding functions according to the three-dimensional scheduling attributes of the intent blocks in the execution queue.

[0068] Specific limitations regarding the task intent scheduling and execution device for dynamic scenarios can be found in the limitations of the task intent scheduling and execution method for dynamic scenarios described above, and will not be repeated here. Each module in the aforementioned task intent scheduling and execution device for dynamic scenarios can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0069] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a task intent scheduling execution method oriented towards dynamic scenarios. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0070] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0071] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the method described above.

[0072] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0073] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A task intent scheduling and execution method for dynamic scenarios, characterized in that, The method includes: An intent block library is constructed, which stores multiple atomic and reusable intent blocks. Each intent block predefines a function type and binds a three-dimensional scheduling attribute. The function type corresponds to the basic operation or advanced behavior of the industrial robot. The three-dimensional scheduling attribute includes capability level, task abstraction level, and main input / output modality. The system receives the target task of the industrial robot and invokes a task model that matches the target task. The task model includes a directed acyclic graph where nodes are intent blocks, and the connection relationships between nodes represent the execution dependencies of each intent block. The scenario type of the industrial robot is determined based on the real-time sensor data, and a corresponding scheduling strategy is selected based on the scenario type. Based on the scheduling strategy and the three-dimensional scheduling attributes of the intent blocks of each node in the directed acyclic graph, the intent blocks to be executed in this round are selected. The intent blocks to be executed are conflict-resolved, and an execution queue is generated based on the conflict resolution results. The industrial robot calls the corresponding hardware and software resources to execute the corresponding functions according to the three-dimensional scheduling attributes of the intent blocks in the execution queue. In the three-dimensional scheduling attributes, the capability levels include a perception layer, an action layer, and a reasoning and prediction layer; the task abstraction levels include a target layer, a planning layer, and an operation layer; and the main input-output modalities include visual, auditory, and tactile.

2. The method according to claim 1, characterized in that, The method further includes: After the intent blocks in the queue to be executed are completed, it is determined whether all node intent blocks in the directed acyclic graph have been executed. If not completed, the next round of intent blocks to be executed is selected, and the target task is determined to be completed when all intent blocks of all nodes in the directed acyclic graph have been executed.

3. The method according to claim 1, characterized in that, The method further includes: If no task model matches the target task, an intent block matching the target task requirements is selected from the intent block library or dynamically generated. Three-dimensional scheduling attributes are bound to the selected or generated intent blocks, and a directed acyclic graph is constructed based on the execution dependencies of each intent block. This directed acyclic graph is then used as the new task model corresponding to the target task.

4. The method according to claim 1, characterized in that, The process of determining the scene type of the industrial robot based on real-time sensor data includes: The environmental information and robot body state information collected by the sensors are integrated into a scene feature vector; The scene feature vector is matched with a preset scene rule base to determine whether the scene type is a high-security scene, a low-cognitive-load scene, or a normal scene.

5. The method according to claim 1, characterized in that, The step of selecting the corresponding scheduling strategy based on the scenario type includes: If the scenario type is a high-security scenario, the scheduling strategy is to prioritize filtering intent blocks with redundant input / output modalities or high reliability characteristics. If the scenario type is a low cognitive load scenario, the scheduling strategy is to enable the intent block at the inference and prediction layer capability level and prepare multimodal interpretation information. If the scenario type is a regular scenario, the scheduling strategy is to filter intent blocks according to the topological order of the directed acyclic graph.

6. The method according to claim 1, characterized in that, Based on the scheduling strategy and the three-dimensional scheduling attributes of the intent blocks of each node in the directed acyclic graph, the intent blocks to be executed in this round are selected as follows: Select node intent blocks from the directed acyclic graph where all preceding dependent intent blocks have been executed, and use them as candidate intent blocks. Based on the requirements of the current scheduling strategy for the three-dimensional scheduling attributes, the candidate intent blocks are matched for attributes, and the candidate intent blocks that pass the attribute matching are determined as the intent blocks to be executed in this round.

7. The method according to claim 1, characterized in that, The conflict resolution of the intent block to be executed includes: The security weight priority of each intent block is preset. When mutual exclusion or resource contention is detected in the execution logic of an intent block, the high-priority intent block is retained, while the low-priority intent block is suppressed or delayed.

8. A task intent scheduling and execution device for dynamic scenarios, used to implement the task intent scheduling and execution method for dynamic scenarios as described in any one of claims 1 to 7, characterized in that, The device includes: The intent block definition module is used to build an intent block library, which stores multiple atomic and reusable intent blocks. Each intent block predefines a function type and binds three-dimensional scheduling attributes. The function type corresponds to the basic operation or advanced behavior of the industrial robot. The three-dimensional scheduling attributes include capability level, task abstraction level, and main input / output modality. The task model invocation module is used to receive the target task of the industrial robot and invoke the task model that matches the target task; the task model includes a directed acyclic graph with nodes as intent blocks, and the connection relationship between nodes represents the execution dependency relationship of each intent block; The task intent scheduling module is used to determine the scene type of the industrial robot based on the real-time sensor data, select the corresponding scheduling strategy based on the scene type, and filter the intent blocks to be executed in the current round based on the scheduling strategy and the three-dimensional scheduling attributes of the intent blocks of each node in the directed acyclic graph. The task intent execution module is used to resolve conflicts in intent blocks to be executed and generate an execution queue based on the conflict resolution results. The industrial robot calls the corresponding hardware and software resources to execute the corresponding functions according to the three-dimensional scheduling attributes of the intent blocks in the execution queue.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

Images