Closed-Loop Activity Program Correction Using Simulator Feedback

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing closed-loop activity program generation systems lack the ability to automatically correct errors during execution, leading to incomplete or failed tasks once an instruction fails.

Innovation Solution

A method and system for generating closed-loop human activity programs that utilize a planner to create an initial program, verify actions using a verification code, execute in a simulator, and employ a corrector to modify the program based on simulator feedback for error correction, incorporating error correction experience to optimize the program dynamically.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If open-loop activity program generation is used, then generation speed is improved, but error correction capability deteriorates

Engineering Contradiction:

Improvegeneration speedVSAvoiderror correction capability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent implements a closed-loop generation system where the simulator provides feedback on execution status to the generator. The generator receives execution results and uses this feedback to automatically correct errors in the activity program, transforming the open-loop process into a closed-loop system that continuously improves reliability through feedback-driven correction.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent performs preliminary verification of activity programs before execution by simulating them in advance. The simulator pre-executes the program to detect potential errors, and the generator uses these preliminary results to correct issues before actual execution, preventing errors from propagating through the entire program.

Inventive Principle:

Principle #10Preliminary action

2Reliability

If closed-loop activity program generation is used, then error correction capability is improved, but generation complexity increases

Engineering Contradiction:

Improveerror correction capabilityVSAvoidgeneration complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent merges the generator and simulator into an integrated closed-loop system where both components work together seamlessly. The generator creates activity programs, the simulator executes and verifies them, and the generator automatically corrects errors based on simulator feedback, combining multiple functions into a unified system that manages complexity through integration.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The system performs self-correction of errors automatically without external intervention. The generator uses execution feedback from the simulator to identify and correct errors in the activity program autonomously, making the system self-sufficient and reducing the need for manual complexity management.

Inventive Principle:

Principle #25Self-service

3Manufacturing precision

If manual design of activity programs is used, then program quality is improved, but time cost increases

Engineering Contradiction:

Improveprogram qualityVSAvoidtime cost

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent replaces manual design processes with an automated generator that creates activity programs based on input parameters. The generator uses algorithms and rules to automatically produce programs, substituting the mechanical manual design process with an automated computational system that maintains quality while significantly reducing time investment.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system copies and adapts successful activity program patterns from a database of verified programs. The generator retrieves similar programs from the database and automatically adapts them to current requirements, using copying of proven patterns to maintain high quality while accelerating generation speed compared to creating programs from scratch.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS20260111304A1Generation method and system for closed-loop human activity program

Publication Date: 2026.04.23 SHENZHEN UNIV

AI summary

Generation method and system for closed-loop human activity program. The method involves inputting initial scene information and activity description, generating initial activity program through planner, checking verification code of corresponding action of the activity program, and sending current action to simulator for execution when the current action is executable, or modifying the activity program using verification code when the current action is unexecutable; checking execution status of the activity program, and if the corresponding action is successfully executed in simulator, obtaining post-execution status returned from the simulator, or if the corresponding action is executed unsuccessfully, modifying the activity program again using corrector according to error information returned from the simulator; and outputting the final modified activity program. The method can dynamically correct the activity program during the execution process, and greatly improve the enforceability of the activity program.