Robot Force Control Tuning for Robust Assembly Under Pose Uncertainty
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic assembly systems face challenges in tuning force control parameters for general assembly tasks due to difficulties in detecting and correcting complex misalignments, requiring manual trial and error, which is time-consuming, expensive, and potentially dangerous, and existing simulation-based systems are limited in applicability and require significant human expertise.
Innovation Solution
A method using numerical optimization in a simulation environment to evaluate different combinations of force control parameters, incorporating random pose uncertainties, to autonomously tune parameters, ensuring robustness and safety by simulating various assembly scenarios and updating parameter distributions iteratively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual tuning of force control parameters is performed on real robotic systems, then the assembly task can be made robust to positioning uncertainties, but the process becomes time-consuming, expensive, and potentially dangerous
Solution Approach 1:
The patent creates a virtual copy of the robotic assembly system through high-fidelity simulation. The simulation model replicates the physical system's dynamics, including force control behavior and contact mechanics, allowing parameter tuning to be performed in the virtual environment rather than on the actual robot. This copying approach eliminates the time loss and safety risks associated with manual tuning on real systems while maintaining the reliability benefits of force control parameter optimization.
Solution Approach 2:
The patent performs force control parameter tuning in advance through automated simulation-based optimization before deploying the parameters to the real robotic system. The system pre-computes optimal parameters by running multiple simulations with varied initial conditions and using optimization algorithms to converge on robust parameter sets. This preliminary action in the virtual environment eliminates the need for time-consuming and potentially dangerous iterative tuning on the actual robot during operation.
2Extent of automation
If existing simulation-based parameter tuning systems are used, then some automation is achieved, but they require significant human expertise and are limited to specific assembly task types
Solution Approach 1:
The patent develops a universal simulation-based parameter tuning framework that can handle multiple types of assembly tasks through a common force control simulation model. The system uses a generalized contact dynamics model and optimization approach that adapts to different assembly scenarios (peg-in-hole, planar part installation, etc.) without requiring task-specific customization. This universality enables automated parameter tuning across diverse assembly tasks while eliminating the need for significant human expertise in guiding the selection of force control parameters.
Solution Approach 2:
The patent implements an autonomous parameter tuning system where the simulation framework automatically selects force control parameters, runs simulations, evaluates performance, and iteratively optimizes parameters without human intervention. The system self-corrects and self-improves through automated optimization algorithms that analyze simulation results and adjust parameters accordingly. This self-service capability eliminates the need for human expertise in the tuning process while achieving high levels of automation applicable to general assembly tasks.
3Ease of manufacture
If force control parameters are tuned without considering pose uncertainties in simulation, then the tuning process is simpler, but the results do not transfer well to real-world operation
Solution Approach 1:
The patent incorporates pose uncertainties into the simulation-based parameter tuning process in advance, before deploying parameters to the real robotic system. The simulation model includes random variations in part poses, fixture errors, and grasping uncertainties to realistically represent real-world conditions. By cushioning against these uncertainties beforehand through robust optimization, the system ensures that tuned parameters remain effective when transferred to actual operation, eliminating the need for simpler but less reliable tuning approaches that ignore uncertainties.
Data Source
AI summary
A method for tuning the force control parameters for a general robotic assembly operation. The method uses numerical optimization to evaluate different combinations of the parameters for a robot force controller in a simulation environment that is built based on a real-world robotic setup. This method performs autonomous tuning for assembly tasks based on closed loop force control simulation, where random samples from a distribution of force control parameter values are evaluated, and the optimization routine iteratively redefines the parameter distribution to find optimal values of the parameters. Each simulated assembly is evaluated using multiple simulations including random part positioning uncertainties. The performance of each simulated assembly is evaluated by the average of the simulation results, thus ensuring that the selected control parameters will perform well in most possible conditions. Once the parameters have been optimized, they are applied to real robots to perform the actual assembly operation.