Robot Contact Trajectory Optimization Using Virtual Contact Forces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robotic motion planning techniques struggle with planning contact-interaction trajectories due to non-smooth dynamics introduced by physical contacts, which precludes the use of gradient-based solvers and requires computationally inefficient testing of multiple trajectories, especially for complex tasks.
Innovation Solution
A relaxed contact model is introduced that uses virtual forces to model contact dynamics, allowing smooth optimization techniques and efficient computation of physically accurate trajectories without the need for parameter tuning, by representing contact interactions through virtual forces that are gradually penalized and diminished during optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If contact-implicit trajectory optimization (CITO) with smooth contact models is used to enable motion planning without predefined contact schedules, then the ability to plan contact-rich complex motions is improved, but physical inaccuracies are introduced due to relaxations (penetrations and contact forces at a distance)
Solution Approach 1:
The patent introduces virtual contact forces as an intermediary mechanism that bridges the gap between smooth optimization requirements and physical contact accuracy. These virtual forces act as a mediator that enables gradient-based optimization while gradually converging to physically accurate contact dynamics, resolving the contradiction between computational smoothness and physical fidelity
Solution Approach 2:
The patent dynamically adjusts the penalty parameter during optimization iterations. Initially, the penalty is set to allow smooth virtual contact forces for convergence, then progressively increased to enforce physical accuracy constraints. This parameter evolution resolves the contradiction by transitioning from adaptability-focused optimization to accuracy-focused refinement
2Manufacturing precision
If penalty parameters for contact model relaxations are tuned to accurately approximate real contact dynamics, then manufacturing precision of the trajectory is improved, but the complexity of parameter tuning increases and requires re-tuning for task or robot changes
Solution Approach 1:
The patent implements self-tuning of penalty parameters through an automated iterative process. The system automatically adjusts penalty values based on optimization progress and contact detection, eliminating the need for manual parameter tuning. This self-service mechanism resolves the contradiction by making the system adaptive without increasing operational complexity
Solution Approach 2:
The patent incorporates feedback loops that monitor optimization convergence and contact dynamics accuracy, automatically adjusting penalty parameters based on observed performance. This feedback-driven adaptation resolves the contradiction by enabling accurate contact modeling without requiring external tuning expertise
3Ease of operation
If virtual forces with distance-based contact forces are used to facilitate optimization convergence, then the ease of operation for trajectory generation is improved, but computational inefficiency arises when testing multiple trajectories
Solution Approach 1:
The patent dynamically adapts the virtual contact force model during optimization, transitioning from distance-based virtual forces that facilitate convergence to physically accurate contact models as optimization progresses. This dynamic adaptation resolves the contradiction by initially prioritizing ease of operation for convergence, then shifting to computational efficiency for refinement
Data Source
AI summary
An apparatus and a method for planning contact-interaction trajectories are provided. The apparatus is a robot that accepts contact interactions between the robot and the environment. The robot stores a dynamic model representing geometric, dynamic, and frictional properties of the robot and the environment, and a relaxed contact model to representing dynamic interactions between the robot and the object via virtual forces. The robot further determines, iteratively until a termination condition is met, a trajectory, associated control commands for controlling the robot, and virtual stiffness values by performing optimization reducing stiffness of the virtual force and minimizing a difference between the target pose of the object and a final pose of the object moved from the initial pose. Further, an actuator moves a robot arm of the robot according to the trajectory and the associated control commands.