Robot Contact Trajectory Planning With Relaxed Virtual Force Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic motion planning techniques struggle to efficiently plan contact-interaction trajectories due to the non-smooth dynamics introduced by physical contacts, which preclude the use of gradient-based solvers and require computationally impractical predefined contact schedules or tuning of contact models.
Innovation Solution
A relaxed contact model that uses virtual forces to represent contact dynamics, allowing smooth optimization techniques by iteratively adjusting penalty values on virtual stiffness, minimizing virtual forces, and ensuring physically accurate trajectories without sensitivity to initialization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a predefined contact schedule is used to plan contact-interaction trajectories, then the motion planning can be performed with gradient-based solvers, but the approach becomes computationally impractical for complex motion planning
Solution Approach 1:
The patent introduces a virtual contact force model as an intermediary between the robot and environment. This virtual model allows gradient-based optimization to reason about contacts without requiring predefined contact schedules, enabling efficient computation for complex motions while maintaining physical accuracy through iterative refinement of contact forces.
Solution Approach 2:
The patent changes the parameterization of contact dynamics by representing contact forces as continuous variables that can be optimized alongside state and input trajectories. This parameter change enables the use of gradient-based solvers for contact-rich complex motions without requiring predefined contact schedules, resolving the contradiction between computational efficiency and planning complexity.
2Productivity
If contact-implicit trajectory optimization with smooth contact model is used, then motion planning of contact-rich complex motions is enabled without predefined contact schedule, but physical inaccuracies occur due to penetrations and contact forces at a distance
Solution Approach 1:
The patent makes the contact model dynamic by iteratively adjusting the stiffness parameter during optimization. The virtual contact force model transitions from a purely computational tool to a physically accurate representation through successive refinement, where the contact stiffness is adapted to match real contact dynamics while maintaining the benefits of smooth optimization.
Solution Approach 2:
The patent implements feedback by using the optimization results to refine the contact model parameters. The virtual contact forces computed during optimization provide feedback that is used to adjust the contact stiffness and other parameters in subsequent iterations, progressively improving physical accuracy while maintaining computational efficiency.
3Manufacturing precision
If contact model parameters are tuned to accurately approximate real contact dynamics, then physical accuracy is improved, but the tuning process becomes difficult and requires re-tuning when task or robot changes
Solution Approach 1:
The patent enables the contact model to self-adjust by automatically tuning its parameters during the optimization process. The virtual contact force model uses the optimization feedback to self-calibrate the stiffness and other parameters, eliminating the need for manual tuning and making the system adaptable to different tasks and robots without additional complexity.
Data Source
Figure 1A
Figure 1B
Figure 1C
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.