Robot Trajectory Retargeting for Real-Time Mobility Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robotic devices face challenges in navigating complex environments with difficult terrain and unforeseen obstacles while efficiently managing computational resources, as existing systems struggle to adapt movements in real-time effectively.
Innovation Solution
A robotic system that pre-computes a library of 'template behaviors' such as running, walking, or jumping, and adapts these movements in real-time by combining them with environmental and kinematic information to determine 'retargeted trajectories', using computational shortcuts to achieve fluid motion and adapt to changing conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the robot performs real-time nonlinear trajectory optimizations with densely sampled data, then the movement precision and adaptability are improved, but the computational resources are excessively consumed
Solution Approach 1:
The patent pre-computes and stores optimal trajectories offline before robot operation. These pre-computed trajectories are stored in a database and can be quickly retrieved during real-time operation, eliminating the need for computationally intensive real-time optimization while maintaining high movement precision
Solution Approach 2:
The patent segments the trajectory computation process into two distinct phases: an offline phase where complex nonlinear optimizations are performed and trajectories are stored, and an online phase where the robot simply retrieves and executes pre-computed trajectories. This segmentation separates the computationally intensive tasks from real-time operation
2Manufacturing precision
If the robot uses detailed template behaviors with long time horizons and densely sampled data, then the movement quality is improved, but the real-time computational speed is reduced
Solution Approach 1:
The patent performs detailed trajectory computations and stores them as template behaviors offline before operation. During real-time operation, the robot retrieves these pre-computed templates and makes only minor adjustments, achieving high movement quality without real-time computational burden
Solution Approach 2:
The patent creates copies of optimal trajectories as reusable template behaviors that can be quickly retrieved and executed. These template copies contain the results of complex offline optimizations and can be applied repeatedly without re-computation
3Adaptability or versatility
If the robot makes frequent real-time adjustments to account for environmental deviations, then the adaptability is improved, but the computational intensity increases
Solution Approach 1:
The patent implements feedback by comparing the robot's actual state with the pre-computed template trajectory and making corrective adjustments. The system uses sensed deviations from the planned trajectory to generate correction commands, maintaining adaptability without requiring full re-optimization
Solution Approach 2:
Instead of performing complete trajectory re-optimization in real-time, the patent applies partial corrections to the pre-computed trajectories. Only the portions of the trajectory affected by environmental deviations are adjusted, reducing computational intensity while maintaining adaptability
Data Source
AI summary
Systems and methods for determining movement of a robot about an environment are provided. A computing system of the robot (i) receives information including a navigation target for the robot and a kinematic state of the robot; (ii) determines, based on the information and a trajectory target for the robot, a retargeted trajectory for the robot; (iii) determines, based on the retargeted trajectory, a centroidal trajectory for the robot and a kinematic trajectory for the robot consistent with the centroidal trajectory; and (iv) determines, based on the centroidal trajectory and the kinematic trajectory, a set of vectors having a vector for each of one or more joints of the robot.