Mobile Robot Navigation with Dynamic DoF Trajectory Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile manipulator platforms face challenges in autonomous navigation due to high-dimensional mobility spaces and decoupled architectures, which complicate autonomy and mechatronic design, limiting their ability to efficiently reposition arms, manipulate objects, and optimize end-effector positioning.
Innovation Solution
An autonomous navigation system for mobile robots with a Hybrid Ground Autonomous Manipulator Vehicle (HGAMV) that dynamically adjusts the state search space, integrating sensors, a replanning supervisor, pose planner, and trajectory planner to optimize the robot's degrees of freedom and minimize a cost function, allowing for seamless transitions between mobile and fixed platforms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a decoupled architecture is used for mobile manipulator platforms, then the mobile base and robotic arm can be designed separately, but the autonomous navigation and manipulation capabilities are compromised due to sequential handling of movements
Solution Approach 1:
The patent merges the mobile base and robotic arm into a unified kinematic chain, treating them as a single integrated system rather than separate units. This allows simultaneous optimization of navigation and manipulation tasks, enabling the robot to perform autonomous operations that require coordinated movement of both base and arm, such as moving while manipulating objects or repositioning the arm to avoid obstacles.
2Adaptability or versatility
If all degrees of freedom are always enabled for the robot, then the robot has maximum flexibility to reach any target, but the computational complexity and energy consumption increase significantly
Solution Approach 1:
The patent implements dynamic adjustment of degrees of freedom based on the current task and robot state. The system selectively enables or disables specific DoFs during operation, allowing the robot to use only the necessary degrees of freedom for each particular task. This reduces computational complexity by lowering the dimensionality of the search space and decreases energy consumption by activating only the required actuators, while still maintaining the ability to reach any target when needed.
3Reliability
If the search state space includes all dimensions (position, velocity, acceleration) for a high-degree-of-freedom robot, then complete motion control is achieved, but the computational speed decreases due to the high dimensionality
Solution Approach 1:
The patent segments the search space by identifying and separating the relevant dimensions from the irrelevant ones for each specific task. Instead of uniformly considering all 27 dimensions (9 for position, 9 for velocity, 9 for acceleration) for every motion planning problem, the system divides the search space to include only the necessary dimensions, thereby reducing computational complexity while maintaining complete control over the required motion aspects.
Solution Approach 2:
The patent extracts and removes unnecessary dimensions from the search space based on the current task requirements. By identifying which degrees of freedom and which kinematic dimensions (position, velocity, or acceleration) are actually needed for the current operation, the system extracts only those relevant parameters from the full state space, significantly reducing the computational burden while preserving complete control over the essential motion parameters.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An apparatus, including: an interface operable to receive sensor data and generate a map representation of an environment of a robot; processing circuitry operable to: plan a sequence of states to direct the robot to a task goal of the robot based on the map representation and a kinematic state of the robot for a plurality of degrees of freedom; determine a time-dependent trajectory of the robot to the task goal based on the sequence of states by dynamically enabling or disabling one or more of the plurality of degrees of freedom; and generate a movement instruction to control movement of the robot based on the time-dependent trajectory.