Robot Motion Planning for Accurate Pose Estimation Under Occlusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual programming of robotic movements is tedious, time-consuming, and error-prone, and often fails to ensure accurate object pose estimation, especially in environments with complex geometries or occluded objects, limiting the flexibility and accuracy of robotic tasks.
Innovation Solution
A system that automatically generates motion plans for robots by using in-process sensor observations to ensure accurate object pose estimation, allowing for flexible task execution in challenging environments without the need for manual programming or specific fixtures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual programming is used to dictate robotic movements, then the robot can perform tasks with predetermined paths, but the programming becomes tedious, time-consuming, and error-prone while lacking flexibility for different workcells
Solution Approach 1:
The system enables robots to automatically generate their own motion plans by performing self-measurement and self-planning. The robot uses onboard sensors to measure object poses and automatically generates motion plans without human intervention, making the system self-sufficient and eliminating tedious manual programming
Solution Approach 2:
The system implements a feedback loop where the robot measures object poses using sensors, uses this measurement information to generate motion plans, executes the plans, and then measures again to verify accuracy. This closed-loop feedback enables automatic adaptation and eliminates the need for predetermined manual programming
2Measurement precision
If manual planning is created for one workcell, then the robot can execute tasks with specific accuracy requirements, but the plan cannot be reused for other workcells with different physical properties and layouts
Solution Approach 1:
The system transitions from static predetermined paths to dynamic adaptive planning. Motion plans are generated in real-time based on actual object poses measured by sensors, allowing the system to adapt to different workcells, object positions, and geometries while maintaining measurement precision requirements
Solution Approach 2:
The system changes the approach from fixed parameters to variable parameters. Instead of using predetermined paths with fixed coordinates, the system generates motion plans with parameters derived from actual object measurements, enabling the same system to adapt to different workcells and object configurations while maintaining accuracy
3Ease of operation
If the robot observes objects from certain viewpoints to estimate pose, then the pose estimation can be obtained, but the accuracy deteriorates significantly when objects are occluded or have complex geometries
Solution Approach 1:
The system performs preliminary measurement actions by positioning the robot and sensor to capture object poses before generating motion plans. This preliminary measurement provides accurate initial pose information even for occluded or complex objects, enabling subsequent planning to account for these challenging geometries
Solution Approach 2:
The system introduces an intermediate measurement step between object placement and task execution. Sensors capture object poses from multiple viewpoints and the system processes this intermediate measurement data to generate accurate motion plans, serving as a mediator that bridges the gap between occluded objects and accurate pose estimation
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for planning robotic movements to perform a given task while satisfying object pose estimation accuracy requirements. One of the methods includes generating a plurality of candidate measurement configurations for measuring an object to be manipulated by a robot; determining respective measurement accuracies for the plurality of candidate measurement configurations; determining a measurement accuracy landscape for the object including defining a high measurement accuracy region based on the respective measurement accuracies for the plurality of candidate measurement configurations; and generating a motion plan for manipulating the object in the robotic process that moves the robot, a sensor, or both, through the high measurement accuracy region when performing pose estimation for the object.