Robotic Grasp Servoing With Deep Learning Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robots face challenges in successfully grasping objects due to variability in environmental conditions and inaccurate robotic actuation, leading to inefficiencies in end effector motion and grasp attempts.
Innovation Solution
A deep neural network, such as a convolutional neural network (CNN), is trained to predict the probability of successful grasping based on candidate motion data and environmental images, enabling iterative updates to motion control commands and improving servoing performance of grasping end effectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional robotic actuation methods are used for grasping, then the system structure remains simple, but the reliability of successful grasp is low due to environmental variability and inaccurate actuation
Solution Approach 1:
The patent implements a feedback mechanism where the deep neural network continuously predicts grasp success probability based on sensor data and motion commands, then uses this prediction to iteratively update motion control commands. This closed-loop feedback system enables the robot to adapt to environmental variations and compensate for actuation inaccuracies, significantly improving grasp reliability without requiring complex mechanical modifications
Solution Approach 2:
The patent replaces traditional mechanical control systems with a data-driven deep neural network approach. Instead of relying on complex mechanical sensing and actuation systems, the invention uses software-based prediction and iterative command optimization to achieve robust grasping, substituting mechanical complexity with computational intelligence
2Reliability
If iterative updates to motion control commands are implemented, then servoing performance is improved, but the loss of time for computation increases
Solution Approach 1:
The patent applies partial action by performing iterative updates only when necessary for achieving successful grasp, rather than continuously optimizing all parameters. The system generates a sequence of motion commands with incremental improvements, stopping iterations when the predicted success probability meets a threshold or when time constraints are approached, thus balancing servoing performance with computational efficiency
Data Source
AI summary
Deep machine learning methods and apparatus related to manipulation of an object by an end effector of a robot. Some implementations relate to training a deep neural network to predict a measure that candidate motion data for an end effector of a robot will result in a successful grasp of one or more objects by the end effector. Some implementations are directed to utilization of the trained deep neural network to servo a grasping end effector of a robot to achieve a successful grasp of an object by the grasping end effector. For example, the trained deep neural network may be utilized in the iterative updating of motion control commands for one or more actuators of a robot that control the pose of a grasping end effector of the robot, and to determine when to generate grasping control commands to effectuate an attempted grasp by the grasping end effector.