Robotic Touch Sensing With Bayesian Fusion for Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current household robots lack the ability to autonomously perform tasks such as cleaning by manipulating objects and cleaning surfaces, as they are limited by inadequate perception and object recognition capabilities, particularly in combining touch and visual sensor data for goal-oriented tasks.
Innovation Solution
A robotic apparatus equipped with touch sensors and distance sensors that use Bayesian inference models to build an environmental map, allowing for the fusion of touch and visual data to perform tasks like cleaning by moving and manipulating objects autonomously.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If household robots use simple sensor systems for navigation, then device complexity is reduced, but perception precision and object recognition capability deteriorate
Solution Approach 1:
The patent combines multiple sensor types (touch sensors, distance sensors, image sensors) into an integrated perception system. The fusion engine merges data from these different sensors to achieve comprehensive environmental understanding and precise object recognition, resolving the contradiction between simple device complexity and high measurement precision.
Solution Approach 2:
The robotic system implements multi-functional sensors that serve multiple purposes. For example, touch sensors not only detect contact but also provide information about object texture, hardness, and shape. This multi-functionality allows the system to achieve high perception precision without proportionally increasing device complexity.
2Extent of automation
If robots use advanced touch perception and sensor fusion for autonomous tasks, then task automation capability is improved, but device complexity increases
Solution Approach 1:
The fusion engine automatically processes and integrates sensor data without requiring complex external control systems. The system performs self-localization, object recognition, and task planning autonomously by processing its own sensor inputs, thereby improving automation capability while managing device complexity through self-service architecture.
Solution Approach 2:
The system implements continuous feedback loops where sensor data from touch, distance, and image sensors is constantly processed to update the robot's understanding of its environment and adjust its actions accordingly. This feedback mechanism enables autonomous task execution while keeping the control architecture manageable through iterative refinement.
3Device complexity
If robots rely solely on visual sensors for object recognition, then device complexity is reduced, but perception accuracy in contact tasks deteriorates
Solution Approach 1:
The patent merges visual sensor data with touch sensor data to create a comprehensive perception system. The fusion engine combines information from both sensor types, allowing the robot to leverage the range and speed of visual sensors while compensating for their lack of tactile feedback through touch sensor inputs, thereby improving contact perception precision without excessive complexity.
Solution Approach 2:
The system adds the tactile dimension to the visual dimension of perception. By incorporating touch sensors that detect contact force, texture, and shape, the system transitions from two-dimensional visual information to multi-dimensional sensory data including tactile properties, significantly improving perception accuracy in contact tasks.
Data Source
AI summary
An apparatus such as a robot capable of performing goal oriented tasks may include one or more touch sensors to receive touch perception feedback on the location of objects and structures within an environment. A fusion engine may be configured to combine touch perception data with other types of sensor data such as data received from an image or distance sensor. The apparatus may combine distance sensor data with touch sensor data using inference models such as Bayesian inference. The touch sensor may be mounted onto an adjustable arm of a robot. The apparatus may use the data it has received from both a touch sensor and distance sensor to build a map of its environment and perform goal oriented tasks such as cleaning or moving objects.