Robot Finger Deformable Sensors for Tactile Pose and Force Sensing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Robotic end effectors lack the sensitivity to determine the shape and proper grasping force of objects, often leading to damage or dropping of targets due to insufficient tactile feedback.

Innovation Solution

Incorporating deformable sensors with deformable membranes and internal sensors on robot fingers, allowing for detection of contact, geometry, and pose of objects, enabling a sense of touch similar to humans.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional pressure sensors are used in end effectors, then contact detection is provided, but the sensitivity and information about object geometry and proper grasping force are insufficient

Engineering Contradiction:
Improvetactile feedback precisionVSAvoidobject geometry and pose information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The end effector is divided into multiple fingers, each equipped with its own deformable sensor. This segmentation allows independent detection of contact forces and geometry at each finger, providing distributed tactile feedback that captures both contact magnitude and object shape information simultaneously.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Deformable membranes are integrated into the fingers of the end effector. These flexible membranes deform in response to contact with objects, and their deformation patterns are captured by internal sensors to determine both contact force and object geometry, thereby providing comprehensive tactile feedback without information loss.

Inventive Principle:
Principle #30Flexible shells and thin films

2Strength

If rigid end effectors are used, then structural strength is maintained, but the ability to adapt to different object shapes and apply appropriate grasping force is reduced

Engineering Contradiction:
Improveend effector structural strengthVSAvoidgrasping adaptability
Core Design Contradiction:
StrengthVSAdaptability or versatility

Solution Approach 1:

The end effector transitions from a rigid structure to a dynamic system where fingers can deform independently through integrated deformable sensors. This dynamic capability allows the end effector to adapt its shape to match various object geometries while maintaining structural integrity through controlled deformation, thereby achieving both strength and versatility.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The stiffness and deformability parameters of the end effector fingers are adjusted through the integration of deformable membranes with controlled mechanical properties. This allows the system to optimize the balance between structural strength and grasping adaptability by selecting materials and designs that provide appropriate compliance for different manipulation tasks.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If deformable sensors are integrated into robot fingers, then tactile feedback and object geometry detection are improved, but device complexity increases

Engineering Contradiction:
Improvecontact and geometry detection precisionVSAvoidsensor integration complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

Multiple sensing functions (contact detection, geometry detection, pose estimation) are merged into a single deformable sensor system. The deformable membrane serves as both the structural element and the sensing element, with internal sensors capturing deformation patterns that simultaneously provide information about contact force, object shape, and orientation, thereby reducing overall system complexity.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The deformable sensor system performs multiple functions simultaneously: it detects contact forces, determines object geometry, and estimates object pose. This multi-functionality eliminates the need for separate sensors for each measurement type, reducing device complexity while maintaining high measurement precision across all tactile feedback parameters.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enables secure grasping and manipulation of objects by providing accurate tactile feedback, preventing damage and ensuring proper handling through precise force control.

Implementation Method 1

each finger (112, 118) includes a deformable sensor (200, 500) that detects contact and a geometry and/or pose of the object (150). In a particular embodiment, the deformable sensor (200, 500) includes a deformable membrane (220, 520) and an internal sensor (230).

Methodology Applied
Scientific EffectDeformation: Deformation

Data Source

PatentEP3879374B1Robot arm assemblies including fingers having deformable sensors
Publication Date: 2024.03.27 TOYOTA JIDOSHA KK
  • EP3879374B1 patent drawingFigure 1
  • EP3879374B1 patent drawingFigure 2A~2B
  • EP3879374B1 patent drawingFigure 3~4

AI summary

A robot arm assembly for detecting a pose and force associated with an object is provided. The robot arm assembly includes an end effector having a plurality of fingers, and a deformable sensor provided on each finger. The deformable sensor includes a housing, a deformable membrane coupled to the housing, an enclosure filled with a medium, and an internal sensor disposed within the housing having a field of view directed through the medium and toward an internal surface of the deformable membrane. A processor is configured to receive an output from each internal sensor, the output including a contact region of the deformable membrane as a result of contact with the object. The processor determines an amount of displacement of the contact region based on the output from each internal sensor, and determines the pose and the force associated with the object based on the amount of displacement.