Abstract:Purpose The structural adaptive ability of the soft robot is fully demonstrated in the grasping task of the soft hand. A soft hand can easily realize the envelope operation of the object without planning. With the continuous development of robot applications, researchers are no longer satisfied with the ability of the soft hand to grasp. The purpose of this paper is to perceive the object's shape while grasping to provide a decision-making basis for more intelligent robot applications. Design/methodology/approach This paper proposes a dual-signal comparison method to obtain the fingertip position. The dual signal includes the displacement calculated by the static model without considering the external load change and the displacement calculated by the bending sensor. The dual-signal comparison method can use the obvious change trend difference between the above two signals in the hover and contact states to identify the touch position. The authors make the soft hand scan around the object through touch operation to detect the object's shape, and the tracks of every touch fingertip position can envelop the object's shape. Findings The experimental results show that the dual-signal comparison method can accurately identify the contact moment of soft fingers. This detection method makes the soft hand develop the shape detection ability. The soft hand in the experiment can perceive squares, circles and a few other complex shapes. Originality/value The dual-signal comparison method proposed in this paper can detect a touch action by using the signal change trend when the working condition suddenly changes with the rough robotic model and sensing, thus improving the utilization value of the measured signal. The problems of large model errors and inaccurate sensors also negatively impact the use of other soft robots. It is generally difficult to achieve good results by directly using these models and sensors with the thinking of rigid robot analysis. The dual-signal comparison method in this paper can provide some reference for this aspect.

Haptic Perception Of Unknown Object By Robot Hand: Exploration Strategy And Recognition Approach

An Anthropomorphic Exploration Strategy of Unknown Object Based on Haptic Information of Dexterous Robot Hand

Unknown Objects Grasp Planning Using Force/Torque Measurements

A Haptic Shared Control Architecture for Tracking of a Moving Object.

Surface material recognition through haptic exploration using an intelligent contact sensing finger

Sliding Touch-based Exploration for Modeling Unknown Object Shape with Multi-fingered Hands

Recognition of the Three-Dimensional Shape of Objects Grasped for Pesa Multi-Fingered Robot Hand

Simple Kinesthetic Haptics for Object Recognition

Multifunctional Tactile Sensors for Object Recognition

A multisensory-feedback tactile glove with dense coverage of sensing arrays for object recognition

A Multisensory Tactile System for Robotic Hands to Recognize Objects

Recent progress on tactile object recognition

A Multifunctional Tactile Sensory System for Robotic Intelligent Identification and Manipulation Perception

Development of Robotic Hand Tactile Sensing System for Distributed Contact Force Sensing in Robotic Dexterous Multimodal Grasping

Enhanced robotic tactile perception with spatiotemporal sensing and logical reasoning for robust object recognition

3-D Object Reconstruction Using Tactile Sensing for Multi-fingered Hand

Adaptive visual–tactile fusion recognition for robotic operation of multi-material system

Shape perception of soft hand based on dual-signal comparison contact detection

A soft robotic system imitating the multimodal sensory mechanism of human fingers for intelligent grasping and recognition

Open-Set Object Recognition Using Mechanical Properties During Interaction

Model Recovery Of Unknown Objects From Discrete Tactile Points