Robot Dynamic Manipulation: Perception of Deformable Objects and Nonprehensile Manipulation Control (Springer Tracts in Advanced Robotics, 144)

✍ Scribed by Bruno Siciliano (editor), Fabio Ruggiero (editor)

• Publisher: Springer
• Year: 2022
• Tongue: English
• Leaves: 263
• Category: Library

✦ Synopsis

This book collects the main results of the Advanced Grant project RoDyMan funded by the European Research Council. As a final demonstrator of the project, a pizza-maker robot was realized. This represents a perfect example of understanding the robot challenge, considering every inexperienced person's difficulty preparing a pizza. Through RoDyMan, the opportunity was to merge all the acquired competencies in advancing the state of the art in nonprehensile dynamic manipulation, which is the most complex manipulation task, considering deformable objects. This volume is intended to present Ph.D. students and postgraduates working on deformable object perception and robot manipulation control the results achieved within RoDyMan and propose cause for reflection of future developments. The RoDyMan project culminating with this book is meant as a tribute to Naples, the hosting city of the project, an avant-garde city in robotics technology, automation, gastronomy, and art culture.

✦ Table of Contents

Foreword
Preface
Contents
Contributors
Acronyms
Common Symbols
Perception
Deformation Modelling for a Physics-Based Perception System
1 Brief Introduction
2 Background on Deformation Models
2.1 Mesh-Based Approaches
2.2 Mesh-Free Approaches
2.3 Hybrid Approaches
3 FEM Elastic Model
3.1 Finite Element Modelling
3.2 Linear Elasticity
3.3 The Corotational Approach
4 Fracture Model
4.1 Fracture Detection
4.2 Fracture Propagation and Remeshing
4.3 Comments on the Results
5 Interaction Model
5.1 Collision Detection
5.2 Collision Response
6 Discussion and Conclusion
References
Non-rigid Tracking Using RGB-D Data
1 Brief Introduction
2 Related Work and Motivations
2.1 Registration Using Implicit Physical Modeling
2.2 Registration Using Explicit Physical Modeling
2.3 Handling Topological Changes, Fractures and Cuts
2.4 Multiple Objects Registration
2.5 Motivations and Contributions
2.6 Overview of the System
3 Visual Segmentation
3.1 Grabcut Segmentation
3.2 Temporal Coherence and Real-Time Issues
4 Segmented and Sampled Point Cloud
5 Rigid Registration
6 Point Cloud Matching for Non-rigid Registration
6.1 Nearest Neighbor Correspondences
6.2 Computation of External Forces
6.3 Weighting Forces Using Contours
7 Solver
7.1 Experimental Results
7.2 Results for Tracking on Synthetic Data
7.3 Results on Real Data
8 Fractures
8.1 A Pure Physics-Based Approach
8.2 Experimental Results
8.3 Comments on the Results
9 Multiple Objects
9.1 Preliminary Parallel Visual Segmentation
9.2 Parallel Rigid Pose Estimation
9.3 Parallel Point Cloud Matching for Non-rigid Registration
9.4 Resolution
9.5 Experimental Results
9.6 Comments on the Results
10 Application to Robotic Manipulation
10.1 Trajectory Planning and Control
10.2 Experimental Set-Up
10.3 Comments on the Results
11 Application to Elasticity Parameter Estimation and Contact Force Estimation
11.1 Related Work
11.2 Elasticity Parameter Estimation
11.3 Contact Force Estimation
11.4 Experimental Results
11.5 Comments on the Results
12 Discussion and Conclusion
References
Smoothed Particle Hydrodynamics-Based Viscous Deformable Object Modelling
1 Brief Introduction
2 Theoretical Background about Navier-Stokes's Theorem and SPH
2.1 Navier-Stokes' Theorem for Continuous Materials
2.2 SPH Formulation
3 Viscosity Property and Various Viscosity Methods for SPH
4 Other Components for the SPH-based Modelling
4.1 Kernel Functions
4.2 Incompressible Fluid
5 Simulations
5.1 Accuracy and Time Analysis
5.2 Couette Flow Experiment
5.3 Poiseuille Plane Flow Experiment
5.4 Comparison with Conventional Viscosity Methods
5.5 Additional Simulations
6 Discussion and Conclusion
References
Perception and Motion Planning for Unknotting/untangling of Ropes of Finite Thickness
1 Brief Introduction
2 Problem Statement
3 Preliminaries and Associated State-of-the-art
3.1 Rope Models
4 Proposed Algorithm
4.1 Rope Model
4.2 Self-collision Avoidance Model
5 Motion Planner
6 Discussion and Conclusion
References
Nonprehensile Manipulation Planning and Control
Pizza-Peel Handling Through a Sliding Nonprehensile Manipulation Primitive
1 Brief Introduction
2 State of the Art
3 Pizza-Peel Manipulation Task
3.1 Dynamic Model
3.2 Controller Design and Stability Analysis
3.3 Numerical Simulation
4 Discussion and Conclusion
References
Holonomic Rolling Nonprehensile Manipulation Primitive
1 Brief Introduction
2 Dynamic Model of Nonprehensile Holonomic Rolling Manipulation Systems
3 Input-State Feedback Linearisation
3.1 Hypotheses on the Shapes and Input-State Linearisation
3.2 Case Studies
4 Passivity-Based Approach
4.1 Background on Passivity-Based Control
4.2 Control Design for Nonprehensile Systems
4.3 Case Studies
5 Discussion and Conclusion
References
Nonholonomic Rolling Nonprehensile Manipulation Primitive
1 Brief Introduction
2 The Hula-Hoop Problem
2.1 Contact Kinematics
2.2 Dynamic Model
2.3 Controller Design and Stability Analysis
2.4 Numerical Simulation
3 Ballbot
3.1 Lagrangian Dynamics of the Ballbot
3.2 Passivity Based Control Design
3.3 Numerical Examples
4 Discussion and Conclusion
References
A Coordinate-Free Framework for Robotic Pizza Tossing and Catching
1 Introduction
2 Grasp Constraints
3 Kinematics
4 Dynamics
4.1 Variable Inertia Rigid Body Orientation Dynamics
4.2 Rigid Body Translational Dynamics
4.3 RoDyMan Arm Manipulator Dynamics
4.4 Object and Manipulator Combined Dynamics
5 Trajectory Generation
5.1 Theory
5.2 Generating Hand Frame Trajectories
6 Control Law
6.1 Tossing
6.2 Catching
7 Simulation
8 Discussion and Conclusion
References
Planning Framework for Robotic Pizza Dough Stretching with a Rolling Pin
1 Brief Introduction
2 Related Research
3 Framework for a Pizza Dough Stretching Behaviour
4 Pizza Dough Recognition
4.1 Image Processing for Sensor Data
4.2 Description for a Status of a Pizza Dough
5 Construction of a Planner for Pizza Dough Stretching
5.1 Cost Value Function
5.2 Actions for a Deformable Object
5.3 Transition Originated from an Action
5.4 LUT Method
6 Path Generation for a Rolling Pin
7 Inverse Kinematics for the RoDyMan Robot
8 Simulations
8.1 Modelling of a Deformable Object
8.2 Pizza Dough Transition Look-up-table
9 Discussion and Conclusion
References