The human hand's unique biomechanical structure and neuromuscular control combine to produce amazing dexterous capabilities in a way that is still not fully understood. The Anatomically Correct Testbed (ACT) hand is a robotic system that is designed as a physical simulation of the human hand, and can help us examine and potentially uncover the roles of biomechanics and neural control in achieving dexterity. In this dissertation, I utilize the ACT hand and other robotic systems to explore the underlying sources of human hand dexterity, with the goal of understanding the fundamental differences between robotic and human hands in terms of (i) mechanical joint/tendon structure and (ii) control strategies. To begin, I develop comprehensive mechanical models that describe the musculoskeletal and tendon mechanics of the fingers and thumb of the human hand. Then, I work to isolate the contributions of biomechanical structure and neuromuscular control toward human dexterity. I have developed and implemented control strategies for achieving fine object manipulation first with the robotic hand of a space humanoid, Robonaut 2, and then with the ACT hand. I examined the unique control challenges, including uncontrollable joints and the requirement of accurate internal models, that arise due to the human hand's complex musculotendon structure and the potential advantages offered by the human hand's design, such as passive joint coupling to facilitate grasp shape adaptation and force production capabilities that are ideally suited for common manipulation tasks. Finally, inspired by the neuromuscular control strategies of the human hand, I have developed a novel hierarchical control strategy for the ACT hand and experimentally demonstrated improved grasp stability and manipulation capabilities compared to conventional robotic control laws. Through an in-depth exploration of human hand biomechanics and neuromuscular control, theoretical control analysis of robotic and human hands, and experimental demonstration of fine object manipulation, this work uncovers crucial insights into the sources of human hand dexterity that have the potential to drive innovative design and control strategies and bring robotic and prosthetic hands closer to human levels of dexterity.

Human Inspired Dexterity in Robotic Manipulation provides up-to-date research and information on how to imitate humans and realize robotic manipulation. Approaches from both software and hardware viewpoints are shown, with sections discussing, and highlighting, case studies that demonstrate how human manipulation techniques or skills can be transferred to robotic manipulation. From the hardware viewpoint, the book discusses important human hand structures that are key for robotic hand design and how they should be embedded for dexterous manipulation. This book is ideal for the research communities in robotics, mechatronics and automation. Investigates current research direction in robotic manipulation Shows how human manipulation techniques and skills can be transferred to robotic manipulation Identifies key human hand structures for robotic hand design and how they should be embedded in the robotic hand for dexterous manipulation

“The Human Hand as an Inspiration for Robot Hand Development” presents an edited collection of authoritative contributions in the area of robot hands. The results described in the volume are expected to lead to more robust, dependable, and inexpensive distributed systems such as those endowed with complex and advanced sensing, actuation, computation, and communication capabilities. The twenty-four chapters discuss the field of robotic grasping and manipulation viewed in light of the human hand’s capabilities and push the state-of-the-art in robot hand design and control. Topics discussed include human hand biomechanics, neural control, sensory feedback and perception, and robotic grasp and manipulation. This book will be useful for researchers from diverse areas such as robotics, biomechanics, neuroscience, and anthropologists.

The human hand and its dexterity in grasping and manipulating objects are some of the hallmarks of the human species. For years, anatomic and biomechanical studies have deepened the understanding of the human hand’s functioning and, in parallel, the robotics community has been working on the design of robotic hands capable of manipulating objects with a performance similar to that of the human hand. However, although many researchers have partially studied various aspects, to date there has been no comprehensive characterization of the human hand’s function for grasping and manipulation of everyday life objects. This monograph explores the hypothesis that the confluence of both scientific fields, the biomechanical study of the human hand and the analysis of robotic manipulation of objects, would greatly benefit and advance both disciplines through simulation. Therefore, in this book, the current knowledge of robotics and biomechanics guides the design and implementation of a simulation framework focused on manipulation interactions that allows the study of the grasp through simulation. As a result, a valuable framework for the study of the grasp, with relevant applications in several fields such as robotics, biomechanics, ergonomics, rehabilitation and medicine, has been made available to these communities.

This is a cornerstone publication in robotic grasping. The authors have developed an internationally recognized expertise in this area. Additionally, they designed and built several prototypes which attracted the attention of the scientific community. The purpose of this book is to summarize years of research and to present, in an attractive format, the expertise developed by the authors on a new technology for grasping which has achieved great success both in theory and in practice.

Achieving human-like dexterity for common daily tasks with a robotic hand is a challenging task for reinforcement learning. To combat the high dimensionality of the state-action space, we propose pre-training policies on low-level vision tasks. Our system, called DexV2A, first trains a neural network on structural vision tasks like edge detection, center point estimation and surface normal estimation to embed useful visual features into the network weights. When this network is transferred to a dexterous manipulation policy, it offers an advantageous initialization for task learning. We conduct experiments on four diverse manipulation tasks with a 30-DoF dexterous robotic hand in simulation. We show that for the tasks of opening, closing, pouring, and stirring, DexV2A improves policy learning over policies trained without any visual pre-training. Our experimental results demonstrate the effectiveness of our approach, and emphasizes the potency of visual pre-training over learning via direct experience

The 9-volume set LNAI 14267-14275 constitutes the proceedings of the 16th International Conference on Intelligent Robotics and Applications, ICIRA 2023, which took place in Hangzhou, China, during July 5–7, 2023. The 413 papers included in these proceedings were carefully reviewed and selected from 630 submissions. They were organized in topical sections as follows: Part I: Human-Centric Technologies for Seamless Human-Robot Collaboration; Multimodal Collaborative Perception and Fusion; Intelligent Robot Perception in Unknown Environments; Vision-Based Human Robot Interaction and Application. Part II: Vision-Based Human Robot Interaction and Application; Reliable AI on Machine Human Reactions; Wearable Sensors and Robots; Wearable Robots for Assistance, Augmentation and Rehabilitation of Human Movements; Perception and Manipulation of Dexterous Hand for Humanoid Robot. Part III: Perception and Manipulation of Dexterous Hand for Humanoid Robot; Medical Imaging for Biomedical Robotics; Advanced Underwater Robot Technologies; Innovative Design and Performance Evaluation of Robot Mechanisms; Evaluation of Wearable Robots for Assistance and Rehabilitation; 3D Printing Soft Robots. Part IV: 3D Printing Soft Robots; Dielectric Elastomer Actuators for Soft Robotics; Human-like Locomotion and Manipulation; Pattern Recognition and Machine Learning for Smart Robots. Part V: Pattern Recognition and Machine Learning for Smart Robots; Robotic Tactile Sensation, Perception, and Applications; Advanced Sensing and Control Technology for Human-Robot Interaction; Knowledge-Based Robot Decision-Making and Manipulation; Design and Control of Legged Robots. Part VI: Design and Control of Legged Robots; Robots in Tunnelling and Underground Space; Robotic Machining of Complex Components; Clinically Oriented Design in Robotic Surgery and Rehabilitation; Visual and Visual-Tactile Perception for Robotics. Part VII: Visual and Visual-Tactile Perception for Robotics; Perception, Interaction, and Control of Wearable Robots; Marine Robotics and Applications; Multi-Robot Systems for Real World Applications; Physical and Neurological Human-Robot Interaction. Part VIII: Physical and Neurological Human-Robot Interaction; Advanced Motion Control Technologies for Mobile Robots; Intelligent Inspection Robotics; Robotics in Sustainable Manufacturing for Carbon Neutrality; Innovative Design and Performance Evaluation of Robot Mechanisms. Part IX: Innovative Design and Performance Evaluation of Robot Mechanisms; Cutting-Edge Research in Robotics.