This work considers a robot equipped with an anthropomorphic hand and aims at providing it with efficient autonomous in-hand manipulation skills. While fine in-hand action planning algorithms have interesting state-of-the-art solutions, we built a competitive high-level control layer to plan the complete in-hand manipulation activity. Our solution generates a sequence of subgoals from an initial to a final configuration provided by the task, thus decomposing in-hand manipulation into simple transitions that can be easily planned by the low-level algorithms. We use a Markov decision process (MDP) to generate the sequence, taking into account the object influence and the desired final subgoal. We use a simple state representation for the sugoals: canonical grasp types from a taxonomy, enabling fast and on-line computation. The transitions between grasp types are modelled as probabilities of success. The simple formulation of the sequence leaves the complete configurations and transitions to be planned by the low-level layer, which can ask for a different subgoal path if required. The MDP can generate the appropriate behaviour if the in-hand action skills of the robot are known. They can be learnt by self-exploration of the robot if possible. Otherwise, the behaviour can be directly learnt from human demonstration. We boost the learning process using an empirical guess of the transition probabilities and an active learning algorithm. We implemented our solution on a real platform. The planning of in-hand manipulation relies on the grasp sequence generated which probability of success is used as an insight of the task achievability for the initial grasp choice.

Grasping, the skill to hold objects and tools while doing in-hand manipulation, still is in many cases an unsolvable problem for robotics, but a natural act for humans. An efficient grasping requires not only human-like robotic hands with articulated fingers but also tactile, force, and kinesthetic sensors for the precise control of the forces and motions exerted during the manipulation. As a fully autonomous robotic dexterous manipulation is too difficult to develop for changing and unstructured environments, an alternative approach is to combine the low-level robot computer control with the higher-level perception and task planning abilities of a human operator equipped with an adequate human-computer interface (HCI). This thesis presents theoretical and experimental contributions to the development of an upgraded haptic-enabled anthropomorphic Ring Ada dexterous robotic hand and a biology-inspired synergistic real-time control system for teleoperated grasping of different objects using a CyberTouch HCI data glove. A fuzzy logic controller module was developed to efficiently control the underactuated Ring Ada' robotic hand during grasping. A machine learning classification system was developed to recognize grasped objects. Experiments have convincingly demonstrated that our novel Ring Ada robotic hand equipped with kinematic position sensors and touch sensors is able to efficiently grasp different lightweight objects through teleoperation.

This book introduces a novel model-based dexterous manipulation framework, which, thanks to its precision and versatility, significantly advances the capabilities of robotic hands compared to the previous state of the art. This is achieved by combining a novel grasp state estimation algorithm, the first to integrate information from tactile sensing, proprioception and vision, with an impedance-based in-hand object controller, which enables leading manipulation capabilities, including finger gaiting. The developed concept is implemented on one of the most advanced robotic manipulators, the DLR humanoid robot David, and evaluated in a range of challenging real-world manipulation scenarios and tasks. This book greatly benefits researchers in the field of robotics that study robotic hands and dexterous manipulation topics, as well as developers and engineers working on industrial automation applications involving grippers and robotic manipulators.

Manipulation using dextrous robot hands has been an exciting yet frustrating research topic for the last several years. While significant progress has occurred in the design, construction, and low level control of robotic hands, researchers are up against fundamental problems in developing algorithms for real-time computations in multi-sensory processing and motor control. The aim of this book is to explore parallels in sensorimotor integration in dextrous robot and human hands, addressing the basic question of how the next generation of dextrous hands should evolve. By bringing together experimental psychologists, kinesiologists, computer scientists, electrical engineers, and mechanical engineers, the book covers topics that range from human hand usage in prehension and exploration, to the design and use of robotic sensors and multi-fingered hands, and to control and computational architectures for dextrous hand usage. While the ultimate goal of capturing human hand versatility remains elusive, this book makes an important contribution to the design and control of future dextrous robot hands through a simple underlying message: a topic as complex as dextrous manipulation would best be addressed by collaborative, interdisciplinary research, combining high level and low level views, drawing parallels between human studies and analytic approaches, and integrating sensory data with motor commands. As seen in this text, success has been made through the establishment of such collaborative efforts. The future will hold up to expectations only as researchers become aware of advances in parallel fields and as a common vocabulary emerges from integrated perceptions about manipulation.

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

Our hand research has focused on enhancing the dexterity of robotic hands and understanding the nature of dexterous manipulation. The premise of the research is that incorporating task level understanding into a manipulation system simplifies robot planning and increases autonomy. The study of task level strategies for dexterous manipulation has led to development of several novel techniques for controlling the fingertip forces during manipulation and fingertip motion planning. The insights into increased autonomy have led to the development of a novel technique for teleoperating robot hands. The traditional technique of teleoperating a robot hand is to use a Dataglove or exoskeleton master; there is a direct mapping from the human hand to the robot hand. This approach has several limitations which we have addressed by using a simpler control interface with a joystick or keyboard. Enhancing the robot hand's autonomy allows for simpler control strategies and gives it greater functionality than by traditional means. Control of the hand is shared between the user and the robot. We have developed a prototype teleoperation system using a Utah/MIT hand. Our research will ultimately have application in medicine and industry, for enhancement of prosthetic hands and the development of more complex robotic grippers.

Robotics dexterous manipulation with a number fingers has witnessed a number of developments. Hand dexterous manipulation is a difficult task, this due to a number of axioms, the first is related to complex mechanics of motion, while the last is related to the inter-related topics of intelligence and cognitive sciences needed to achieve tasks. This book is projecting fundamental topics related to dexterous Multifinger robotics manipulation. The book is summarizing large number of scientists and researchers efforts over years to assemble robotics hands with high level of dexterity, mobility, and adequate manipulability needs. Over years, there are large number of hand design attempts, some of which have been designed with high level of dexterities. Actuation and sensing have been cumbersome to achieve full skillful abilities for better manipulation. There have been some recent efforts to make use of artificial muscle type for hands actuation. Hands with artificial muscles for actuations is a new venue within robotics hands. Electroactive polymers have emerged with great promise and enabled the improvement of unique mechanisms that are biologically inspired.