Dextrous object manipulation is a crucial task for the hands of the space humanoid Robonaut 2 (R2), and requires accurate control of fingertip positions and forces. We present a novel Cartesian control for the fingers and thumb of the R2 hand. The controller is designed such that the singularities in the fingers are avoided, and distal joint stiffness is added in the thumb for full controllability. We then present a higher-level object stiffness control law for explicit control of object position, orientation, and grasp forces. The complete algorithm is tested experimentally on the R2 hand, with results demonstrating tracking performance and robustness against disturbances.

The second generation Robonaut hand has many advantages over its predecessor. This mechatronic device is more dexterous and has improved force control and sensing giving it the capability to grasp and actuate a wider range of tools. It can achieve higher peak forces at higher speeds than the original. Developed as part of a partnership between General Motors and NASA, the hand is designed to more closely approximate a human hand. Having a more anthropomorphic design allows the hand to attain a larger set of useful grasps for working with human interfaces. Key to the hand's improved performance is the use of lower friction drive elements and a redistribution of components from the hand to the forearm, permitting more sensing in the fingers and palm where it is most important. The following describes the design, mechanical/electrical integration, and control features of the hand. Lessons learned during the development and initial operations along with planned refinements to make it more effective are presented.

NASA and General Motors have developed the second generation Robonaut, Robonaut 2 or R2, and it is scheduled to arrive on the International Space Station in early 2011 and undergo initial testing by mid-year. This state of the art, dexterous, anthropomorphic robotic torso has significant technical improvements over its predecessor making it a far more valuable tool for astronauts. Upgrades include: increased force sensing, greater range of motion, higher bandwidth, and improved dexterity. R2's integrated mechatronic design results in a more compact and robust distributed control system with a fraction of the wiring of the original Robonaut. Modularity is prevalent throughout the hardware and software along with innovative and layered approaches for sensing and control. The most important aspects of the Robonaut philosophy are clearly present in this latest model's ability to allow comfortable human interaction and in its design to perform significant work using the same hardware and interfaces used by people. The following describes the mechanisms, integrated electronics, control strategies, and user interface that make R2 a promising addition to the Space Station and other environments where humanoid robots can assist people.

A mobile version of the NASA/DARPA Robonaut humanoid recently completed initial autonomy trials working directly with humans in cluttered environments. This compact robot combines the upper body of the Robonaut system with a Segway ™ Robotic Mobility Platform yielding a dexterous, maneuverable humanoid ideal for interacting with human co-workers in a range of environments. This system uses stereovision to locate human teammates and tools and a navigation system that uses laser range and vision data to follow humans while avoiding obstacles. Tactile sensors provide information to grasping algorithms for efficient tool exchanges. The autonomous architecture utilizes these pre-programmed skills to form complex behaviors. The initial behavior demonstrates a robust capability to assist a human by acquiring a tool from a remotely located individual and then following the human in a cluttered environment with the tool for future use.

Engineers at the Johnson Space Center recently combined the upper body of the National Aeronautics and Space Administration (NASA) / Defense Advanced Research Projects Agency (DARPA) Robonaut system with a Robotic Mobility Platform (RMP) to make an extremely mobile humanoid robot designed to interact with human teammates. Virtual Reality gear that immerses a human operator into Robonaut's working environment provides the primary control pathway for remote operations. Human/robot interface challenges are addressed in the control system for teleoperators, console operators and humans working directly with the Robonaut. Multiple control modes are available for controlling the five fingered dexterous robot hands and operator selectable depending on the type of grasp required. A relative positioning system is used to maximize operator comfort during arm and head motions. Foot pedals control the mobility base. Initial tasks that include working with human rated tools, navigating hallways and cutting wires are presented and show the effectiveness of telepresence control for this class of robot.

The Johnson Space Center has developed a new mobile manipulation system with the combination of a Robonaut upper body mounted onto a Segway mobile base. The objective is to study a fluid and coordinated control of dexterous limbs on a mobile robot. The system has been demonstrated interacting with people, tools, and urban interfaces built for humans. Human interactions have included manually exchanging objects with humans, following people, and tracking people with hand held objects such as flashlights. Like other configurations of the Robonaut family, the upper body provides dexterity for using tools such as wire cutters, shovels, space flight gear, and handling flexible tethers and fabrics. The Segway base is a custom version called the Robotic Mobility Platform (RMP) built for DARPA, and provided to NASA for this collaborative effort. The RMP's active balance gives Robonaut a relatively small footprint for its height, allowing it to pass through doors and elevators built for humans, and use wheelchair accessible ramps and lifts. Lessons learned from this development are presented to improve the design of future mobile manipulation systems, and the Segway base provides mobility to Robonaut for Earth based testing.

Tactile data from rugged gloves are providing the foundation for developing autonomous grasping skills for the NASA/DARPA Robonaut, a dexterous humanoid robot. These custom gloves compliment the human like dexterity available in the Robonaut hands. Multiple versions of the gloves are discussed, showing a progression in using advanced materials and construction techniques to enhance sensitivity and overall sensor coverage. The force data provided by the gloves can be used to improve dexterous, tool and power grasping primitives. Experiments with the latest gloves focus on the use of tools, specifically a power drill used to approximate an astronaut's torque tool.

The NASA/DARPA Robonaut system is evolving from a purely teleoperator controlled anthropomorphic robot towards a humanoid system with multiple control pathways. Robonaut is a human scale robot designed to approach the dexterity of a space suited astronaut. Under teleoperator control, Robonaut has been able to perform many high payoff tasks indicating that it could significantly reduce the maintenance workload for human's working in space. Throughout its development, Robonaut has been augmented to include new sensors and software resulting in increased skills that allow for more shared control with the teleoperator, and ever increasing levels of autonomy. These skills range from simple compliance control, and short term memory, to, most recently, reflexive grasping and haptic object identification using a custom tactile glove, and real-time visual object tracking.

A highly anthropomorphic human scale robot hand designed for space based operations is presented. This five finger hand combined with its integrated wrist and forearm has fourteen independent degrees of freedom. The device approximates very well the kinematics and required strength of an astronaut's hand when operating through a pressurized space suit glove. The mechanisms used to meet these requirements are described in detail along with the design philosophy behind them. Integration experiences reveal the challenges associated with obtaining the required capabilities within the desired size.

Techniques for determining and correcting threaded part alignment using force and angular position data are developed to augment currently limited techniques for aligning threaded parts. These new techniques are based on backspinning a nut with respect to a bolt and measuring the force change that occurs when the bolt "falls" into the nut. Kinematic models that describe the relationship between threaded parts during backspinning are introduced and are used to show how angular alignment may be determined. The models indicate how to distinguish between the aligned and misaligned cases of a bolt and a nut connection by using axial force data only. In addition, by tracking the in-plane relative attitude of the bolt during spinning, data can be obtained on the direction of the angular misalignment which, in turn, is used to correct the misalignment. Results from experiments using a bolt held in a specialized fixture and a three fingered Stanford/JPL hand are presented.

The described mathematical formulations answer the questions "how strong is enough?", "strong in which direction?", and "where is the weakest link?" that robot designers face in the early stages of matching tasks with robotic mechanisms. This work produces 3D visualizations of robot strengths and weaknesses across their dexterous workspaces, aiding interactive manipulator joint and kinematic design. Strength is defined here as the maximum force or moment that the robot can exert on the environment at its point of resolution. This strength is a 6/spl times/1 vector for spatial mechanisms, comprising forces and moments resolved in a global frame. The three factors studied in this work are the design of the manipulator's drive trains, the chain's kinematic parameters, and the articulation of kinematic redundancies as self motion. The mathematical formulation is implemented as a minimum of a set of serial strengths, within which can exist parallel sets as appropriate for the mechanism's kinematics. The strength model is first developed for serial chains, broadened to the statically indeterminant cases of parallel chains, and then applied to a bifurcated system of chains in development at NASA. This work's primary objective is to bring to bear in robot design the mathematical principles usually reserved for robot modeling and control. After a robot has been built, it is often too late to fix a mechanism's many weaknesses. It is our belief that the Jacobian has as much to teach robot designers as it has taught students of dynamics and control theory.

Techniques for determining and correcting threaded part alignment using minimal force and angular position data are developed to augment currently limited techniques for aligning threaded parts. These new techniques ore based on backspinning a nut with respect to a bolt and measuring the force change that occurs when the bolt "drops" into the nut. Kinematic models that describe the relationship between threaded parts during backspinning are introduced and are used to show how angular alignment may be determined. The models indicate how to distinguish between the aligned and misaligned cases of a bolt and a nut connection by using axial force data only. In addition, by tracking the in-plane relative attitude of the bolt during spinning, data can be obtained on the direction of the angular misalignment which, in turn, is used to correct the misalignment. Results from experiments using a bolt held in a specialized fixture and a Stanford/JPL hand are presented.

A technique for determining threaded part alignment using minimal force and angular position data is developed. This technique is based on backspinning a nut with respect to a bolt and measuring the force spike that occurs when the bolt "drops" into the nut. Kinematic models that describe the relationship between threaded parts during backspinning are presented and are used to show how angular alignment may be determined. The models indicate that it may be possible to distinguish between the aligned and misaligned cases between a bolt and a nut by using axial force data only. In addition, by tracking the in-plane relative attitude of the bolt during spinning, data can be obtained on the direction of the angular misalignment. Results from experiments using a bolt held in a Salisbury hand are presented.

A master-slave system is developed to evaluate the effectiveness of telepresence in space telerobotics applications. An operator uses the master's telepresence and virtual reality equipment to control the slave. The slave is a dual-arm, dual-hand robot equipped with a stereo camera platform designed to provide an operator-centered perspective of the remote environment. The initial integration and tests of the system presented several operational issues that are resolved through a variety of shared control techniques and optimized algorithms. The resulting system provides a very flexible capability and is used to perform a range of tasks: grasping and handling toots, manipulating electronics controls, manipulating soft flexible material, and performing planetary geology tasks that involve a variety of manipulation and tool-use skills. Using this system an operator is able to complete most of these tasks in less than 2 minutes.

An analysis of an improved joint drive system for the shuttle remote manipulator system (SRMS), capable of sustaining a single actuator failure, is presented. This system employs a differential gear train with dual input actuators driving a single load. The mathematical model for the drive system includes: gearbox flexibility and damping, motor damping, high gear ratio, and high load impedance. The nonlinear dynamic equations for the system reduce to linear form without approximation. Effects of gearbox placement and load sharing are examined. A feedback system that improves overall response is discussed. Simulation results show that the design is able to sustain a single electromechanical actuator failure without impeding the joint's performance. Plans for a hardware implementation to verify the system are given.>