In this paper, we present a planning and control framework for dynamic, whole-body motions for dynamically stable shape-accelerating mobile manipulators. This class of robots are inherently unstable and require careful coordination between the upper and lower body to maintain balance while performing arm motion tasks. Solutions to this problem either use a complex, full-body nonlinear dynamic model of the robot or a highly simplified model of the robot. Here we explore the use of centroidal dynamics which has recently become a popular approach for designing balancing controllers for humanoid robots. We describe a framework where we first solve a trajectory optimization problem offline. We define balancing for a ballbot in terms of the centroidal momentum instead of other approaches like ZMP or angular velocity that are more commonly used. The generated motion is tracked using a PD-PID cascading balancing controller for the body and torque controller for the arms. We demonstrate that this framework is capable of generating dynamic motion plans and control inputs with examples on the CMU ballbot, a single-spherical-wheeled balancing mobile manipulator.

A control algorithm is developed to enable dynamically stable spherical-wheel robots (ballbots) with arms to detect a heavy object of unknown mass, navigate to it, lift it, transport it, and place it in a desired location semi-autonomously. Previous work has successfully demonstrated two-wheeled dynamically stable mobile manipulator robots transporting heavy objects. We report here the first ballbot to reliably achieve such a task. A successful semi-autonomous lift and transport of a 15 kg heavy box whose actual mass was unknown was achieved using a combination of feedforward and feedback control laws based on a quasi-static center of mass computation. The ballbot's pan and tilt sensor turret tracked fiducial markers on the box. Ballbot-to-human and human-to-ballbot exchanges of a 10 kg heavy object was achieved while dynamically balancing.

We give a system overview and initial experimental results for SIMbot (spherical induction motor based ballbot), a human-sized dynamically stable mobile robot with only two active moving parts. SIMbot is a qualitatively new kind of robot with the potential to drastically reduce mechanical complexity in human-sized mobile robots. SIMbot uses a novel spherical induction motor as its drive mechanism, which is mechanically simpler than other ballbot drive mechanisms. We demonstrate the feasibility of this innovation with experimental results for balancing, stationkeeping, short point-to-point motions, and recovery from initial lean angles.

This paper describes the design, construction, and operation of a closed-loop spherical induction motor (SIM) ball wheel for a balancing mobile robot (ballbot). Following earlier work, this new design has a smaller rotor and higher torques due to the use of six stators in a skewed layout. Actuation and sensing kinematics as well as control methods are presented. In its current implementation, torques of up to 8 Nm are produced by the motor with rise and decay times of 100 ms. Results are presented supporting its potential as a prime mover for mobile robots.

This work presents a method for assisting people in getting out of chairs with a dynamically stable mobile robot. A user study is described in which force profiles and joint trajectories were recorded for a human-human sit-to-stand experiment. These data were used to develop an impedance-based controller designed to allow the ballbot to help people out of chairs. The control strategy is experimentally verified on the ballbot, a balancing mobile robot, exhibiting lean angles of over 15° and applying more than 120 N of assistive force. To our knowledge, this is the first mobile robot to assist people in standing.

The ballbot is a human sized, dynamically stable mobile robot that balances on a single, spherical wheel. The current framework for navigation and control makes the assumption that the robot is operating on a level surface without any center-of-mass offset; however, in practice, such a dynamically stable robot has to be able to successfully navigate sloped surfaces and with such offsets. This work develops the equations of motion for the ballbot system on a sloped surface with a center-of-mass offset. The equilibria of this system are analyzed, and a compensation strategy is formulated that allows the ballbot to operate in the presence of slopes and center-of-mass offsets. This work also develops estimation algorithms for slope and center-of-mass offset angles during station-keeping and trajectory following. Results for compensation and estimation are demonstrated experimentally.

This work presents a method to generate dynamically feasible trajectories for a balancing robot in the presence of obstacles, both static and moving. Intended for use on a ballbot, these trajectories respect the dynamics of the robot, and can be generated in milliseconds. Trajectories were experimentally verified on the ballbot in unstructured indoor environments at speeds up to .7 m/s and distances of up to 25 m. The method presented provides a tractable solution for indoor ballbot navigation, enabling safe movement through unstructured environments.

This paper reports, to our knowledge, the first spherical induction motor (SIM) operating with closed loop control. The motor can produce up to 4 Nm of torque along arbitrary axes with continuous speeds up to 300 rpm. The motor's rotor is a two-layer copper-over-iron spherical shell. The stator has four independent inductors that generate thrust forces on the rotor surface. The motor is also equipped with four optical mouse sensors that measure surface velocity to estimate the rotor's angular velocity, which is used for vector control of the inductors and control of angular velocity and orientation. Design considerations including torque distribution for the inductors, angular velocity sensing, angular velocity control, and orientation control are presented. Experimental results show accurate tracking of velocity and orientation commands.

This work presents a method to find analytic and feasible trajectories for underactuated, balancing robots. The target application is fast computation of achievable trajectories for the ballbot, a dynamically stable mobile robot which balances on a single spherical wheel. To this end, assumptions are made to form the system as differentially flat, and a method of deriving feasible trajectories is shown. These trajectories are tracked on the real robot, demonstrating both point to point motions and recovery from nonzero initial conditions.

We describe an experimental system for the evaluation of teleoperation performance. The system was used in two experiments where operators were assigned the task of discovering a buried object while minimizing contact forces. The studies i) demonstrated the advantage of haptic feedback in the discovery task, and ii) compared different methods for haptic feedback, including our new bimanual method, Touch and Guide in Tandem (TAGIT). Results show that TAGIT enables the effective workspace of the teleoperator to be expanded while minimizing forces from exploration and contact, reducing their variability, and reducing task completion times.

This paper reports a planar induction motor that can output 70 N translational thrust and 9 Nm torque with a response time of 10 ms. The motor consists of three linear induction armatures with vector control drivers and three optical mouse sensors. First, an idea to combine multiple linear induction elements is proposed. The power distribution to each element is derived from the position and orientation of that element. A discussion of the developed system and its measured characteristics follow. The experimental results highlight the potential of its direct drive features.

The ballbot with arms is an underactuated balancing mobile robot that moves on a single ball. Achieving desired motions in position space is a challenging task for such systems due to their unstable zero dynamics. This paper presents a novel approach that uses the dynamic constraint equations to plan shape trajectories, which when tracked will result in optimal tracking of desired position trajectories. The ballbot with arms has shape space of higher dimension than its position space and therefore, the procedure uses a user-defined weight matrix to choose between the infinite number of possible combinations of shape trajectories to achieve a particular desired trajectory in position space. Experimental results are shown on the real robot where different motions in position space are achieved by tracking motions of either the body lean angles, or the arm angles or combinations of both.

This paper presents controllers called motion policies that achieve fast, graceful motions in small, collision-free domains of the position space for balancing mobile robots like the ballbot. The motion policies are designed such that their valid compositions will produce overall graceful motions. An automatic instantiation procedure deploys motion policies on a 2D map of the environment to form a library and the validity of their composition is given by a gracefully prepares graph. Dijsktra's algorithm is used to plan in the space of these motion policies to achieve the desired navigation task. A hybrid controller is used to switch between the motion policies. The results of successful experimental testing of two navigation tasks, namely, point-point and surveillance motions on the ballbot platform are presented.

Robots using ball(s) as spherical wheels have the advantage of omnidirectional motion. Some of these robots use only a single ball as their wheel and dynamically balance on it. However, driving a ball wheel is not straightforward. These robots usually use one or more motor-driven rollers or wheels in frictional contact with the ball wheel. Some slippage can occur with these schemes, so it is important to measure the actual rotation of the ball wheel for good control. This paper proposes one method for measuring the three dimensional rotation of a sphere, which is applicable to the case of a ball wheel. The system measures surface speed by using two or more optical mouse sensors and transforms them into the angular velocity vector of the ball, followed by integration to give the rotational angle. Experiments showed the correctness of this method, yielding an error of approximately 1%, with a 10 ms response.

Developed by Prof. Ralph Hollis in the Microdynamic Systems Laboratory at Carnegie Mellon University, Ballbot is a dynamically stable mobile robot moving on a single spherical wheel providing omni-directional motion. Unlike statically stable mobile robots, dynamically stable mobile robots can be tall and skinny with high center of gravity and small base. The ball drive mechanism is a four motor inverse mouse-ball setup. An Inertial Measuring Unit (IMU) and encoders on the motors provide all information needed for full-state feedback. Ballbot has three legs that provide static stability when powered down and is capable of auto-transitioning from the statically stable state to the dynamically stable state and vice versa. It is also capable of yaw rotation about its vertical axis. An absolute encoder provides the relative angle between the IMU and the ball drive unit.

Human-Robot Physical Interaction is an important attribute for robots operating in human environments. The authors illustrate some basic physically interactive behaviors with dynamically stable mobile robots using the ballbot as an example. The ballbot is a dynamically stable mobile robot moving on a single spherical wheel. The dynamic stability and robust controllers enable the ballbot to be physically moved with ease. The authors also demonstrate other behaviors like human intent detection and learn-repeat behavior on the real robot.

The ballbot is a dynamically stable mobile robot that moves on a single spherical wheel and is capable of omnidirectional movement. The ballbot is an underactuated system with nonholonomic dynamic constraints. The authors propose an offline trajectory planning algorithm that provides a class of parametric trajectories to the unactuated joint in order to reach desired static configurations of the system with regard to the dynamic constraint. The parameters of the trajectories are obtained using optimization techniques. A feedback controller is proposed that ensures accurate trajectory tracking. The trajectory planning algorithm and tracking controller are validated experimentally. The authors also extend the offline trajectory planning algorithm to a generalized case of motion between non-static configurations.

Unlike statically stable wheeled mobile robots, dynamically stable mobile robots can have higher centers of gravity, smaller bases of support and can be tall and thin resembling the shape of an adult human. This paper concerns the ballbot mobile robot, which balances dynamically on a single spherical wheel. The ballbot is omni-directional and can also rotate about its vertical axis (yaw motion). It uses a triad of legs to remain statically stable when powered off. This paper presents the evolved design with a four-motor inverse mouse-ball drive, yaw drive, leg drive, control system, and results including dynamic balancing, station keeping, yaw motion while balancing, and automatic transition between statically stable and dynamically stable states.

In this paper we present our efforts to design a system for feeding back useful haptic information from a highly dynamic running robot to a remote operator using a haptic device. Without adding additional sensors, the legs of the robot are used as feelers to give the operator the capability to both explore and manipulate the robot's environment and to gather meaningful information about properties not captured by visual feedback like weight, movability and structure of an encountered object. We show the capabilities of the system in a user study with both trained and untrained operators.

Multi-wheel statically-stable mobile robots tall enough to interact meaningfully with people must have low centers of gravity, wide bases of support, and low accelerations to avoid tipping over. These conditions present a number of performance limitations. Accordingly, we are developing an inverse of this type of mobile robot that is the height, width, and weight of a person, having a high center of gravity, that balances dynamically on a single spherical wheel. Unlike balancing 2-wheel platforms which must turn before driving in some direction, the single-wheel robot can move directly in any direction. We present the overall design, actuator mechanism based on an inverse mouse-ball drive, control system, and initial results including dynamic balancing, station keeping, and point-to-point motion

We present what we believe to be the first free-roaming planar motors. Unlike conventional planar motors requiring a tether for air, power, and signals, these planar motor forcers are tetherless and self-contained. We have designed and operated such forcers utilizing both mechanical bearings and air bearings with air supplied by an on-board pump. Considerations of on-board power, bearing means, wireless communications, and systems integration are discussed in the paper. Performance results for the air bearing model are reported, and work needed to develop future practical autonomous precision planar mobile robots based on tetherless planar motors is given.

This paper describes the design and operation of a coarse-fine teleoperation system with bilateral force reflection. The system uses a 6-DOF magnetic levitation device (MLD) master to control a 6-DOF fine-positioning MLD slave mounted on a 6-DOF coarse positioner arm. Coarse-fine teleoperation expands the workspace of the slave MLD while retaining its frictionless characteristics, fine position resolution and high bandwidth. Several modes of operation are explored. In unilateral coarse-fine mode (UCF) the master controls the wrist position which in turn controls the coarse positioner. In master deadband rate control mode (MDRC) the master position controls the coarse positioner velocity while wrist position is used to reflect contact forces to the master. In symmetric bilateral control (SBC) mode the wrist and master servo to each other's positions providing force reflection during fine manipulation. Operation of each system mode is investigated during a simple assembly task.

We report an experimental high-fidelity system for making psychophysical measurements on human operators performing real, virtual and real-remote 3D haptic manipulation tasks. Operators interact with task environments through six-degree-of-freedom (6-DOF) Lorentz magnetic levitation haptic devices. This arrangement allows the operator to exert and experience real, virtual, and real-remote forces/torques using the same 6-DOF master device. In the virtual task scenario, interactions are rendered haptically. In the real task scenario, the manipulation of the haptic device interacts by direct mechanics with a real environment. In the remote-real scenario, interactions with a remote real task environment are mediated through a 6-DOF Lorentz magnetic levitation slave device carried by a 6-DOF robot arm. In all three scenarios, visual feedback is provided by a graphical display. The system records accurate positions/orientations and forces/torques as a function of time. These records can be parsed automatically and analyzed off-line to evaluate operator performance.

We describe the mechanical and electronic design of a precision, 2-DOF robot manipulator. The manipulator is one of a wide variety of modular robotic agents in the minifactory, a rapidly deployable precision assembly system under development in our laboratory. The manipulator, in cooperation with another type of 2-DOF robot, termed a courier, can perform 4-DOF assembly operations emulating a SCARA robot. This arrangement provides increased precision, higher throughput, smaller footprint, and increased flexibility relative to the SCARA robot.

We describe an experimental arrangement for comparison of user performance during a real and a virtual 3D peg-in-hole task. Tasks are performed using a unique six-degree-of-freedom (6-DOF) magnetic levitation haptic device. The arrangement allows a user to exert and experience real and virtual forces using the same 6-DOF device. During the virtual task, a peg and hole are rendered haptically, and visual feedback is provided through a graphical display. During the real task, a physical peg is attached to the underside of the haptic device. Using only real forces/torques, the peg is inserted into a hole in a plate attached to a force/torque sensor, while positions/orientations are measured by the haptic device. positions/orientations and forces/torques are recorded for both modes. Preliminary results indicate increased task time, larger total forces and more failures occur with the virtual task. Recorded data reveal user strategies that are similar for both tasks. Quantitative analysis of the strategies employed should lead to identification of significant factors in haptic interface design and haptic rendering techniques.

A distributed cooperative coverage algorithm DC/sub R/ is presented, which is derived from an earlier complete single-robot algorithm, CC/sub R/. DC/sub R/ executes independently on each robot in a team where the individual robots do not know the initial locations of their peers and applies to systems of robots operating in a rectilinear environment that use only intrinsic contact sensing to determine the boundaries of the environment. Due to the reactive nature of CC/sub R/, the natural extension to DC/sub R/ preserves the completeness properties of the single-robot algorithm, and the outline of a completeness proof of DC/sub R/ is also presented. DC/sub R/ has been implemented in simulation, and directions for future work are presented which will make the algorithm more suited to physical robot systems.

We document our initial efforts to instantiate visually-guided cooperative behaviors between robotic agents in the minifactory environment. The minifactory incorporates high-precision, 2-DOF robotic agents to perform micron-level precision 4-DOF assembly tasks. Here we utilize two minifactory agents to perform visual servoing. We present a detailed description of the control and communication systems used to coordinate the agents. To provide a suitable communications infrastructure, we describe the development of a new interagent communication system which uses low-latency protocols carried by a commercial 100 Mb Ethernet network. Finally, we conclude by presenting experimental results front our first coordinated multi-agent task, the visually guided positioning of a small medical device.

The need for high precision robotic systems has increased in recent years. Examples include automated assembly of complex electromechanical products such as disk drives, cameras, and medical devices. In this paper we present an optical sensor which can sense motion with better than 0.15 /spl mu/m resolution while also offering the advantages of insusceptibility to ambient lights, high bandwidth, and a large dynamic range. The sensor is used for calibration and coordination tasks in a novel agile robotic system (minifactory) which employs cooperative 2-DOF robots. The sensor can also be used for testing, calibrating, and position sensing in many other high precision robotic applications.

A high performance six degree-of-freedom magnetic levitation haptic interface device has been integrated with a physically-based dynamic rigid-body simulation to enable realistic user interaction in real time with a 3-D dynamic virtual environment. The user grasps the levitated handle of the device to manipulate a virtual tool in the simulated environment and feels its force and motion response as it contacts and interacts with other objects in the simulation. The physical simulation and the magnetic levitation controller execute independently on separate processors. The position and orientation of the virtual tool in the simulation and the levitated handle of the maglev device are exchanged at each update of the simulation. The position and orientation data from each system act as impedance control setpoints for the other, with position error and velocity feedback on each system acting as virtual coupling between the two systems. The setpoints from the simulation are interpolated by the controller at the faster device control rate so that the user feels smooth sliding contacts without chattering due to the slower updates of the simulation. The simple feedback coupling between the two systems enables the overall stiffness and stability of the combined system to be tuned easily and provides realistic haptic user interaction. Sample task simulation environments have been programmed to demonstrate the effectiveness of the haptic interaction system.

Industrial Sawyer motor technology has existed for nearly three decades, traditionally being operated as open-loop positioners. Such systems can attain micron level motion resolution and open-loop repeatability of 10 /spl mu/m. This paper outlines a optical sensing technique for Sawyer motors which is capable of sub-micron resolution. Such a sensor can easily be integrated into a motor enabling closed-loop control and sub-micron repeatability.

Planar linear motors have been shown to be capable of fast accurate 2-DOF motions, making them useful for assembly tasks. However, the lack of sensing capability has limited their applications. Previous planar motor sensors were precise but bulky and sensed only a single direction of motion. We present a small integrated 3-DOF sensor capable of 0.2 /spl mu/m position resolution (1/spl sigma/) which has been integrated with an existing planar linear motor without changing its overall size. Physical design of the sensor is presented as well as the custom electronics used to produce a low-noise DC output from the small signal levels of the sensor. Finally, we present results that demonstrate the resolution and accuracy of the complete sensor system.

Planar linear motors (Sawyer motors) have been used in industry as open-loop stepping motors, but their robustness and versatility has been limited. Using a sensor integrated into such a motor, a closed-loop 3-DOF controller has been implemented. The software-based control system consists of a commutator for computing amplifier currents from actuator forces, a force resolution function for solving the redundant actuation and saturation problems, and an observer for producing a velocity estimate, together with a PID controller. Experiments are performed using a 2-axis laser interferometer to show that the controller has sub-micron resolution, 2 /spl mu/m peak-to-peak repeatability, and settling times after trajectories of about 20 ms. Limitations of the PID controller are discussed and ideas for improvements are presented.

This work examines force modeling and software-based commutation for closed-loop control of planar linear motors, motivated by the need for a robust and versatile planar robot for precision assembly. The approach taken is to make measurements of the static and dynamic force capabilities of the motor as directly as possible, and determine the applicability of simple models commonly used. Measurements of force ripple, linearity with current, force reduction with skew angle, and eddy current damping forces are presented. The high-frequency current changes required for high-speed motion are shown to make the system sensitive to both the latency and update rate of the commutator and to limit the force generation capabilities at high velocities. Although this effect is caused by multiple sources, it is shown that it is well modeled as a scalar with units of time.

Abstract: "This paper serves to present our initial thoughts on the design of an architecture for highly flexible, modular, and distributed precision assembly systems. Through the use of a unified design, simulation, programming, and monitoring environment coupled with self- representing cooperative agents this architecture will significantly simplify the process of factory design and deployment. Our demonstration of this architecture takes the form of a 'table top' sized miniature assembly system targeted at the partial assembly of high-density magnetic storage devices."

A new configuration of cooperative 2-DOF robots is presented for use in precision assembly applications. This configuration consists of direct drive planar linear motors traveling on a tabletop platen stator surface combined with overhead 2-DOF manipulators. We outline desirable features of a precision assembly system and use them to compare the proposed configuration to typical configurations that use 4-DOF robots. Advantages of the new approach are used to motivate further research needed to fully develop precision assembly systems based on this concept.

We have been working on developing a visual/haptic interface for virtual environments. The authors have previously (1996) proposed a WYSIWYF (What You See Is What You Feel) concept which ensures a correct visual/haptic registration so that what the user can see via a visual interface is consistent with what he/she can feel through a haptic interface. The key components of the WYSIWYF display are (i) vision-based tracking, (ii) video keying, and (iii) physically-based simulation. The first prototype has been built and the proposed concept was demonstrated. It turned out, however, that the original system had a bottleneck in the vision tracking component and the performance was not satisfactory (slow frame rate and large latency). To solve the problem of our first prototype, we have implemented a fast tracker which can track more than 100 markers in video-rate. In this paper, new experimental results are shown followed by the improvements of the vision-based tracking component.

A high-performance magnetic levitation haptic interface has been developed to enable the user to interact dynamically with simulated environments by holding a levitated structure and directly feeling its computed force and motion responses. The haptic device consists of a levitated body with six degrees of freedom and motion ranges of /spl plusmn/5 mm and /spl plusmn/3.5 degrees in all directions. The current device can support weights of up to 20 N and can generate a torque of 1.7 Nm. Control bandwidths of up to 50 Hz and stiffnesses from 0.01 to 23 N/mm have been achieved by the device using a digital velocity estimator and 1 KHz control on each axis. The response of the levitated device has been made successfully to emulate virtual devices such as gimbals and bearings as well as different dynamic interactions such as hard solid contacts, dry and viscous friction, and textured surfaces.

Planar linear motors, e.g. Sawyer motors, operate in an open-loop stepping-manner. This mode of operation makes them: 1) susceptible to loss of steps, 2) unable to reject external disturbances, 3) unable to provide controlled forces, and 4) unable to provide high stiffness. These limitations, in turn, restrict their usefulness in a wide range of robotic applications. Suitable position sensing and control technology, when added to such motors, can to a large degree eliminate these problems. In this paper we present two new sensor technologies for planar motor systems: one uses an AC magnetic technique, the other uses an optical fluorescence technique. The magnetic sensor has achieved 1 /spl mu/m position resolution and is compact and easy to fabricate. The optical fluorescence sensor has the advantage of complete insensitivity to nearby motor fields. Either technology has the potential to greatly improve future robotic systems that are based on planar linear motors.

A new approach to the design of teleoperation systems is presented. It is proposed that the teleoperation slave be a coarse-fine manipulator with a fine-motion wrist identical to the teleoperation master. By using a combination of position and rate control, such a system would require only small operator hand motions but would provide low mechanical impedance, high motion resolution and force feedback over a substantial volume. A new teleoperation system, consisting of a conventional manipulator and two identical magnetically levitated wrists, has been developed using this approach and is described in this paper. Aspects of mechanical, system and computational design are discussed. It is shown that the best way to position the slave is by decoupling position and rate control, with the conventional robot controlled in rate mode and its wrist in position mode. Kinesthetic feedback is achieved through wrist-level coordinated force control. Transparency is improved through feedforward of sensed hand forces to the master and environment forces to the slave. To maintain stability, the amount of damping in the system is controlled by the sensed environment forces. Experimental results demonstrating excellent performance are presented.

Assembly accuracies of 0.05-0.1 mm, needed for electronic product manufacture, are attainable with current automated manufacturing equipment. Aggressive electronic system designs will require manufacturing accuracies in the range of 0.005-0.01 mm. A system and strategy is described here to fabricate circuit boards requiring alignment accuracies of 7.5 /spl mu/m using existing automation equipment with enhancements and typical manufacturing line fixtures. The system, configured as a pilot workcell, consisted of an IBM 7576 coarse positioning robot, a fine positioning manipulator, an optical sensing system and a unique bracing method to reduce environmental disturbances. The strategy was to use a coarse/fine placement technique with sensing to align, stack and bond individual test laminates with patterns of 100 /spl mu/m holes. The results showed that pairs of holes were consistently aligned to 2-5 /spl mu/m which surpasses the 7.5 /spl mu/m manufacturing requirement.>

The authors discuss design considerations and control issues of a fine-motion device which provides very high performance, precision, and speed (12-g acceleration, 0.2 mu m resolution). Coupled with a standard industrial robot and external sensing, improvements in precision up to two orders of magnitude can be achieved. Software enables it to be operated from a serial port of an IBM PC or PS/2. The moving element of the positioning head can be commanded to move over a +or-1 mm range in x and y, and rotated up to +or-1.75 degrees in theta about the z axis. The device provides variable compliance with compliance set by loop gain in a digital controller. It incorporates a novel three-degree-of-freedom motor, a pair of internal position sensors to enable controlled motion in the plane, and an air bearing to support the moving elements.>

Previous work on coarse-fine robots has emphasized improved positioning resolution or improved compliant behavior. The authors stress the use of the coarse-fine system's redundant degrees of freedom for bracing of either the coarse or the fine manipulator to execute an assembly task reliably. They refer to real bracing if the constraining forces and torques are supplied by mechanical means, and to virtual bracing if they are supplied by a control system. An application of the coarse-fine robot bracing approach to precision laminate registration and stacking for electronic packaging is presented. Results show that assembly errors of only a few microns can be readily obtained, approximately an order of magnitude improvement over the unbraced coarse-fine case, and nearly two orders of magnitude better than those attainable by coarse manipulation alone.>

The authors present a force-reflecting teleoperation system. The system consists of a conventional manipulator equipped with a magnetically levitated teleoperation master identical to the wrist. Aspects of mechanical, system, computational, and controller design are discussed. An approach to the design of force-reflecting teleoperation controllers is also discussed. Admittance matrices and the parameterization of all stabilizing teleoperation system controllers are used to guarantee teleoperation system performance and stability. Some preliminary experimental results are presented.>

A high-performance six-degree-of-freedom magnetically levitated fine-motion wrist with programmable compliance is described. Design considerations, a discussion of the major elements of the device, and issues of modeling, kinematics, dynamics, and control are presented. A prototype wrist which has been built and controlled successfully is discussed. Experimental results, including high bandwidth position control, compliant control, and the emulation of several mechanisms through software gain setting which establish the use of magnetically levitated robot wrists as an option for manipulation tasks requiring high precision and fine compliant motions, are included.>

A novel method for controlling the endpoint position of a coarse-fine manipulator using a fast vision sensor is presented. With the vibrational disturbances caused by both the manipulator and its environment can be attenuated for disturbance frequencies up to 8 Hz using active feedback control. At the same time, the bandwidth of the control response extends to 30 Hz. By visually sampling the workpiece surface within the operating range of the fine manipulator, a relative displacement in x and y can be estimated by correlating successive image frames. Because of the high computational complexity and the speed requirements, a dedicated correlation processor based on three digital signal processors has been developed for each axis. Results of simulations and experiments with the correlation processor are presented.>

Special-purpose sensors and hybrid mechanical systems have been developed to assemble advanced electronic circuits. State-of-the-art robot systems that are used for electronic assembly and accuracy requirements for the developing surface mount technology are reviewed, showing the need for improved robot positioning capability. Examples of laboratory experiments that demonstrated successful precision assembly or testing using an ordinary robot, a micropositioner, and an optical sensing system are given. These involve fiber-optic sensing, machine-vision endpoint sensing, and hybrid tactic-vision endpoint sensing.>

Reference is made to a previous paper by the authors and A.P. Allan (1988) in which the motivation for and the design of a magnetically levitated robot wrist have been described. The kinematic and dynamic transformations that are required for the control of this 'magic' wrist are presented. The wrist's dynamics are formulated in terms of the Euler quaternion, and the advantages of such a formulation, as well as control schemes based on it, are discussed. The prototype magic wrist was successfully controlled with a simple PD (proportional differential) feedback compensator, both in position and compliance modes. The result obtained, including synthesis of several mechanisms, establish the feasibility of the approach proposed in the original paper.>

We describe in this paper a small, fast, precise positioner which can be attached to the terminal link of a general purpose robot for assembly and inspection tasks. The device is a very compact combination of two-dimensional linear motor, flexure springs, position sensor and digital controller. Peak velocities and accelerations of 24.8 cm/sec and 8.7 g are achieved. Positioning resolution is 0.5 μm using the built-in sensor, and compliance can be varied dynamically by the digital controller loop gain. Design considerations, discussion of the major elements of the device, and performance measurements are presented.