This paper describes a collision avoidance system using Whole Arm Proximity (WHAP) sensors on an articulated robot arm. The capacitance-based sensors generate electric fields which completely encompass the robot arm and detect obstacles as they approach from any direction. The robot is moved through the workspace using a velocity command generated either by an operator through a force-sensing input device or a preprogrammed sequence of motions. The directional obstacle information gathered by the WHAP sensors is then used in a matrix column maximization algorithm that automatically selects the sensor closest to an obstacle during each robot controller cycle. The distance from this sensor to the obstacle is used to reduce the component of the command input velocity along the normal axis of the sensor, allowing graceful perturbation of the velocity command to prevent a collision.

This paper introduces the notion of virtual fixtures for use in telepresence systems and presents an empirical study which demonstrates that such virtual fixtures can greatly enhance operator performance within remote environments. Just as tools and fixtures in the real world can enhance human performance by guiding manual operations, providing localizing references, and reducing the mental processing required to perform a task, virtual fixtures are computer generated percepts overlaid on top of the reflection of a remote workspace which can provide similar benefits. Like a ruler guiding a pencil in a real manipulation task, a virtual fixture overlaid on top of a remote workspace can act to reduce the mental processing required to perform a task, limit the workload of certain sensory modalities, and most of all allow precision and performance to exceed natural human abilities. Because such perceptual overlays are virtual constructions they can be diverse in modality, abstract in form, and custom tailored to individual task or user needs. This study investigates the potential of virtual fixtures by implementing simple combinations of haptic and auditory sensations as perceptual overlays during a standardized telemanipulation task.

In the areas of virtual reality, telemanipulation, and robot teaching, objects must be manipulated in 3-D space. This paper proposes a 3-D pointing device that uses tactile sensors. First, the concept and construction of the pointing device is described. The key concept of our device is the sensing of the touch pattern of the user's fingers on the device surface and recognition of the number, positions, and pressures of the contact points. Then, the motion commands for manipulating the object are interpreted from this information. The body of the pointing device is a fixed ball or cylinder.

In this paper we describe an algorithm for generating virtual forces in a bilateral teleoperator system. The virtual forces are generated from a world model and are used to provide real-time obstacle avoidance and guidance capabilities. The algorithm requires that the slave's tool and every object in the environment be decomposed into convex polyhedral primitives. Intrusion distance and extraction vectors are then derived at every time step by applying Gilbert's polyhedra distance algorithm, which has been adapted for the task. This information is then used to determine the compression and location of nonlinear virtual spring-dampers whose total force is summed and applied to the manipulator/teleoperator system. Experimental results validate the whole approach, showing that it is possible to compute the algorithm and generate realistic, useful pseudo forces for a bilateral teleoperator system using standard VME bus hardware.

In this paper we examine the role of high-frequency vibrations in the performance of a task in which subjects had to alternately tap in between two strips without applying excessive forces. We conducted three different experiments. In the first two experiments, subjects performed the task using two different actuated styli: a lightweight stylus and a massive stylus. In the first experiment, subjects performed the task using the lightweight stylus with and without masking (provided by continuous actuation of the solenoid). The number of successful taps decreased by a factor of 6.0 when the subjects used the stylus with masking. In the second experiment the massive stylus was used in two different modes. In one mode the solenoid was not actuated and in the second mode, the display mode, the solenoid produced a burst of vibrations when triggered by an electrical switch that closed when the stylus was in contact with the target strips. The performance of the subjects increased by a factor of 1.6 when they used the stylus with the display. These results indicate that people use high-frequency information during manipulation tasks and that a high-frequency display can be used to convey this information.

In early 1993 the space robot technology experiment ROTEX flew with the space-shuttle Columbia (spacelab mission D2 on flight STS-55 from April 26 to May 6). A multisensory robot on board the space-craft successfully worked in autonomous modes, teleoperated by astronauts, as well as in different telerobotic ground control modes. These include on-line teleoperation and tele-sensor-programming, a task-level oriented programming technique involving `learning by showing' concepts in a virtual environment. The robot's key features were its multisensory gripper and the local sensory feedback schemes which are the basis for shared autonomy. The corresponding man-machine interface concepts using a 6 dof non-force- reflecting control ball and visual feedback to the human operator are explained. Stereographic simulation on ground was used to predict not only the robot's free motion but even the sensor based path refinement on board; prototype tasks performed by this space robot were the assembly of a truss structure, connecting/disconnecting an electric plug (orbit replaceable unit exchange ORU), and grasping free-floating objects.

This paper discusses the requirements and an implementation of a virtual environment for telerobotic shared control. The knowledge base that the computing system uses to perform decision making is called the World Model. Visual information from the World Model is displayed graphically as rendered, shaded, texture mapped animated polygons. The core of the shared control system is the computational engine that maintains, displays, visualizes and controls interaction with the World Model. The computational engine can be interrogated at several different levels. During operation: multilevel interaction defined by the program (mouse, keyboard, multi-modal input devices). The visualization system offers a highly interactive operating environment, enabling users to change objects, display attributes, and elements of the World Model in real time.

The ATOP Laboratory at JPL is equipped with a two Universal Motor Controllers (UMCs) for real time control calculations. A high degree of software complexity is not practical for implementation on the UMCs due both to memory and speed limitations as well as minimal high level language support. An analysis of the minimum level of functionality required at the remote site of a teleprogramming system was performed. This minimum set of functionality was then implemented using the UMCs for real-time control calculations and a UNIX based computer communicating via a serial I/O line. The combination of the UMCs, the robotic manipulator, and the UNIX machine form the remote site of the teleprogramming system. Because the serial I/O line cannot provide real time communication, the architecture was designed to be insensitive to unpredictable serial I/O transfer times. A demonstration teleprogramming experiment performed in the ATOP Laboratory at JPL is presented. This paper is intended as an aid to researchers wishing to reproduce laboratory results or perform research in teleprogramming and supervisory control.

Recent advances in real-time 3-D graphics and graphical user interface (GUI) technologies enable development of advanced operator interfaces for telerobotic systems. In particular, we have employed preview/predictive displays with calibrated graphics overlay and X/Motif- based GUIs for efficient and reliable ground-controlled space telerobotic servicing under communication time delay. High fidelity preview/predictive displays have been achieved by an operator-interactive camera calibration and object localization technique that enables reliable matching of simulated 3-D graphics models with the remote site task environment. The developed graphical operator interface supporting telerobotic operations with high-fidelity preview/predictive displays have been successfully utilized in demonstrating a ground- simulated ORU (Orbital Replacement Unit) changeout remote servicing task by remotely operating a robot arm at NASA Goddard Space Flight Center from the Jet Propulsion Laboratory under a varying time delay of up to several seconds. The positioning alignment accuracy achieved by this technique with four camera views was about +/- 5 mm for a tool insertion in the servicing task.

Human operators can have difficulty with accurately controlling the position of a remote manipulator due to, in part, a lack of knowledge of the position of the manipulator with respect to its desired position, destination, or target object to be manipulated. The objective and contribution of the research presented in this paper is to provide a Manipulator Position Display (MPD) to assist a human operator with the operation of remote manipulator systems such as the space shuttle remote manipulator system. The MPD provides the operator with six degree of freedom position cues in a graphical display format for manipulator operations. The display provides these cues based on position (both translation and orientation) data from the manipulator system data stream. Preliminary experimental results have shown that the display significantly improved operator performance for a simulated space manipulation task, while reducing training requirements and operator workload.

For space robots telemanipulation long communication delay and operator load are serious problems. We have developed a useful off-line robot motion planning system. The operator enters commands using object-level instructions, and low-level actions like path planning, sensor feedback, etc. are generated and executed automatically. The command generation and path planning algorithms in this system are original. The path planning algorithm can generate a collision free path in a few seconds using ready-made suggested paths which we call path- templates. The system greatly simplifies the operator's task and also frees the operator from dealing with the communication delay inherent in space robot telemanipulation.

The control feel of the master controller in a telemanipulation system has a critical effect on human-machine system performance. Many theoretical and practical human factors issues on hand controller design and selection remain to be explored. In this paper, two types of hand controllers, elastic versus isometric rate controllers, were studied in a 6 degree-of-freedom pursuit tracking experiment. Twenty-six subjects participated in the experiment. The results showed that the elastic rate controller facilitated significantly lower tracking error than the isometric rate controller, presumably due to richer proprioceptive feedback afforded by the elastic device. It was also found that, although the elastic controller was superior to the isometric controller in general, the magnitude of performance differences between the two controller decreased as subjects gained more experience, supporting the theory that human manipulation shifts from more closed-loop behavior towards open loop skills.

Integrated Services Digital Network (ISDN) communication technology has made it possible to conduct broadband teleconferencing and teleoperation in a method which is much less costly than satellite systems. Fujita Research has taken advantage of this medium to conduct teleoperation tests of a robot system across the Pacific Ocean. Through our application, researchers in Encino, California were able to conduct simple pick and place tasks using a 6 degree-of-freedom robot in Tokyo, Japan. Problems of time delay and low update rates prompted preliminary investigations on the use of visual enhancements. The use of predictive overlays and stereo visual cues were investigated using the ISDN set-up and results confirm past studies which indicate their usefulness in improving teleoperation task performances. This research may eventually be applied to remote monitoring and operation systems of robots on several construction sites or in hazardous environments.

The complexity of telerobotic operations in a cluttered environment is exacerbated by the need to present collision information to the operator in an understandable fashion. In addition to preventing movements which will cause collisions, a system providing some form of virtual force reflection is desirable. With this goal in mind Lawrence Livermore National Laboratory (LLNL) has installed a kinetically similar master/slave system and developed a whole arm collision avoidance system which interacts directly with the telerobotic controller. LLNL has also provided a structure to allow for automated upgrades of workcell models and provide collision avoidance even in a dynamically changing workcell.

Personnel at the Robotics and Process Systems Division (RPSD) of the Oak Ridge National Laboratory have extensive experience designing, building, and operating teleoperators for a variety of settings, including space, battlefields, nuclear fuel reprocessing plants, and hazardous waste retrieval. In the course of the last decade and a half, the RPSD designed, built, and operated 4 telemanipulators (M-2, ASM, LTM, CESAR arm) and operated another half dozen (M-8, Model 50, TOS SM-229, RM-10, PaR 5000, BilArm 83A). During this period, human factors professionals have been closely integrated with RPSD design teams, investigating telemanipulator feedback and feed forward, designing cockpits and control rooms, training users and designers, and helping to develop performance specifications for telemanipulators. This paper presents a brief review of this and other work, with an aim towards providing perspectives on some of the human factors aspects of telemanipulation.

The development of the symbolic inverse of a manipulator Jacobian and its use in controlling the manipulator is discussed. The ease of symbolically inverting a matrix is directly related to the simplicity of the expressions in the matrix. Therefore, a general method for obtaining a simplified form of the Jacobian of a manipulator in any of the intermediate link frames is developed and presented using an example. Also, a general method is presented for the verification of a manipulator Jacobian using the forward kinematics. A hand-waving discussion of the inversion of the Kraft slave Jacobian using symbolic manipulation software is given and the final form of the inverse is presented. An approximate solution to the inverse Jacobian problem for the Kraft slave using symbolic inverses is presented. This solution is used in dealing with the only reachable singularity of the Kraft slave. Finally, a teleoperational control system using the inverse Jacobian to calculate the inverse kinematics is discussed.

This paper addresses the performance of force-reflecting interfaces (`haptic displays'). We suggest that an important measure of performance is the dynamic range of achievable impedances, and that an impedance is achievable if it satisfies a robustness property such as passivity. The effects of sampling, inherent interface dynamics, and sensor quantization on dynamic range are discussed. Techniques for achieving very high impedances are presented.

This paper addresses the problem of optimal motion control for teleoperated surgical robots, which must maneuver in constrained workspaces. The control problem is determining how best to use the available degrees of freedom of a surgical robot to accomplish a particular task, while respecting geometric constraints on the work volume, robot mechanism, and the specific task requirements. We present a method of formulating desired motions in the task space (task space goals) as instances of a quadratic optimization problem, optionally subject to additional linear constraints. The control formalism applies both to kinematically deficient as well as kinematically redundant mechanisms. Specifically, we discuss the problem as it relates to a representative set of tasks in teleoperated navigation of a laparoscopic camera attached to a surgical robot.

This paper presents a controller design framework for a telerobotic system in which the dynamic behaviors of the master robot and the slave robot are functions of each other. These functions, which are set by the designer based upon the application, are described.

This paper presents an overview of the developed modified impedance control (MIC) concept and provides real-time sensor based robot capabilities (gross and fine motion-skills). These skills are mandatory requirements for the telerobotics systems to execute unstructured tasks in an unstructured/uncertain environment (i.e., space, defence operations, etc.). A special emphasis is given to the applications of Space Station Freedom operations. The first part of the paper focuses on the (free) gross motion of the robot manipulators and provides a real-time collision avoidance approach for complete posture (all links) of a robot manipulator in an unstructured environment based on MIC concept. The second part of the paper presents contact motion skills. The experimental demonstrations show that the MIC concept is very efficient, satisfactory, and an important candidate for manipulation control of complex tasks since it allows real-time collision avoidance as well as contact motion capability.

This paper describes a real-time controller architecture being implemented within the Advanced Teleoperation Laboratory at the California Institute of Technology's Jet Propulsion Laboratory. We address system performance/integration trade-off issues with respect to the system's computational needs and the present day technology, while considering a variety of specification goals in addition to performance, e.g., the flexibility to enable fast prototyping of new real-time control algorithms and control/sensor interfaces.

With the development of robotic technology, many robots have been applied to various fields, such as factories, construction sites, and so on. However, most of the robots have been applied to tasks which require them to repeat the same operation for the known environment automatically. Execution of tasks with unknown environments is still a difficult problem for today's autonomous robot. A robotic system which cooperates with a human operator seems to be an effective approach to solve the problem. In this paper, we are going to discuss a control strategy for a robotic system directly maneuvered by an operator like a human amplifier. The results in this paper could be applied to a telerobotic system. We propose an alternative control algorithm based on `virtual tool' dynamics for the mechanical system. The control algorithm is designed without explicit models of the environment and the human operator. By controlling the mechanical system so as to imitate the dynamics of a tool, the algorithm specifies both the human force amplification ratio and the maneuverability of the system.

Robotics technology is being developed to assist in remediating high-level radioactive wastes stored in large tanks at the Department of Energy's Hanford, Washington facility. Remediation activities will likely employ a remotely controlled long reach manipulator, capable of extending over 40 feet from its base, to deploy remediation end-effectors and sensors. Waste inside the tanks consists of hazardous vapor, liquid, sludge, solidified crystal, and various steel pipes and miscellaneous hardware. On November 10 - 12, 1992, a demonstration by Sandia National Laboratories of environmental sensor scanning and robot manipulated steel pipe cutting was performed on a tank mockup located at the Hanford facility. This paper describes the Sandia developed technology implemented at this demonstration.

Research into force reflecting remote manipulation has recently started to move away from common error systems towards explicit force control. In order to maximize the benefit provided by explicit force reflection the designer has to take into account the asymmetry of the bandwidths of the forward and reflecting loops. This paper reports on a high performance system designed and built at Oxford University and Harwell Laboratories and on the preliminary results achieved when performing simple force reflecting tasks. The input device is based on a modified Stewart Platform, which offers the potential of very high bandwidth force reflection, well above the normal 2 - 10 Hz range achieved with common error systems. The slave is a nuclear hardened Puma industrial robot, offering a low cost, reliable solution to remote manipulation tasks.

Using a design method for whole-hand input developed by the authors, the teleoperation of a six-legged mobile robot with manipulator arms has been simulated. Evaluations of three different control structures compare performance between conventional input, exemplified by a set of dials, and whole-hand input using a VPL DataGlove. The evaluations reveal a range of appropriate use of whole-hand input compared to conventional input. In the first example, whole-hand input is superior, in the second, whole-hand input provides performance on par with conventional devices, and in the last, whole-hand input is inferior to conventional devices.

The demand for cost effective space science missions has resulted in the design of spacecrafts whose lives can be extended by changing certain on-orbit replaceable components. To increase cost savings, robots can be used to reduce astronaut time during the servicing and accomplish preparatory and mundane tasks such as the insertion of foot restraints and the opening of stowage areas. The work described in this paper demonstrates the feasibility of a semi- autonomous system capable of assisting astronauts using high fidelity mockups of actual space hardware.

Experiments were conducted with a neutral-buoyancy robot to test whether vehicle station keeping and end effector disturbance compensation significantly affect human teleoperation performance. The vehicle used for experiments, called the Submersible for Telerobotic Astronautical Research (STAR) is a free-flying underwater telerobot equipped with a three degree of freedom arm, a stereo pan/tilt camera platform, and a vision-based navigation system. Using visual feedback from a fixed onboard camera, test subject performed a Fitts- type tapping task with the arm while the vision navigator and control system held the vehicle steady relative to a visual reference target. This paper describes the testbed vehicle, experiments, data analysis, and conclusions.

Several operators performed two solar maximum repair subtasks while viewing the workspace with either three monocular (non-stereo) black and white TV cameras or with one pair of stereo black and white TV cameras. For the monocular TV viewing, the three camera views were presented on three TV monitors, and a second operator controlled the pan, tilt and zoom lens power of two of the cameras. For the stereo TV viewing, neither pan, tilt, nor zoom adjustments were allowed, and no second operator was present. Operators were instructed to perform the subtasks as quickly as possible. In all cases, operators completed the tasks more quickly with stereo TV viewing than with multiple monocular TV viewing. Operators reported the learning of the subtasks to be easier with stereo viewing, and reported the most difficult aspect of learning with monocular viewing to be learning which monitor to observe during the various phases of the subtasks.

Responsiveness is the ability of a telemanipulator to recreate user trajectories and impedance in time and space. For trajectory production, a key determinant of responsiveness is the ability of the system to accept user inputs, which are forces on the master handle generated by user hand acceleration/deceleration (a/d) impulses, and translate them into slave arm acceleration/deceleration. This paper presents observations of master controller a/d impulses during completion of a simple target acquisition task. Power spectral density functions calculated from hand controller a/d impulses were used to assess impulse waveform. The relative contributions of frequency intervals ranging up to 25 Hz for three spatially different versions of the task were used to determine which frequencies were most important. The highest relative power was observed in frequencies between 1 Hz and 6 Hz. The key frequencies related to task difficult were in the range from 2 Hz to 8 Hz. Differences were also observed from 9 Hz to 12 Hz. The results provide clues to the source of the performance inhibition.

The extra vehicular activity helper/retriever (EVAHR) is a prototype for an autonomous free- flying robotic astronaut helper. The ability to grasp a moving object is a fundamental skill required for any autonomous free-flyer. This paper discusses an algorithm that couples resolved acceleration control with potential field based obstacle avoidance to enable a manipulator to track and capture a rigid object in (imperfect) zero-g while avoiding joint limits, singular configurations, and unintentional impacts between the manipulator and the environment.

The EVAHR (extravehicular activity helper/retriever) robot is being developed to perform a variety of navigation and manipulation tasks under astronaut supervision. The EVAHR is equipped with a manipulator and dexterous end-effector for capture and a laser range imager with pan/tilt for target perception. Perception software has been developed to perform target pose estimation, tracking, and motion estimation for rigid, freely rotating, polyhedral objects. Manipulator grasp planning and trajectory control software has also been developed to grasp targets while avoiding collisions. A software simulation of the EVAHR hardware, orbital dynamics, collision detection, and grasp impact dynamics has been developed to test and measure the performance of the integrated software. Performance measurements include grasp success/failure % and time-to-grasp for a variety of targets, initial target states, and simulated pose estimation computing resources.

Robotic capabilities needed for future space exploration mission tasks must be assessed as part of resource-demand balancing and planning. The same statement can be made for other types of resources; and, in fact, the approach presented in this paper applies to those resources as well. The approach, which has been developed over the past 4 years, is based upon integration involving the following methodological components: (1) comprehensive representative schemata for processes, objects, and states of affairs; (2) several types of analysis for processes (as opposed to robotic objects per se); (3) factor space analyses to relate requirements to capacities to design concepts to designs; and (4) evolved practices in the arenas of trade studies and decision aids. The preliminary results of applying the approach are presented.

Fault detection and fault tolerance are increasingly important for robots in space or hazardous environments due to the dangerous and often inaccessible nature of these environs. We have previously developed algorithms to enable robots to autonomously cope with failures of critical sensors and motors. Typically, the detection thresholds used in such algorithms to mask out model and sensor errors are empirically determined and are based on a specific robot trajectory. We have noted, however, that the effect of model and sensor inaccuracy fluctuates dynamically as the robot moves and as failures occur. The thresholds, therefore, need to be more dynamic and respond to the changes in the robot system so as to maintain an optimal bound for sensing real failures in the system versus misalignment due to modeling errors. In this paper, we analyze the reachable measurement intervals method of computing dynamic thresholds and explore its applicability to robotic fault detection.

This paper describes a robotic system which accepts motion and control commands which can be generated autonomously. The system developed has been designed to perform an autonomous grapple based on guidance control feedback provided by an image from a single camera mounted on the slave robot's end effector. The vision system consists of three parts. The first signature based, trained on an arbitrary grapple interface (i.e., no special targets are required for guidance), and provides estimates for the 3D attitude by interpolating sampled signature correlations. These signatures are essentially the distribution of line orientations obtained by radial integration of the FFT of a pre-processed edge image. The second part estimates the range and bearing of the interface based on the first and second moments of the preprocessed edge image of the interface. And the third stage of the algorithm verifies the results.

Autonomous mobile robots must respond to external challenges and threats in real time. One way to satisfy this requirement is to use a fast low level intelligence to react to local environment changes. A fast reactive controller has been implemented which performs the task of real time local navigation by integrating primitive elements of perception, planning, and control. Competing achievement and constraint behaviors are used to allow abstract qualitative specification of navigation goals. An interface is provided to allow a higher level deliberative intelligence with a more global perspective to set local goals for the reactive controller. The reactive controller's simplistic strategies may not always succeed, so a means to monitor and redirect the reactive controller is provided.

The extreme environmental conditions of low Earth orbit impact the design and control of space manipulators, requiring solutions not found in terrestrial applications. Among these conditions are the thermal and vacuum states of low Earth orbit, which are considered here. System models are offered to track thermal state as various internal and external heat transfer mechanisms act on the system. Sensory requirements for thermal management are proposed, with applications described for on line control and off line trajectory design.

Cooperating manipulators in dual-arm robots usually result in a mechanically overconstrained system having more actuator resources than freedom of motion, or an excess of actuator inputs. Decision-making criteria are necessary to make effective use of dual-arm robotic system resources, as well as to enhance performance for either telerobotic or autonomous dual-arm robots. The decision-making criteria are based on parameters inherent to the two cooperating manipulators, and parameters intrinsic to cooperative modes of operation in dual- arm robots. There are many criteria that need to be evaluated and balanced in a complete decision-making system for dual-arm robots. A set of decision-making criteria for dual-arm robots is proposed from which a subset is examined in detail. A graphical simulation of a 17 degree-of-freedom robot performing cooperative dual-arm robotic operations is used to interactively study and evaluate multiple decision-making criteria for dual-arm robots.

One basic robotics task is to position a robot tool point over the geometric center of a symmetric object. This paper presents a fully autonomous control command development technique of linear complexity that solves the robot tool point centering and alignment problem. It is based on non-visual sensor imaging and starts with only a partial image of the target in its field of view. A new capacitive proximity and imaging sensor, called a capaciflector, is being used to obtain 2D discrete images of objects with moderate surface complexity. Considered are spacecraft parts whose imaged surfaces are flat with a simple geometrical shape such as a solid rectangle or a circle.

A real-time laser scanner for object recognition, inspection, and robot control is presented in this paper. Range and intensity images are generated in perfect registration at a rate of 10 Mega-samples per second. Images have a resolution of 483 lines each having 512 pixels. Owing to its compatibility with the video standard RS-170, the range camera can be directly interfaced to an image processor through a video frame grabber with either an analog or digital input. The angular field of view is 30 degree(s) X 30 degree(s). The stand-off distance is 0.5 m and the operational depth of field is 1 m. Furthermore, cooperative targets can be measured up to several meters away from the camera. The range resolution is 12 bits. Under normal operating conditions, a range precision of 0.2 mm and 2 mm are achieved at 0.5 m and 1 m respectively. The testing of the prototype is well under way and soon, images of a variety of representative objects will be available.

This paper describes the significance in human, scientific, and technical terms of the first experiment of robotic telesurgery effected between the Telerobotics Laboratory of the Politecnico di Milano and the Jet Propulsion Laboratory, NASA, in Pasadena California, on 7 July 1993. An Italian surgeon controlled from the U.S. A. an Italian robot in the Telerobotics Laboratory in such a way that the robot performed a biopsy, on a model containing the organs of a pig, carrying out an aspiration of organic material and two incisions for the commencement of the surgical operation of laparoscopy. Transmission was effected by means of a double satellite link with three stations -- one in Italy, one in New York and one in Pasadena -- and two geostationary satellites, the first over the Atlantic and the second over the United States.

A person with limited arm and hand function could benefit from technology based on teleoperation principles, particularly where the mechanism provides proprioceptive-like information to the operator giving an idea of the forces encountered in the environment and the positions of the slave robot. A test-bed system is being prepared to evaluate the potential for adapting telemanipulator technology to the needs of people with high level spinal cord injury. This test-bed uses a kinematically dissimilar master and slave pair and will be adaptable to a variety of disabilities. The master will be head controlled and when combined with auxiliary functions will provide the degrees of freedom necessary to attempt any task. In the simplest case, this mapping could be direct, with the slave amplifying the person's movements and forces. It is unrealistic however to expect that the operator will have the range of head movement required for accurate operation of the slave over the full workspace. Neither is it likely that the person will be able to achieve simultaneous and independent control of the 6 or more degrees of freedom required to move the slave. Thus a set of more general mappings must be available that can be chosen to relate to the intrinsic coordinates of the task. The software structure to implement the control of this master-slave system is based on two IBM PC computers linked via an ethernet.