Nandagopal Ramachandran
Department of Industrial and Systems Engineering
Virginia Polytechnic Institute and State University
�

Abstract

Putting man in space and having him work there is costly. This paper discusses the alternative for putting man in space, teleroperators. The successful implementation of teleoperator technology for space teleoperation requires designing of the teleoperator by dividing it into subsystems and then designing each subsystem. Each design is finally integrated into a complete design of the teleoperator to be used in space. This paper discusses the design of these subsystems and also discusses the problems associated with these designs. Finally it also discusses some solutions to these problems.
�

Background

May be the first teleoperator in history would be the ferocious looking automaton designed by Baron Von Kempelen to play chess. A chess expert, who was hidden beneath the chess table, by using various mechanical linkages, manipulated this automaton. Those were the days when man didn�t have to work in the hazardous environments. Now man has to work in space, deep seas, and other hostile environments. Man is building machines that could replace him in these environments and do the tasks in these environments in the same way as humans would do. These machines act as auxiliaries of man in which wires and metal replace nerves and muscle tendons. These man-machine systems are called as the �Teleoperators�.

If the telerobot performs tasks without human assistance or supervision then the telerobot operates as a robot. If the telerobot performs tasks by mimicking the movements of the human operator, through a master-slave system then the telerobot is operating as a teleoperator. Telerobots used in space merge both teleoperator and robot capabilities of the telerobot to perform tasks.

In space there are targets which are so far away that a proxy can only explore them. Here comes the importance of teleoperators in space applications. Then there are tasks that are difficult for the astronauts to perform there also teleoperators come for the help of man. Teleoperators can be used in space to accomplish more efficiently and safely these tasks that would otherwise be impossible for the humans to accomplish. Teleoperators are used in space to do tasks like inspection, maintenance, repair, module changeout, clean up, tending science experiments, capturing and despinning satellites, etc.

A typical space teleoperator system consists of a control station that could be located in a spacecraft or in earth and telerobot. The control station is the place where the human who controls the remote manipulator is located. He has hand controllers and joysticks to control the remote manipulator. The control station is located on the earth if the teleoperator is used for deep space exploration and in the spacecraft if the teleoperator is used in earth orbit operations like space station construction. The teleoperator system can be modeled as in the following diagram.

The control station includes interfaces that the operator uses to comprehend the remote task and control the telerobot. Control station includes joysticks, data gloves, hand controllers and powerful computer to convert the operator�s movement into signals and signals from the remote site into data that the operator can comprehend. The control station is connected to the remote station through data links of high bandwidth.

The remote site includes the remote station computer, which receives commands from the control station and performs the tasks with the help of manipulators. The manipulators have several degrees of freedom usually greater than six for efficient reproduction of human behavior in the remote environment. The remote station robot has sensors for sensing force, position, and mobility, cameras for visual feedback and other auxiliary control systems.

Teleoperator Subsystems

To fulfill man�s objective in space the teleoperator must be able to propel itself in space from one place to another, communicate with it�s status both operational and positional to the operator and above all it should be able to effectively substitute man in the environment being explored. To achieve this, the teleoperator should have an array of subsystems that makes it emotional, physical, and above all economic substitute of man. The different subsystems needed for effective teleoperation in space are,

1.Actuator subsystem
2.Sensory subsystem
3.Control subsystem
4.Physical subsystem.

The actuator subsystem consists of the effector portion of the teleoperator. It facilitates the reproduction of manipulation and dexterous activities that the human operator is doing at the control station. The sensory subsystem is the emotional part of the teleoperator. It helps in sensing the remote environment to give the human operator a feel of telepresence [5]. The control subsystem includes the human operator in loop and it controls all other subsystems to complete the task at hand. The physical subsystem helps in maintaining the teleoperator in position and supply power for various manipulative activities. The complete subsystems can be modeled as in the following diagram.

Successful teleoperation needs the effective designing of all the subsystems based on certain principles, which will be discussed later.

Design Principles

The teleoperator subsystems are to be designed using certain principles. The principles discussed here does not take into consideration mechanical aspects of design like, force, mass, rigidity of the links, etc. These principles take into consideration the design philosophy for teleoperator performance, system effectiveness, etc. The design principles are stated below are for general teleoperator design and not for space teleoperation in particular. Certain principles will be less applicable to space teleoperation. The following are the principles that are to be kept in mind while designing a teleoperator for space application.

1.All of human�s sensory abilities are to be incorporated into the system. The human senses are vision, audition, haptic, olfactory and gustatory. In space teleoperation visual and haptic senses are only needed. Since the other senses are not needed in space teleoperation they need not be implemented in space teleoperation.

2.The responses of the teleoperator to the movements of the human should not be sluggish and too sensitive. If it is sluggish the movements of the teleoperator will be scaled down in comparison to the movements of the human and if it is sensitive the movements of the teleoperator will be scaled too high in comparison to the movements of the human operator. A moderate scaling of the teleoperator arm in comparison with the human arm is desirable. In space teleoperation scaling is an important factor as large area is to be covered by the teleoperator and there is the other extreme of having to do minute manipulative tasks.

3.There should be visual correspondence [4] for visual feedback display. To achieve visual correspondence the visual display should move along with the movement of the head of the human, i.e., when the human turns his head he should see a different part of the work environment. Head mounted displays can be used for visual feedback.

4.If the force feedback is given directly to the human through force reflecting feedback gloves and others the force should be scaled down to such an extend that the strength of the force should not tire the human. In space teleoperation the forces involved can be very high to very low and so to extremes of scaling is required. More often visual displays of force feedback are used in space teleoperation.

5.Onboard computers should be programmed for autonomy. The teleoperator should be able to do tasks with supervisory control [5].

6.Due to time lag, in space applications like planetary exploration preview control [4] should be employed in the teleoperator.

7.The teleoperator should be able to do a variety of tasks. It should be designed to be a general-purpose machine [4] to have launch economy [1].
Based on the above stated principle each subsystem is to be designed.

Actuator Subsystem

The actuator subsystem carries out manipulations and dexterous activities that are ordered by the human operator. The classical master-slave system is the best choice for space teleoperation. The actuator subsystem essentially consists of a master part and a slave part. The master part of the actuator subsystem is located at the control station and slave part is located at the remote station. The master-slave system is a bilateral system, which provides two ways between operator and remote manipulator [2]. The slave should have at least six degrees of freedom to effectively reproduce the tasks that human is doing. The actuator subsystem should have the following capabilities,

1.It should be able to transform the six dimensional motion of the human hand to the six dimensional motion of the slave arm.
2.It should be able to reflect the force acting on the slave arm back to the master arm.

In space due to micro gravity the weight of the actuator subsystem is not at all a major factor. The total weight of the actuator subsystem is determined by the launching economy criteria [1]. To achieve launch economy the weight of the actuator subsystem has to be as small as possible. So while designing the actuator subsystem the materials are to be selected in such a way that the weight of the subsystem is least to achieve launch economy.

Different master-slave actuator subsystems are conceived depending on the way in which torque and force are generated in the slave. The different types of master-slave actuator subsystems considered are mechanical actuator subsystem, hydraulic actuator subsystem, and electrical actuator subsystem. The following are the actuator subsystems,

1.Mechanical actuator subsystem, in which the force and torque are generated with the help of mechanical linkages from the master. For this kind of actuator subsystem the operator needs to be in close proximity of the teleoperator. In space teleoperation the use of mechanical linkages is not feasible due to this constraint of close proximity between the human operator and teleoperator.

2.Hydraulic and pneumatic actuator subsystem, in which force and torque are generated with the help of air pistons and hydraulic pumps. This subsystem needs air and fluids for their operation. In space this constraint causes a major problem. Hydraulic systems are bulky and heavy this will have an effect on launch economy. So the use of hydraulic system in space is not feasible.

3.Electrical actuator subsystem, in which electrical current and motors are used to generate force and torque. This actuator system is the obvious choice in space teleoperation in which there is interplanetary exploration. Solar panels and fuel cells are used in this type of subsystem to generate electric power. If the teleoperator is located in the earth orbit and attached to a spacecraft, then electric power can be derived directly from the spacecraft itself. Due to it�s advantages in space teleoperation conditions the electrical actuator subsystem is the obvious choice for space teleoperation.
In an electrical actuator subsystem electrical solenoids and servomotors are used to generate force and torque. The servomotors at the master end capture the motion of the human operator and transmit them to the remote slave through radio links. The master captures the movements of the human by the help of solenoids and electrical servomotors, and then these movements are converted into analog signals. The analog signals are again converted into digital signals by the computer at the control station and send across to the remote station through radio links where these digital signals are converted back to analog signals by the remote station computer and fed to the slave. The slave uses solenoids and electrical servomotors to convert these signals back to the movements that the human at the control station had produced. Thus with the help of this master-slave the movements of the human are reproduced exactly by the teleoperator. The actuator subsystem is bilateral and so there is communication in the other direction also. Any force that is encountered by the slave is captured and converted into digital signals and transmitted to the control station. At the control station the same procedure is used to convert the signals back to the motion of the master arm. The whole actuator subsystem can be modeled as in the following diagram,

In space teleoperation the best choice of actuator subsystem is electrical manipulator. The exact reproduction of human movements at the teleoperator is not practical and it can involve problems if there is treble in the human arm these treble can appear at the teleoperator and can cause errors during manipulation in space, which involves intricate movements of the tools used. So the movements of the human arm at the control station has to be scaled so that the treble in the human arm is not transmitted to the remote teleoperator. Then the exact reproduction of forces that are acting at the teleoperator end causes problems of fatigue for the operator as large amount of forces are involved in space teleoperation. So the forces involved have to be scaled down or forces are to be displayed outside using television monitors.

Sensory Subsystem

Teleoperator sensory subsystem is the emotional part of the teleoperator-human system. Teleoperator sensors vary a lot from one telerobot to another. At the sensory extreme the teleoperator far away from earth orbit, like the planetary explorers, may be controlled by a human operator who is sitting on earth station and surrounded by a bank of electronic consoles and displays that convey to him the sight, sound, and feel of the alien environment providing for telepresence [5]. The basic principle of sensory subsystem design for effective teleoperation in space is that for effectiveness man has to project his presence in the alien space and he should be able to get a feel of the remote environment. The main function of the sensory subsystem is to faithfully produce the physical properties of the remote environment that are needed to attain telepresence [5] and for the man to do his job well.

In space the most important senses that are to be sensed and provided as feedback to human operator are visual, and force and tactile senses. There are sensors that are used to avoid collision with other object as in proximity sensors and navigation sensors. There are navigation sensors and status sensors, which provides feedback regarding the condition the teleoperator is in. But of all these the most important sensors in space teleoperation are visual feedback and force feedback sensors.

Visual Feedback Sensors

To see the target environment, remote manipulator, and their spatial relationships directly, visual feedback devices are used. In space teleoperation two types of visual feedback methods can be implemented, direct vision and remote television visual feedback.

Human factors considerations demand the following two considerations for implementing direct viewing in space teleoperation visual feedback,

1.Good lighting of the remote site that is being worked upon. In space the lighting is at the extreme ends. Sometimes when the sun is shining directly and the light intensity is maximum. This causes strain for the eyes. Then there is the condition in which the sun is blocked by other bodies in space and the intensity of light is minimum. This also causes strain to the eyes. So good lighting is difficult to get in space teleoperation. The problem of low intensity of light can be overcome by artificial lighting of the remote environment. For maximum intensity the solution suggested is that the operator should wear a light protecting glass on his eyes. This problem of lighting exists only in low earth orbit where the operator and telerobot are in close proximity and the telerobot is attached to the spacecraft. In the case of teleoperation in interplanetary operations this problem doesn�t exist as the operator is out of direct visual reach of the telerobot.

2.A good view path that let the operator see without strain. The light reflecting from the target being manipulated and the manipulator causes the strain. There should not be any obstruction in the line of gaze of the operator.
For direct viewing windows are provided in the spacecraft but with invention of high-resolution cameras and 3D television the usage of direct vision visual feedback is reduced as it has lot of limitations. But the direct viewing is used as an augmentation to the visual feedback using cameras and television

When it comes to deep space teleoperation like interplanetary exploration direct vision feedback is not possible. In such cases remote cameras and televisions are to be used to provide visual feedback. In such a system cameras are fitted to the teleoperator, which senses the remote environment and converts the signals into digital form and sends it back to the control station, where the digital signals are decoded and fed to a 3D television. The 3D television provides a three dimensional view of the remote environment.

In visual feedback involving cameras and television, even if there is deficiency of light in the environment the visual feedback that can be obtained will be adequate due to the possibility of artificial lighting. On the other extreme if there is excess light also the visual feedback will be of high quality. The camera can change aperture, insert filters and so on and get a good image of the remote environment and can discard the unwanted data.

The transfer of visual data from the remote environment to the control station requires high bandwidth communication channel. The problem of high bandwidth for visual data can be overcome by using a computer at the remote site to compress data using data compression algorithms. The compressed data is send to the control station where the same data compression algorithm is used to decompress the data from the remote site. The decompressed data is fed to the television located at the control station and the human operator gets a visual feedback of the remote environment.

Then there are minor problems in remote television visual sensing like possibility of operator disorientation due unnatural spatial relationships between the camera and the remote manipulator, limited resolution and sensitivity to radiation. The problem of operator disorientation can be overcome by implementing the 3D viewing of the remote environment using more than one cameras placed at different angles with respect to the teleoperator. The two cameras sense the environment at different angles and the signals from each camera is mixed using a mixer to get a three dimensional perception of the work environment. At the control station 3D televisions are used to get the three dimensional view of the remote work environment.

Force feedback Sensors

Next to vision force feedback is the most important type of feedback information needed in space teleoperation. Cables and servomotors force slave arms to follow master arm and the vice versa. Cables and servomotors act as the primary force-sensing device in the space teleoperation. The slave arms encounters some force, when it pushes against some objects or grasps some objects. When the slave arm the servomotors attached to encounters this force the slave arm generates signals that corresponds to the force encountered in the remote manipulator. This signal is transmitted to the control station. At the control station a similar servomotor develops a corresponding motion as that encountered in the remote manipulator. This motion is reflected in the force reflecting master arm and the human operator feels the same force encountered in the slave arm.

Another type of force feedback device that should be used in space teleoperation is the tactile force feedback device. The tactile force sensor senses the tactile force that are encountered in the remote manipulator and send it back to the master arm and human operator. Tactile feedback gives a sense of how big a force is applied to an object while gripping it, by the remote manipulator. It also gives a feedback on the shape and texture of the target object being manipulated. The tactile sensors that can be implemented in the remote manipulator are an array of tiny force transducers distributed densely over the teleoperator hands and fingers. The force transducers are piezoelectric crystals, which emit electric current when pressure is applied over them. The electric signals, which are emitted, depend on the shape of the object and the force being applied on the object. These signals are fed back to the control station. The tactile force can be displayed on a screen or provided on the hand of the human operator through the master control by the help of electric devices like capacitors and resistors. These devices generate small amount of electric shock, which will correspond to the force being applied on the object being manipulated by the teleoperator. NASA has used data-gloves for tactile feedback displays in space teleoperation.

Time Delay in Force Feedback

Due to time delay of the signals transmitted the feedback is obtained with a time delay. Time delay is caused by speed limitation of radio waves, processing time for on board and task interactive computers. In teleoperation in low earth orbit when the teleoperation is carried out in close proximity to the human operator the problem of time delay is not there. Time delay becomes a major problem in deep space and interplanetary operations. The roundtrip delay [1] causes even the smallest task to be finished in a long time. This happens because of the need to avoid instability [1] the human operator has to adopt a strategy in which the human operator gives a command and then wait for the feedback to arrive and then again give the next command. This causes the implementation of difficult tasks more time consuming.

The problem of time delay can be overcome with the help of predictor displays [1]. In space teleoperation the teleoperator performance can be enhanced if the human operator has some sort of display that estimates the consequences of his actions. Predictor displays look ahead in time by constructing models of the situation. This model runs faster than real time model and its performance are displayed to the human operator. This model is similar to the mental model [5] of the human operator.

As the actual feedback comes this can be given as the feedback to the operator. The actual feedback given to the human operator after it arrives can cause disturbance in the feedback displays. This disturbance is not a major problem in visual feedback that the display of this feedback is not given physically to the human operator but it becomes a major problem in force feedback, which is given directly to the human operator�s hand. The delayed feedback will cause an unexpected disturbance on the operator�s hand, which he cannot ignore and this causes instability in the teleoperator system as a whole. The solutions suggested for the disturbance effect is as follows,

1.Have the force feedback displayed visually on the television used for visual feedback.
2.Provide the time delayed force feedback to the hand that is not operating the master.

From the diagram it can be seen that the both feedback can cause problem due to dual feedback channels and instability. The predictor displays work well in free space where it is easy to predict what will happen in future. In the case of tasks involving assembling and tasks in other unstructured environments it is very difficult to predict what will happen in future time delay will have to be sacrificed for accuracy. The purpose of predictor display is not only to overcome time delay problems but also to warn the human operator of the future consequences of his actions that he might not anticipate.

The complete sensory subsystem can be modeled as in the following diagram,

The dotted lines in the diagram represent predictor displays and double-headed arrow represents two-way communication.

Control Subsystem

Control system includes man in loop. This subsystem can be considered as the brain of the teleoperator system. This subsystem is responsible for the control and manipulation of other subsystems. The control subsystem is further subdivided into subsystems, like communication subsystem, and computer subsystem [4]. So in the discussion of control subsystem the system is divided into two viz., communication and computer subsystems and discussed individually.

1. Communication Subsystem

Noise Interference

The communication using radio waves can cause a lot of noise to interfere in with information signal being passed. This will cause distortion of information being passed and the noise can�t be eliminated in any communication system. The effect of noise can be understood more clearly with following diagram,

Where �a� is the signal transmitted and �� � is the noise that interferes with the system. It is clear that the signal send by the control station and that received by the remote station is the same. To over come this difficulty the system should have an error control loop, which should always undertake error correction. The error control loop will eliminate error by measuring the error that is introduced in the communication channel indirectly. The idea used in error control loop can be modeled as in the following diagram.

The control subsystem sends a signal �a� through the communication link. The actuator receives the signal with noise added to it. The actuator then reports its status, (a+� ) back to the control subsystem. The control subsystem receives the signal with noise added to it. The signal received at the control station will be (a + 2� ). Using computer algorithms the noise, � can be found out as the noise, �a� is known to the control subsystem. Now the control station can send signals by compensating for the noise. This solution to the noise correction problem assumes that the noise, � will remain the same throughout the communication process, but in reality this is not the case. So this solution will reduce the effect of large amount of variations even though it will not be able to completely eliminate the effect of noise.

b.) Computer Subsystem

In teleoperation in space a general-purpose computer will be desirable due to the following functions.

1.Data compression and processing.
2.Lengthy computations
3.Preview and supervisory control.
4.The generation of virtual displays when visual displays is impossible.
5.Forecasting the outcome of specific actions, as in predictor displays.

In teleoperation in space a large general-purpose computer is preferred over number of small, local, digital computers associated with sundry subsystem functions. Teleoperators in space will have thermostats, which control the temperature inside the teleoperator. It will also have some voltage stabilizers, and power regulators. It is difficult to have computers controlling each of these systems, and these computers will not have the flexibility and power of the general-purpose computer.

The teleoperator computer can be a digital computer or analog computer. But analog computers are preferred only for specialized applications. But when memory and flexibility becomes major factors digital computer are preferred over analog computers.

The physical location of the computer depends on the application. In reality the teleoperator will have two or more computers in them to act as back up. On a distant planet the teleoperator will need a local general-purpose computer for supervisory control and data compression prior to transmission. The operator on the earth or space craft will need another computer to for predictor displays and preview control because of the time delays that will be there in deep space teleoperation. This helps to overcome the difficulty of low bandwidth that is available for communication. Many small purpose digital computers have been constructed for manned space flight program. The teleoperator computer technology that is needed for space teleoperation in future can be built on this base.

Physical Subsystem

The physical subsystem helps in maintaining the teleoperator in position and supply power for various manipulative activities. Mobility is the essential attribute for the successful teleoperation. Small thrusters that are built into the remote robot provide mobility to the teleoperator. In order to move from one place to another these thrusters are used. The power that is used for providing this thrust can be derived from chemical energy, nuclear energy, etc, depending on the teleoperation tasks. If the robot is used in deep space applications then nuclear energy is the feasible option, as nuclear energy will last longer in flight than any other forms of energy. But in low earth orbits due to dangers associated with the nuclear energy it is not used.

The manipulation and sensing activities electric current to be carried out. In space teleoperation power can be produced directly from solar energy using solar cells. In deep space teleoperation where solar energy becomes less and less powerful due to dissipation, nuclear energy and fuel cells is used for electric power generation.

All electronic gadgets used in the remote station have to be protected from extreme climates of the space to function properly. The major factor that has to be controlled is temperature. The physical subsystem should have means of heat rejection in space. Heat rejection is achieved by using an evaporator. Radiator can also but the weight of the radiator equipment will cause the launch economy [1] to be reduced and the obvious choice is to use evaporator.

The physical subsystem can be modeled as in the following diagram.

Teleoperator for Space Application

Teleoperation in space and for that matter in any environment requires the system to be built from subsystems. The various subsystems that are required for the teleoperation in space have been discussed and their design requirements and problems were also discussed. Now the subsystems have to be integrated into a single system.

The actuator subsystem, control subsystem and sensory subsystem span both the control station and remote station. The physical subsystem is implemented only in the remote station. The design of each subsystem integrated to from the complete teleoperator system that is used in space can be schematically shown as in the following diagram.

Conclusion

The teleoperator system needed for space teleoperation has been designed by dividing the whole system into several subsystems like actuator subsystem, sensory subsystem, control subsystem, and physical subsystem. Each subsystem is designed individually and after that each designed is integrated into a whole system. The subsystems needed for teleoperation in space is tabulated as follows.
�

The subsystems will have design problems in them each problem that is encountered and the solution suggested is as follows,
�
�
�

The teleoperator subsystems have been designed and the problems associated with these designs have been discussed. Some solutions to these problems have been discussed. The future of space travel and exploitation of other planets lies in the successful development of this field of robotics.

Reference:

1.Skarr, B.S., & Ruoff, F.C. (1994) Teleoperation and Robotics in Space, American Institute of Aeronautics and Astronautics, Inc, Washington, DC, USA.
2.Kim, S.W., (1993) Telemanipulator Technology and Space Telerobotics, SPIE The International Society for Optical Engineering, Washington D.C., USA
3.Wolczek, O (1974) Unmanned Exploration of the Solar System: A Critical Review and Recommendation for the Future, Proceedings of the XXV^th International Astronautical Congress, Amsterdam, Netherlands.
4.Johnsen, E.G, & Corliss, W.R (1975) Human Factors Applications in Teleoperator Design and Operation, Wiley-Interscience, New-York, USA.
5.Sheridan, B.T. (1992) Telerobotics Automation and Human Supervisory Control, MIT Press, Cambridge, Massachusetts, USA.
6.Rahimi, M., & Karwowski, W. (1992) Human-Robot Interaction, Taylor and Francis, London, UK.
7.Webster, G.J (1998) Tactile Sensors for Robotics and Medicine, Wiley-Interscience, New York, USA.
8.Butler, G.V. (1981) Working in Space, American Institute of Aeronautics and Astronautics, New York, USA.
9.Wagenr-Bartak, C.G., Matthews, P.S., & Hill, B.R. (1983) Remote Manipulators in Space, Proceedings of Winter Annual Meeting of the American Society of Mechanical Engineers, Boston, USA
10.Krasnoff, B. (1982) Robots: Reel to Real, Arco Publishing, Inc., New York, USA.
11.Bejczy, A.K., & Corkeer, K (1984) Manual Control Communication in Space Teleoperation, Proceedings of RoManSy �84, MIT Press, Cambridge, Massachusetts, USA.
12.Tachi, S., Tanie, K., Komoriya, K., & Kaneko, M (1984) Teleexistence (I) Design and Evaluation of a Visual Display with Sensation of Presence, Proceedings of RoManSy �84, MIT Press, Cambridge, Massachusetts, USA.
13.Bicker, R. & Maunder, L (1987) Force Feedback in Telemanipulators, Proceedings of the Sixth CISM-IFToMM Symposium on Theory and Practice of Robots and Manipulators, MIT Press, Cambridge, Massachusetts, USA.
14.Heer, E., & Lum, H (1988) Machine Intelligence and Autonomy for Aerospace Systems, American Institute of Aeronautics and Astronautics Inc., Washington DC., USA
15.Wilson, J.W., Miller, J., Konradi, A. & Cucinotta, F.A (1997) Shielding Strategies for Human Space Exploration, NASA Conference Publication 3360, NASA, Hampton, USA.
16.Crouse, K.J. (1985) Vision Requirements for Space Station Applications, Proceedings of SPIE Volume 580, SPIE, Washington, USA.
17. http://ranier.hq.nasa.gov/telerobotics
18. http://www.aist.go.jp/ETL/~5822/distance.html
19. http://ranier.oact.hq.nasa.gov/telerobotics_page/technologies.html
20. http://rmet.jpl.nasa.gov/rmet/telerob/telerob.html
21. http://gypsy.rose.utoronto.ca/projects/telerobotics.html
22. http://www.spar.ca/space/telrbtcs.htm