From Project Persephone

PmWiki: Telebots

Tele-operated robots -- small, mobile, camera-equipped, untethered devices that can monitor and manipulate objects in their environments. The next best thing to being (up) there.

Origin of the term

"Telebot" makes its first appearance in the popular press in 1990 in reference to a remote-controlled submarine salvaging cargo from deep-ocean shipwrecks.1 |

The OpenROV project aims to bring marine telepresence? to the masses

The term has also been used to describe a mobile telepresence? technology for the disabled.2,3 Even in that latter sense, however, the meaning still aligns closely with the Project Persephone sense: the high costs and danger of human space travel (and of deep-seabed exploration) may be thought of as a kind of mobility impairment.

The term telebot was chosen over "robot" to help keep those who are new to Project Persephone from thinking too much in terms of autonomous robotics. At the same time, the "bot" suffix of "telebot" still invokes a sense of something zoomorphic -- able to move, do things, sense things.


Ecosystems could grow telebot limbs and tendons

Project Persephone proposes giving people near-real-time access to telebots in Earth orbit, operating inside (and possibly outside) exovivaria. Many of the imagined uses would be purely utilitarian -- maintenance of exovivarium structure and centralized equipment, "harvesting" of biomass for processing into materials usable for maintenance and repair, and using telebots to repair other telebots. However, many other telebots could have purely recreational uses for paying guests -- exploring the ecosystem, multi-player gaming, arts and crafts, performance art. The boundary between utilitarian work and hobbyist effort is likely to be blurry in some cases -- would using a telebot to decorate another telebot be part of "maintenance" or would it be just for fun?

Ruggedizing at least some of these telebots to survive very high accelerations is one of their design challenges, because Project Persephone also proposes developing equatorial mountain regions to host projectile space launch for long-run reduction of launch costs. This ruggedization challenge may seem insuperable, but in fact there is plenty of evidence that it's far from the hardest problem; see "Practicality of gun-launched telebots", below.

Designing these telebots so that they could build and repair other telebots is yet another design challenge. In this, the Project is somewhat inspired by the "robot ecosystem" designs for self-repairing and even self-replicating systems on the Moon and other planets.4,5,6,7 Exovivaria will grow plant stems and insect exoskeletons that could be used for telebot limbs and other structural elements; cabling may be made from grown plant fibers. The exovivarium payload as launched may therefore include only enough folded-up, ruggedized "bootstrap" telebots to start the basic ecosystem. The remaining complement of telebots may be manufactured, together with the biologically-derived components, from a basic set of more-easily-ruggedized joints, sensors and manipulators that would consume relative little payload mass and volume.

One of the goals of exovivaria design is to make them net-negative orbital debris generators, if only because they collect orbital debris on impact, or slow it down to some extent if the debris passes through. But a debris strike can create secondary debris, and exovivaria would be no exception. In fact, given their high surface-area-to-mass ratio compared to most other space assets, they'd seem to be generating more than their share of secondary debris. Furthermore, telebotics complicates this goal insofar as robotics may require components that are large and/or dense. The ideal secondary debris from exovivaria would be small, low in vaporization temperature (in the heating of debris strikes), low in ballistic coefficient (for rapid deorbiting under LEO drag forces), and high in its rate of deterioration under raw sunlight and atomic oxygen exposure. Making telerobotics small is a first approximation. But more is possible.

The use of biological components should also be helpful in meeting this net-negative orbital debris goal, because biologicals are mostly water. If exposed to the orbital environment, they will dessicate quickly in vacuum, reducing their ballistic coefficient. The heating from impact stresses may start and perhaps even mostly complete the evaporation process. There would remain telerobotic components that can't be biological. However, PV cells can be made very thin, and so can electronic components. Lenses for digital imaging are a challenge, but this challenge could be met by using pinhole-camera optics. Pinhole optics has some serious drawbacks, though these may be met by combining images from multiple cameras. However, lens technology continues to improve.

A metamaterial approach to optics has yielded a thin (thus low-ballistic-coefficient) lens made of silicon nitride whose size has been compared to that of a grain of salt.8 Silicon nitride is hard (outperforming most metal for ball bearings) and with a density falling between aluminum and titanium; it's not going to degrade much under raw sunlight UV or atomic oxygen. Still, it ticks some other boxes: in metamaterial optics, it's flat (low ballistic coefficient) and can apparently be very small.

Practicality of Earth-based teleoperation

The ETS-VIII ("Kiku") demonstrated telebotic orbital construction of antennae

Using ground-based telerobotics for space-based operation, assembly, disassembly and repair has been proven in tests by several nations' space programs. For example, in the late 1990's, on the Japanese satellite ETS-VIII, a microwave antenna featuring a dish with snap-together sections was teleoperatively assembled and disassembled about 50 times, from a university laboratory in Japan.9 The previous ERTS-VII robot arm experiments starting in 1997 and continuing for 5 years had shown that a robot arm could be used in multi-satellite rendezvous and docking (all the more impressive for docking autonomously.)10,11

Canadarm2 has been operated from the ground,12 albeit only by sending sequences of commands, and only for applications like positioning cameras to record an EVA -- it was designed for real-time operation from ISS. In 2008, Dextre (SPDM), operable from the ground, was launched and outfitted for use outside ISS. There are plans to use Dextre in satellite repair, including "satellite surgery", in which the arm wields attached tools to cut through outer layers of a satellite to reach a fuel tank, refueling satellites that were never designed to be refueled.13

Kibo, a Japanese experiment module installed on ISS, features JEMREM, with a major arm and a "Small Fine Arm" (possibly operable from Earth -- the ratio of uplink to downlink speeds, and extensive Japanese experience with on-orbit teleoperation, suggests it was designed for that, not just for operation on ISS.)

Germany has produced two robot arms teleoperated from the ground: ROTEX in 1993, inside a SPACELAB module on a Shuttle mission, and ROKVISS,14,15 a telerobotic arm installed on the Russian section of ISS,16,17 which started operations in 2005. Human operators on the ground were able to use ROTEX to catch a spinning cube, using a predictive display to defeat speed-of-light delays. ROKVISS was used to test very lightweight joints used in teleoperable four-fingered hands; it had a stereocam mounted and was operated when the Shuttle's orbit took it over Germany.18

NASA? is planning a Lunar Surface Manipulator System that could be controlled from Earth.19 NASA is also testing portable telesurgery units underwater, to simulate microgravity.20

Practicality of gun-launched telebots

Small insect-like robots made of composite materials have survived falls of up to 28 meters21. Piezo-electric actuators and shape-memory alloys combined with composite-material hinges can make possible a small telebot with very low mechanical complexity that folds into a package able to survive very high accelerations.22

Limited Autonomy

Exovivaria telebots may feature some limited autonomy. If, for example, a telebot is legged, and is sometimes expected to travel some significant but unobstructed distance within an exovivarium, it would probably make sense to equip it with the ability to respond to a "walk 35 cm" command rather than require individual human commands for each leg movement. Short-term task-teaching by supplying sample human movement may be useful in a pinch23 or when certain simple and repetitive operations that are not system-critical simply become too boring or unproductive to control directly. For low-cost experimentation with limited autonomy, relatively cheap hardware, sold as toys, may be used directly or repurposed. Humanoid robot toys have already been marketed with the ability to act upon very simple voice commands, such as the i-SOBOT.

However, to make telebots significantly autonomous is not a Project Persephone goal. Quite the contrary, if anything. On-orbit robot autonomy is somewhat in conflict with the goals of creating (1) new recreational experiences and (2) useful paid work in support those experiences. Robotic autonomy should perhaps be considered only for automating emergency repair tasks that are required to re-establish communications in the event that on-orbit comm equipment fails with no effective backup channel.


1 Patricia Brennan, "Down to the Deep for a Treasure in Gold", The Washington Post, September 9, 1990

2 "iGrid '98: The Ganymede Telebot, an Enabling Technology for Teleconferencing", STAR TAP: Electronic Visualization Laboratory, 1998

3 Maggie Rawlings, Paul Rossman, Greg Dawe, Alan Cruz, Javier Girado, Jason Leigh, Dan Sandin, "The Ganymede Accessbot: an Enabling Technology for Teleconferencing", Electronic Visualization Laboratory, 1998

4 Ralph Merkle, "NASA and self-replicating systems: implications for the Moon". Appendix I of The Moon: resources, future development, and settlement, David G. Schrunk, Springer, 2008 ISBN 0387360557

5 "Robosphere: Self-Sustaining Robotic Ecologies as Precursors to Human Planetary Exploration", Sylvano Colombano, AIAA

6 "Shape-Shifting Robot Nanotech Swarms on Mars", NASA?

7 Freeman J. Dyson, "The twenty-first century," Vanuxem Lecture delivered at Princeton University, 26 February 1970

8 "Researchers shrink camera to the size of a salt grain",, 29 Nov 2021

9 "Space Experiment of Advanced Robotic Hand System on ETS-VIII", Kazuo Machida (Electrotech. Lab., Agency of Ind. Sci. and Technol.), Yoshitsugu Toda (Electrotech. Lab., Agency of Ind. Sci. and Technol.), Hirotaka Nishida (Fujitsu Ltd.), Masao Kobayashi (Fujitsu Ltd.), Kenzo Akita (USEF), in Proceedings of the Space Sciences and Technology Conference v.42, pp. 101-106 (1998). In Japanese.

10 "Antenna Assembling Experiment on ETS-VII. 2: Results of Basic Experiments", Shin'ichi Kimura (Comm. Research Lab.) Toshiyuki Okuyama (Comm. Research Lab.), Shigeru Tsuchiya (Comm. Research Lab.), Yasufumi Nagai (Univ. of Electro-Communications), Nobuaki Takanashi (NEC), Hajime Morikawa (NEC), Kazuo Nakamura (NEC), in Proceedings of the Space Sciences and Technology Conference v.42, pp. 107-112(1998). In Japanese

11 Dr. Mitsushige Oda, Principal Investigator, "Engineering Test Satellite VII Home Page", JAXA, 2002

12 Bjorn Carey, "Remote Access: Canadarm 2 Gets a Hand From Ground Control",, 4 May 2005

13 Debra Werner, NASA Plans To Refuel Mock Satellite at the Space Station, Space News, 2 April, 2010

14 "ROKVISS" (German), DLR: German Aerospace Center

15 "ROCKVISS" (English), DLR: German Aerospace Center

16 "ROKVISS" (Russian), Energia?

17 "ROKVISS" (English), Energia?

18 Bekey, George A., et al., Robotics: State of the Art and Future Challenges, Imperial College Press, 2008. ISBN 1848160062

19 Taylor Dinerman, "Langley's lunar manipulator: a big Swiss Army knife for planetary operations", The Space Review, July 27, 2009

20 "NASA to Test Portable Robot Surgeon",, April 19, 2007

21 Chris Jablonski, "Resilient cockroach-inspired robot survives large falls, dashes off", Emerging Tech, ZDNet, Oct 13, 2009

22 "Prototyping Folded Robots", Biomimetic Systems Lab^, U.C. Berkeley

23 "Human demonstration data for fast task teaching", Okodi, Samuel; Xin Jiang; Konno, A.; Uchiyama, M.; Intelligent Robots and Systems, 2008. IROS 2008. IEEE/RSJ International Conference on 22-26 Sept. 2008 Page(s):961 - 966

Further reading

Retrieved from
Page last modified on November 30, 2021, at 05:05 AM