More autonomous exploration and mapping of abandoned mines
Bryan Bishop
http://www.cs.cmu.edu/~3D/mines/index.html
http://www.cs.cmu.edu/~3D/
Here's some of the animated results of the robot's exploration:
http://www.cs.cmu.edu/~3D/mines/html/map0.html
This included the production of an open source project:
"Carmen Robot Navigation Toolkit"
http://carmen.sourceforge.net/
"Welcome to CARMEN, the Carnegie Mellon Robot Navigation Toolkit.
CARMEN is an open-source collection of software for mobile robot
control. CARMEN is modular software designed to provide basic
navigation primatives including: base and sensor control, logging,
obstacle avoidance, localization, path planning, and mapping."
Related paper-
Autonomous exploration and mapping of abandoned mines
http://www.ri.cmu.edu/pubs/pub_5317.html
http://www.ri.cmu.edu/pub_files/pub4/thrun_sebastian_2004_1/thrun_sebastian_2004_1.pdf
"Abandoned mines pose significant threats to society, yet a large
fraction of them lack accurate maps. This article discusses the
software architecture of an autonomous robotic system designed to
explore and map abandoned mines. A new set of software tools is
presented, enabling robots to acquire maps of unprecedented size and
accuracy. On 30 May 2003, our robot "Groundhog" successfully explored
and mapped a main corridor of the abandoned Mathies mine near
Courtney, Pennsylvania. This article also discusses some of the
challenges that arise in the subterranean environments and some the
difficulties of building truly autonomous robots."
As you tunnel deeper and deeper into the old mining shafts and
pathways, you easily lose wireless connectivity and communication
starts to become a real pain, so there has to be some protocols for
backtracing after a certain amount of time for a data uplink, or
having a big batch of pingpong balls retrofitted with relay comm
equipment (which the CMU group didn't try).
I have to admit that I've been slacking in my bookmarks and references
to the robotics scene on the web, knowing full well that it's been
doing more than great for the past few years, so I know that there's
been some good updates since that 2003 robot miner even before people
started tossing around the 'open hardware' identifier.
Anyway, I went back and found the abandoned mines websites --
http://www.abandonedmines.gov/
"There are estimates of as many as 500,000 abandoned mines in our
nation." And I'm not going to count the number of abandoned or "at
risk" mines elsewhere in the world. These serve as easy openings into
the mineral contents stored in the geographies, and are largely
unwanted, especially because of human health concerns. With those
500,000 abandoned mines, and if you wanted at least two or three
robots at each, you're looking at a few million robots doing dirty
tasks unfit for human. Besides the ridiculously large technical
support infrastructure needed at each location, I can easily imagine
using plastics and metal machining equipment on site, such as a
rentable "mobile fablab" of sorts, being used to enhance or improve a
basic robotic framework for first exploring and mapping, and then
continuing where previous miners left off. Clearly for some mines,
certain mining instruments and tools are more appropriate than others,
etc. So there's a lot of fun things that could be done here. The
energy requirements for these mobile robots would be a bit staggering
because they wouldn't be able to stay down all day and night mining
away, but instead would be forced to resurface for battery exchanges.
Not terrible. Haven't done the calculations, but I don't know if
photovoltaics would be able to keep up with the energy requirements of
even one robot mining all day. Electric car batteries these days are
doing 50 mile trips, but I don't hear of them being recharged by solar
cells quite yet.
There's also an online tool for checking out mines via the Land Survey
Information System, interactive map:
http://www.geocommunicator.gov/LSIS6/map.jsp
http://www.geocommunicator.gov/GeoComm/index.shtm
I guess they're calling it the "GeoCommunicator". There's also this one:
http://nationalatlas.gov/natlas/Natlasstart.asp
which I find slightly more usable. Click on "mineral operations" on
the right, and then select all of them, redraw the map and you'll see
some colorful information.
There's also more information here:
http://minerals.usgs.gov/minerals/pubs/mapdata/
but this is just PDFs without the raw data in a txt file or something.
But there might be something usable here:
http://mrdata.usgs.gov/
Like the "Material Resource Data System":
http://tin.er.usgs.gov/mrds/
"MRDS describes metallic and nonmetallic mineral resources throughout
the world. Included are deposit name, location, commodity, deposit
description, geologic characteristics, production, reserves,
resources, and references. It includes the original MRDS and MAS/MILS
data. MRDS is large and complex. This service provides a subset of the
database comprised of those data fields deemed most useful and which
most frequently contain some information."
Here's the 30 MB zip file:
http://tin.er.usgs.gov/mrds/mrds.zip
And there's documentation on the previous link.
And of course there's the classic http://mindat.org/ which overlays
its own data set on mining on top of Google Maps.
Through using robots made with money from research grants to jumpstart
materials into a free system, those materials can be used to produce
more robots for further mining operations, although not necessarily
the entire industrial ecosystem needed to manufacture semiconductors,
not at first. Actually, for computation that does not require
spur-of-the-moment decisions, analog control systems can be designed.
But an analog circuit of reasoning on data input is not going to be
able to reconfigure easily. This was pre-digital era tech. Feeding
optical data to capacitors and resistors for processing is very old
school. The robots would be pretty slow at crunching numbers and
wouldn't be able to respond to a large number of situations, but they
would also not require integrated microchip manufacture. Kind of a
crappy tradeoff.
Nathan Cravens
Ideally, chemically engineering metals would be a more preferred, environmental friendly method, and serve the functions of the end user without hefty, time consuming ($$$), distribution channels.
Nathan
Bryan Bishop
> If wireless energy transfer is not possible, 'battery bots' can go in and
> out of the mine specifically to 'hot swap' battery packs from mining units.
Good idea. Of course, some old mines still have electrical wiring for
lighting, but I don't see why you couldn't lay down electrical cabling
for the bots anyway. In the case of a collapse, battery power and an
'oh crap' mode would be great.
> Ideally, chemically engineering metals would be a more preferred,
> environmental friendly method, and serve the functions of the end user
> without hefty, time consuming ($$$), distribution channels.
What? I do not understand.
Nathan Cravens
> Ideally, chemically engineering metals would be a more preferred,
> environmental friendly method, and serve the functions of the end userWhat? I do not understand.
> without hefty, time consuming ($$$), distribution channels.
Turn simple chemical compounds into metals in a test tube, then scale up. This would involve reverse engineering of the chemical structure of a metal.
Nathan
Bryan Bishop
Most metals are mineral elements.
003 Lithium
004 Beryllium
011 Sodium
012 Magnesium
013 Aluminium
019 Potassium
020 Calcium
022 Titanium
023 Vanadium
024 Chromium
025 Manganese
027 Cobalt
028 Nickel
029 Copper
030 Zinc
033 Arsenic
040 Zirconium
042 Molybdenium
047 Silver
048 Cadmium
051 Antimony
056 Barium
076 Osmium
078 Platinum
080 Mercury
081 Thallium
082 Lead
092 Uranium
- Bryan
Nathan Cravens
Its a matter of building from simpler compounds found anywhere to create metals. Is this being done?
Nathan
Bryan Bishop
> Its a matter of building from simpler compounds found anywhere to create
> metals. Is this being done?
http://en.wikipedia.org/wiki/Biomining
"Biomining is a new approach to the extraction of desired minerals
from ores being explored by the mining industry in the past few years.
Microorganisms are used to leach out the minerals, rather than the
traditional methods of extreme heat or toxic chemicals, which have a
deleterious effect on the environment.
Using a bacterium such as Thiobacillus ferrooxidans to leach copper
from mine tailings has improved recovery rates and reduced operating
costs. Moreover, it permits extraction from low grade ores - an
important consideration in the face of the depletion of high grade
ores.
The potential applications of biotechnology to mining and processing
are countless. Some examples of past projects in biotechnology include
a biologically assisted in situ mining program, biodegradation
methods, passive bioremediation of acid rock drainage, and bioleaching
of ores and concentrates. This research often results in technology
implementation for greater efficiency and productivity or novel
solutions to complex problems. Additional capabilities include the
bioleaching of metals from sulfide materials, phosphate ore
bioprocessing, and the bioconcentration of metals from solutions. One
project recently under investigation is the use of biological methods
for the reduction of sulfur in coal-cleaning applications. From in
situ mining to mineral processing and treatment technology,
biotechnology provides innovative and cost-effective industry
solutions."
http://www.niac.usra.edu/files/library/meetings/fellows/mar04/Ragozzine_Darin.pdf
http://en.wikipedia.org/wiki/Bioleaching
"Bioleaching is the extraction of specific metals from their ores
through the use of bacteria. Bioleaching is one of several
applications within biohydrometallurgy and several methods are used to
recover copper, zinc, lead, arsenic, antimony, nickel, molybdenum,
gold, and cobalt."
This is done sometimes with waste water. Requires large scales, but
yeah, it's being done, though I'm not aware of "bioleached metals"
being specifically available on the market.
Eric Hunting
been giving a lot of thought to in the context of space development
and have been working on some articles for recently.
In TMP2 I discuss how, like much of the rest of space colonization,
lunar and planetary settlement cannot effectively be premised on
traditional colonial economic models given the inability, within the
limits of any known and currently anticipated near-term propulsion
technology, to make the bulk export of materials from space to Earth
practical. The Moon and near-Earth asteroids may present a better
situation in this regard given low energy overheads in transit and the
prospect of development -within a century or so- of a robust
terrestrial space elevator system but locations like Mars, most
certainly, will not be 'exploitable' in any traditional colonial
fashion any time soon. The great potential 'profits' possible in space
are only to be realized by people going to live there. Thus space
development must ultimately be premised on a cultural imperative
rather than any economic incentive, becoming the province of those
rare people imaginative -or delusional- enough to look at pictures of
the desolate surface of Mars and actually envision some version of the
Good Life. (which is why my theme music set for Life On Mars in the
TMP2 music lists features a lot of Sami music... it sounds more like
the place) Given this situation the notion of a continuously manned
early phase of lunar and planetary settlement becomes untenable. The
development of a resource and industrial infrastructure capable of
supporting large communities at a high standard of living and safety
takes a very protracted period of time that could never be sustained
at the costs of manned space flight. Though robotics may not be
functionally competitive with human anatomy for some time, it has
distinct advantages in lower support overhead and disposability, with
no minimum in potential system scale and with the costs of unmanned
space flight declining at a far faster rate than for manned space
flight. Astronauts can certainly save time but time is a hell of a lot
cheaper than manned space flight. The most-likely strategy, therefore,
that offers any hope of bringing colonization down to a level
perusable for a relatively modest scale community with the will for it
is one based on both telerobotic pre-settlement and the most
functional use of immediate natural resources for settlement
development in their least processed forms -the equivalent of building
log cabins from the trees at-hand.
Thus I've been giving much thought to the design of telerobotic
outposts, their functional logistics, lowest-cost deployment strategy,
the relationship between telecommunications latency and local systems
intelligence, and the application of small scale robotics to the basic
tasks of wide area resource assay, self-maintenance, habitat
excavation, mining, construction, etc. Basically, the collective
technology that would make it feasible for a relatively small number
of people -a science foundation, a small business, a colonization
advocacy group, or even just one particularly affluent individual
interested in the world's most unique and grand hobby- to remotely
conduct all the preliminary infrastructure development and habitat
construction necessary to establish permanent -and comfortable- human
habitation. The best model train layout ever. The one you can
eventually retire to. And what's great about the concept is that this
is something very open to experimentation and exploration by the
hobbyist robotics community. We can build completely functional
prototype telerobotic outposts on Earth and give space advocacy
something practical to do beside fantasizing and lobbying.
The two keys to this seem to be networking and interusability. Network
communications define the architecture of the telerobotic outpost
while interusability -the ability for a common set of components to be
interchangeable across systems serving many different applications-
affords efficiency in maintenance, economy in systems development and
deployment, and mission survivability based on free cannibalization or
re-purposing of systems. I've envisioned a multi-layered self-managing
but largely passive (as far as individual robotic system are
concerned) IP network infrastructure based on several transmission
technologies and several systems platforms including centralized
surface communications, computing, and command modules (primary Earth
up-link, command systems 'sequencers', and bulk data storage/
processing), distributed self-sufficient transponder units (long and
short-haul WiFi with peer-to-peer relay, low-bandwidth Earth uplink,
surface GPS, asynchronous packet store-and-forward, weather
monitoring, beacon tracking), point-to-point relay stations,
satellites (surface, interorbit, and Earth relay as well as GPS and
radar tracking), modular cable systems with parasite WiFi nodes, piggy-
back networking with RF subcarriers in modular power cable systems, on-
board peer-to-peer networking (robots become passive short-haul relay
nodes in their own WiFi net web), and on-board tracking beacon
systems. So basically the outpost is, collectively, a multi-layered
peer-to-peer web with no apparent hierarchy. Every system node on-line
in one point is accessible to any other in another other point and, of
course, it all funnels back to technicians on Earth as though they
were accessing a VPN where all the systems are laid out as a
hierarchical web of direct systems interfaces, command-level
interfaces, and higher level sequencer interfaces (sequencers would be
programs that control collections of distributed systems). This is a
concept that derives from the idea of the MUOL -modular unmanned
orbital laboratory- which is basically an orbital distributed
automation platform based on web controllers integrated by an IP
backplane supporting multiple service client VPNs -put simply, an
Internet you plug robots into.
If one were to design a platform for total automation of mining today,
it would probably look very much like this, just in a much smaller
area with maintenance more human-dependent. But this would still be
rather sophisticated for the average person. Comprehensive waste
recycling technologies may offer more prospects of personal resource
gathering on an 'appliance' level. However, I have given thought to
the prospects of nanotechnology in mining, which would be one of its
early practical large scale applications and which would parallel many
characteristics associated with biomining.
Nanomining would be extremely simple in management because of the
reliance on a fluid transport medium for all the actual work -just as
with chemical leach-mining or based on bacterial bio-leaching would.
And what makes it an early nanotechnology application is that it would
be based on simpler application-specific disassemblers needing little
control. What this would basically look like in old mines is a sealing-
up of 'work face' sections with portable bulkheads then filling the
gap space with working mediums and injecting disassemblers into it
while using NanoChip 'sorter' ports to remove refined molecular
material from that fluid and package it as feedstock fluid packed in
drums. For new mining, the approach would be akin to leach-mining
through the use of injection well grids with inserted nanoprocessing
well caps. Insert well cap with a metal probe and shallow collar seal,
deploy PV 'solar flower', inject disassemblers, bottle product and
tailings, reinject tailings as extruded granular precipitate, pull up
stakes and move on to the next grid point. This reduces mining to a
one-man operation -and he need never even get his hands dirty. And the
environmental impact, though not insignificant, of course, would be
low. The most effective form of Nanomining, however, would be with
somewhat later technology combining selective molecular disassembly
with diamondoid scaffold assembly and a self-managed root system. Here
your 'well head' extrudes a root structure with active sensing
capability and a communications network distributed along its root
structure. The roots are formed of a fluid permeable scaffold of
diamondoid with a regular geometry that eliminates the problem of
subsidence and is also used to actively sequester tailings as desired
materials are pumped out of the root system. It would also be able to
deploy other roots as a heat sink to increase penetration depth and
even use latent geothermal heat as an energy source. With this
technology a single well head of any modest size could extract
material from an effectively unlimited subterranean area to a depth
where it can no longer effectively unload latent heat fast enough and
with no disturbance to the surface environment. This would not only be
ideal for terrestrial mining but also for asteroids, eliminating the
issue of dealing with free-floating granular material in a
microgravity environment. And it could also become the basis of a
primary source of renewable energy through vast self-constructing
thermocouple networks. Automated excavation would also be possible by
this method using more active control of the direction of root
deployment and the selective disassembly of scaffold structure to make
open tunnels and chambers. This would make the construction of globe-
spanning subways systems and pipelines effectively labor and cost
free, though perhaps not especially faster than conventional
excavation methods.
Eric Hunting
erich...@gmail.com
Bryan Bishop
> fashion any time soon. The great potential 'profits' possible in space
> are only to be realized by people going to live there. Thus space
> development must ultimately be premised on a cultural imperative
> rather than any economic incentive, becoming the province of those
> rare people imaginative -or delusional- enough to look at pictures of
> the desolate surface of Mars and actually envision some version of the
> perusable for a relatively modest scale community with the will for it
> is one based on both telerobotic pre-settlement and the most
> functional use of immediate natural resources for settlement
> development in their least processed forms -the equivalent of building
> log cabins from the trees at-hand.
Yes, many of my friends are with me on the telerobotics scenarios as
well. But what comes between our (even present age) of teleoperabots
and the time when one-way trips start to become advertized? I'm
perfectly all right with a bit of detachment from the robot ecologies
on foreign planets, but that's not quite the narrative flare that one
would expect to be coming from those initiatives.
> Thus I've been giving much thought to the design of telerobotic
> outposts, their functional logistics, lowest-cost deployment strategy,
> the relationship between telecommunications latency and local systems
> intelligence, and the application of small scale robotics to the basic
> tasks of wide area resource assay, self-maintenance, habitat
> excavation, mining, construction, etc. Basically, the collective
I find the Internet Task Force studies on distributed asynchronous
internet protocols to be a satisfying medium for related research. Ah,
there's even a fellow around these parts that focuses on the
SolarNetOne project, which is a wireless solar powered linux network
that faces many of these initiatives out in middle-of-nowhere Africa.
Lots of interplanetary caching (Internet Archive on steroids), cloud
computing transferance of process instantiations, etc. I wonder though
how much work you've put into the design of those telerobotic
compartmentalized functions-- for instance, how does repair and
self-maintenance work, are you using self-replicating systems with
materials from the local environments? Or just the general
architectures?
> technology that would make it feasible for a relatively small number
> of people -a science foundation, a small business, a colonization
> advocacy group, or even just one particularly affluent individual
> interested in the world's most unique and grand hobby- to remotely
> conduct all the preliminary infrastructure development and habitat
> construction necessary to establish permanent -and comfortable- human
> habitation. The best model train layout ever. The one you can
> eventually retire to. And what's great about the concept is that this
> is something very open to experimentation and exploration by the
> hobbyist robotics community. We can build completely functional
> prototype telerobotic outposts on Earth and give space advocacy
> something practical to do beside fantasizing and lobbying.
I haven't considered converting the amateur rocket scene into amateur
'model train layout but-for-Space' advocacy. That's an interesting
method to approach, and I'm sure that the simulations that Paul and I
have been yammering on about would be an interesting first step.
There's already been talk about using software like FreeCity, among
others, and redoing some of the backbone to pull directly from our
open design repositories and to make it somewhat like "Moon Tycoon" I
guess (without the, uh, economics). This is immediately practical,
just not to the detail that we're all hoping for, and then getting
people to convert from the models to the physical "train sets"
implementations is a hard part. I guess you could pitch it to the
space colonies, Keith Henson and Charles F. Radley come to mind at the
moment.
I'd appreciate any information you can braindump on the communication
infrastructure, and simple interfacing thereof. That's a job for an
open source library project for things on top of Arduino and so on,
which if doesn't exist most certainly should. I'm only barely familiar
with (the concept of) radio antennae design, for instance, and haven't
read through 801.11b, etc. But it's rather important.
> concept that derives from the idea of the MUOL -modular unmanned
> orbital laboratory- which is basically an orbital distributed
> automation platform based on web controllers integrated by an IP
> backplane supporting multiple service client VPNs -put simply, an
> Internet you plug robots into.
Isn't this sort of what we already have? Except nobody cooperates in
our present state, and nobody packages it into an entire seed? (Except
perhaps ronja + arduino + open-wrt in some sick and twisted way- plus
everyone who has been obsessing over comm infrastructure collapse for
the past 20 years ..)
> Nanomining would be extremely simple in management because of the
> reliance on a fluid transport medium for all the actual work -just as
> with chemical leach-mining or based on bacterial bio-leaching would.
> And what makes it an early nanotechnology application is that it would
> be based on simpler application-specific disassemblers needing little
> control. What this would basically look like in old mines is a sealing-
> up of 'work face' sections with portable bulkheads then filling the
> gap space with working mediums and injecting disassemblers into it
> while using NanoChip 'sorter' ports to remove refined molecular
> material from that fluid and package it as feedstock fluid packed in
> drums. For new mining, the approach would be akin to leach-mining
In the case of biomining using that setup, there'd have to be some
flush-harvest-extract method. I've been involved in a commercial
super-secret startup on this topic, and we've been exploring the
prospects of different methods for lysis like viruses, sonication,
thermal depolymerization, hexane and other chemicals, mechanical
presses, osmotic pressure, electroporation, electromagnetism,
(bio-)flocculants, all sorts of things. So here's to hoping "renewable
energy" initiatives turn up something good on this front.
On another note, this always brings up the old images of the
dungeoncrawlers for me. I used to write random dungeon generators so
that I could have new levels and stories to play through, and the way
that you generate 'realistic dungeons and caves' (if there are such
things) is to use an exploration-and-search algorithm. Throw in an ant
simulator and you'd see some interesting mining dynamics going on I
bet.
Eric Hunting
On Nov 28, 2008, at 6:37 PM, Bryan Bishop wrote:
>
> Yes, many of my friends are with me on the telerobotics scenarios as
> well. But what comes between our (even present age) of teleoperabots
> and the time when one-way trips start to become advertized? I'm
> perfectly all right with a bit of detachment from the robot ecologies
> on foreign planets, but that's not quite the narrative flare that one
> would expect to be coming from those initiatives.
This is an important point. Everyone wants to get out there yesterday
but there's a very long development span between Columbus and the
Declaration of Independence -between the point when the New World was
a godforsaken wilderness people gambled their lives to make money from
then go home to Europe to live the Good Life and the point when it
became a place people went to stay in search of the Good Life. That
took, not counting the religious kooks, runaway slaves, indentured
servants, and convicts, about three centuries. Of course, part of the
reason this took centuries in the New World was because development
was so ad hoc and never really had an early intention of self-
sufficiency. There was never a concerted effort to establish an
infrastructure to match, through indigenous means, the top potential
for standard of living available in the Old World. Our ancestors had a
very different attitude about wilderness than we commonly do today and
it took them quite a long time of looking at even the verdant natural
splendor of the New World before starting to imagine the Good Life
there. Frank Lloyd Wright used to complain that Americans didn't even
develop a truly indigenous model of the Good Life until well into the
20th century -with him taking much of the credit for it, of course...
In space self-sufficiency is the whole objective since there is -for
the most part- no practical export income possible. This, along with
contemporary technology, would make this transitional development much
faster than it has been historically. Your telerobotic outpost isn't
there to make money. It's there to make a place to live. But it's also
a lot harder in an engineering sense. You don't have such wonder
resources like the simple tree (imagine history with no lumber) and
the harshness of the environments in space demands your resultant
artifacts be much more sophisticated to deal with that. And
teleoperation itself is an inherently slow process -that's the
compromise for its economy. Right now I would guess that, given
current technology, it would take as long as a full human generation
of telerobotic development before you get to that point where your
distant unmanned colony is sophisticated enough that it can build and
launch a spacecraft to come pick you up and bring you home to your
personal underground garden palace. Still, that's a lot better than
three centuries -and also a lot better than not at all given that just
one inevitable accidental human death on the space frontier can, in
the right political climate, kill a whole space program. It would take
about half that time to get to the point of supporting long-term, high
safety margin, but still Earth-dependent human occupation in more
modest accommodations -which would still be cheaper than even a 'lean'
concept like Mars Direct since you don't have to take a lot of stuff
with you on the trip over.
With later technology, for which this sort of project would itself
accelerate the pace of development, the situation much improves and
the amount of time shrinks. The more local intelligence you can build
into your systems the faster everything becomes because that
intelligence compensates for communications latency and leverages the
time management of the individual terrestrial technician -he can set a
lot of tasks on 'automatic pilot' and just wait for the results. First
generations of telerobotic systems will not be particularly
intelligent and so the number of technicians controlling them would be
pretty high -one-to-one for active systems like rovers and excavators-
and their control pretty direct -which makes latency critical and also
makes destinations like the Moon preferable to Mars. Barring some
miraculous breakthrough in AI, this will likely be pared down slowly.
Nanotechnology also potentially aids this tremendously by
miniaturizing and simplifying the tool kit necessary for building a
sophisticated infrastructure, especially in the most critical area of
materials processing, which currently tends to require large,
specialized, and complex systems. Figuring out how to miniaturize this
will be the most challenging engineering task for such a colonization
program. (and one of the key areas yielding great dividends for Post-
Industrial cultural development, since this too is one of the key
bottlenecks for open manufacture) With the benefit of nanotechnology
you get progressively more compact, scalable, and generalized systems.
At a certain point you could reduce the entire industrial
infrastructure for a robust civilization down to about 4 kinds of
compact and physically simple machines; Nanomining plants as I
previously described and which could also become the basis of
structural excavation systems, Nanodigesters which disassemble bulk
pulverized material in a fluid environment and sort it into respective
molecular feedstock (used more for recycling than mining), the
physically very similar Nanofoundries which, again, use a fluid
environment to fabricate artifacts from feedstock (ultimately used
more for the very intricate products beyond the resolution of
mechanical micropositioning), and Nanofabbers which are making
artifacts using the simpler and larger area approach of stationary
nanomachines on a NanoChip tool head moved around mechanically by a
precision translation platform. This sort of technology could reduce
the establishment of a complete infrastructure for colonization to a
decade. Long-term, nanotechnology might reduce the tool kit of
colonization to just one self-delivered piece of self-transforming
material (what I like to call NanoFoam -basically, that scaffold
microstructure of a Nanomining root with the ability to make its own
skin, resulting in a self-transforming structural material that
doesn't just make what you want, it becomes what you want), at which
point colonization becomes a project on par with building your own
vacation home across a couple of years and a means to a systematic
program of cybernetic exogenesis.
>
> I find the Internet Task Force studies on distributed asynchronous
> internet protocols to be a satisfying medium for related research. Ah,
> there's even a fellow around these parts that focuses on the
> SolarNetOne project, which is a wireless solar powered linux network
> that faces many of these initiatives out in middle-of-nowhere Africa.
> Lots of interplanetary caching (Internet Archive on steroids), cloud
> computing transferance of process instantiations, etc. I wonder though
> how much work you've put into the design of those telerobotic
> compartmentalized functions-- for instance, how does repair and
> self-maintenance work, are you using self-replicating systems with
> materials from the local environments? Or just the general
> architectures?
Ultimately, your outpost needs to employ self-replication of its
hardware as early as possible if it is going to remain cheap. This is
the key leveraging potential of automation. But it is certainly a
challenge and cannot be achieved from day-one given current
technology, largely because of limitations in system scale and
resource assay. Accomplishing this level of self-sufficiency will
require a transition in physical form of outposts over time based on
the need of industrial production requiring increasingly 'eutactic'
conditions the more sophisticated the products. The most sophisticated
production will demand conditions comparable to the ultimate permanent
human habitats, I've envisioned three stages of telerobotic settlement
with differing characteristics which correspond to the the first three
of the four stages of overall colonization. These four colonization
stages are;
-Initial outreach exploration
-Systematic assay
-Infrastructure and exploitation
-Permanent settlement
The first stage has largely been accomplished for the Moon and Mars by
the efforts of space agencies but would still require some duplication
by independent means, based largely on the fact that the national
space agencies have only ever produced white elephants that offer
little prospect of cost-effective repurposing. This is why most
commercial launch systems derive from military hardware even when the
technology of the space agencies is effectively free for use by anyone
with an interest in it. It's just, for the most part, expensive over-
elaborate junk in a commercial economic context. So for a telerobotic
settlement campaign this stage would look similar to what has already
been demonstrated; remote sensing probes, satellites, and very small
short-lived surface landers and rovers all implemented on an
unprecedentedly lower cost. Possible novel additions would be sensor-
web systems and impact probes based on very small, numerous, and
simple mass produced hardware. Initial outposts would be concerned
primarily with the deployment of strategic telecommunications and
telemetry facilities intended to provide flight control data for more
precise flight control for incoming vehicles and would tend to be
comprised of individual lander vehicles called 'beachhead' landers
which may also carry with initial motile robot hardware.
The second stage would be focused on establishing outposts that
function as a support infrastructure for systematic resource assay
using long range vehicles as well as initial experimental industrial
production based on the most accessible materials. Initial outposts
would be derived from the location of communications and telemetry
lander vehicles but would be followed by the deployment of similar
facilities with 'rough landed' hardware that is assembled on-site. In
this stage transportation is divided between 'soft' and 'rough'
landing systems, the former based on vertical landing platforms akin
to the beachhead landers and used to deploy relatively fragile or
bulky hardware and the latter based on rocket-chute or parachute
dropped air-bag-cushioned containers and pallets holding tightly
packed modular components. Rough landing systems would be a critical
technology, delivering the vast majority of supplies at the minimum
cost using the simplest trans-orbital vehicles. Most stationary
outpost systems would be based on self-contained modules standing in
the open like appliances and positioned by rover vehicles into
interconnecting clusters. Initial maintenance would be based on open-
environment stationary robot arm systems in service workstations
assisted by more mobile robots. Later, these would be contained in
simple sheds, perhaps in the form of structures akin to corrugated
steel arch huts, using teams of robot arms and larger component
storage structures. Initial free-roaming robots would be few and more
multifunctional with the most common being a recovery vehicle intended
to collect supplies from large drop zones. A variant of this would be
the primary exploration vehicle, designed to deploy communications
transponder modules as it travels around performing systematic
exploration and collecting material samples for more elaborate testing
in self-contained labs at the outposts, establishing key transit
routes in the process. Key locations such as lava tubes or exposed
rock faces would be identified for the next phase of outpost
construction based on excavation. Any initial resource exploitation
and industry in this phase would tend to be based on strip mining of
regolith and the recycling of rough lander discards and old or failed
hardware.
The third phase would be characterized by the intensive development of
excavated facilities and the establishment of key transit routes,
possibly with tracked transportation on a modest scale such as
monorails derived from today's banana monorails used on plantations.
In order to deploy an industrial infrastructure of scale, the outposts
have to move beyond the use of self-contained modular processing units
as deployed in the second stage to much larger factories based on much
smaller component modular systems. These will not likely work well in
the dust and radiation prone ambient environment. And so the outposts
would use complexes excavated by robotic roadheaders as a way to move
these facilities to a sheltered, low-dust, if still mostly non-
pressurized environment. Outfitting these complexes for use would be
based on an approach similar to one I've proposed for adaptive marine
settlement structures and unmanned orbital labs and factories. A grid
structure, possibly based on pins driven into the rock face, is used
to support a light modular frame of uniform geometry that hosts quick-
connect attachment of the rest of the components outfitting the
facilities. All other components in the telerobotic habitats would
attach to this structure, including stationary and tracked robots and
processing systems that can be formed into production lines. This same
architecture would be elaborated later to form the basis of the human
habitat.
This is the key stage for establishing an industrial self-sufficiency
based on only on comprehensive resource exploitation with a
transportation system to support it but also by extensive use of the
'min-a-max' (maximum diversity of utility from a minimum diversity of
components) principle of system design. For instance, most electronics
would employ dynamic gate array components that are far simpler than
typical microprocessors to fabricate and allow a few generic forms of
hardware to serve a large diversity of applications by encoding that
diversity in software. This sort of logic would be employed in the
design and engineering of just about everything going into these
outposts. This allows for generalization in the design of factories,
allowing the same few systems to produce a large diversity of things
on-demand -which brings us back to one of the dividends of this for
post-industrial development on Earth.
Later in this stage the earlier stage outposts that haven't evolved
into this excavated outpost form would be abandoned and the largest of
the new outposts would deploy the first comprehensive pressurized
habitat structures, starting with research farms and CELSS structures
and then habitats intended for human occupation. These would be
initially based on externally supported pneumatic enclosures installed
in the same excavated complexes. Some of the larger robotic rovers
would be adapted into human transports, though still maintaining
telerobotic control features. (all subsequent manned vehicles would
retain this capability)
The fourth stage of colonization would be characterized by the
bulkhead sealing and pressurization of large excavated structures,
possibly with rock faces sealed and stabilized with reinforced
indigenous ceramic materials, heat sealing, or perhaps reinforced
epoxy coatings. They would be outfit with a retrofit grid just like
the machine areas but now this would be used to host interior
finishing components to make a comfortable living environment. Rough
excavation and installation of base structural elements would be done
telerobotically as before but final interior outfitting may be
performed by the first human residents once the chambers are
pressurized. The general design of these human habitats would vary,
some areas still using the simple rectilinear vault grids but others
more specialized for residence and employing large central atrium
garden domes with perimeter dwellings, this organization expanded
radially in a succession of smaller self similar forms down to
individual homes with their own private gardens. If the assay work is
lucky enough to yield some large lava tunnels, these may eventually
become primary habitat space and in places like the Moon and Mars,
with their lower gravity, the spans of these can be tremendous -
kilometers wide and hundreds long. Vast forests and free-standing
buildings could fit inside them. Providing natural light to such
spaces would be done with large heliostat arrays on the surface,
piping light in by optical channels and fiber optic cables linked to
arrays of holographic membrane emitters suspended from the ceiling
like an artificial sky. Eventually this excavated approach to
structures would be combined with built-up surface structures based on
regolith-derived concretes and ceramics. These would likely be built
using a 'mound form' approach where regolith is piled and sculpted
into a desired shape and concrete poured on top of it with the form
dug out when the shell is cured -a technique first employed by German
military engineers in WWII. Their interiors would then be outfit in
the same way as the excavated complexes. These surface domes might be
made virtually transparent using an image-corrected variation of
heliostat lighting based on a geometric grid of exterior light
collectors with matched emitters on the interior, creating the
interior appearance of a transparent geodesic dome when, in fact, the
dome shell might be dozens of meters thick. An early version of this
technology exists today in the form of Litricon; a translucent
concrete with an embedded matrix of aligned optical fibers
manufactured in large bricks.
>
> Isn't this sort of what we already have? Except nobody cooperates in
> our present state, and nobody packages it into an entire seed? (Except
> perhaps ronja + arduino + open-wrt in some sick and twisted way- plus
> everyone who has been obsessing over comm infrastructure collapse for
> the past 20 years ..)
Yes, exactly. This is what makes this architectural approach seem very
plausible. It's not really new technology, We already have a lot of
simple demonstrations; robots you can drive around in a 'sandbox'
environment via a web page, sensor webs on satellite Internet linked
buoys at sea where you can access an on-board web server from anywhere
else on the globe, similar web servers on animal tracking collars,
commercial electronic signage that lets lease clients upload ad copy
on-line, and, of course, all the telecom hardware from home routers to
remote telecom stations with built-in control setting web pages. But
like you said, nothing works together. No one is thinking beyond the
web browser. They're all discrete devices out there on their own in
the aether. We only get close to concerted management of them with
things like WAN administration applications that are just gathering
data from a bird's eye view. There's a lot of potential in IP that
just isn't getting explored. So I looked at all of that and asked
myself, where's the next step? Where would that all go if you started
to use it like a platform for automated systems integration with a
specific purpose? What if you added a simple secure byte-code
protocol on top of the HTTP -a standardized IP MIDI or Fourth? And the
answer was, everything from a home control network, to a conventional
factory automation system, to a personal computer made of net
appliances with no fixed base of hardware, to a rocket launch command
sequencer, to a control network for a space station that lets you
access everything on board from any location you can link-up a tablet
computer, to a robot as big as the planet like Jacque Fresco's
Cybernation, if you're crazy enough...
Eric Hunting
erich...@gmail.com
Bryan Bishop
> On Nov 28, 2008, at 6:37 PM, Bryan Bishop wrote:
> > Yes, many of my friends are with me on the telerobotics scenarios as
> > well. But what comes between our (even present age) of teleoperabots
> > and the time when one-way trips start to become advertized? I'm
> > perfectly all right with a bit of detachment from the robot ecologies
> > on foreign planets, but that's not quite the narrative flare that one
> > would expect to be coming from those initiatives.
>
> This is an important point. Everyone wants to get out there yesterday
> but there's a very long development span between Columbus and the
> Declaration of Independence -between the point when the New World was
> a godforsaken wilderness people gambled their lives to make money from
> then go home to Europe to live the Good Life and the point when it
> became a place people went to stay in search of the Good Life. That
I can't believe I'm arguing history, but really it was a chance for
some escaping prosecution by religious institutions in some cases, in
other words "last resort - just leave" (on a quasi-well funded
seaship, evidently). However, chasing men to the stars doesn't seem
like the right way to go about any of this; much less chasing people
into open manufacturing infrastructures. (I'm imaging a large
loitering Godzilla .. okay, guess not.) Then there's the huge corpus
of science fiction literature where societies send their unwanted
individuals to other planets, as sort of a follow-up to the
British-Australia project, or sending
insert-hated-religious-group-here, etc. But that's not too interesting
to me - I'm more a fan of the one-way trips for singular individuals
(the more the merrier of course). One-way trips, ending in some sort
of super-dignified death (automatic cryonic suspension on board?
Alcor? Ben Best, anyone?), but with the explicit requirement of some
cultural artifact being sent back, such as a narrative to give that
'flare' that I was talking about, might be sufficient to get things
rolling. Heck, even a poetic license granted to a satellite or robot,
sending back "cultural artifacts from our waiting life in space" might
be enough. There are many people who have tried writing computers to
write poems; a long time ago I was working on a game that would
randomly generate its own content so that I may play it forever, in
the style of a rougelike dungeon game, or interactive fiction (the
interface didn't matter too much to me). It was crude and sucked, but
I do know that others have had success with (some) poetry. The poetry
would be based off of sensor readings from a singular satellite, Mars
rover, or whatever. These poems would be packaged into a book "from
the stars", that sort of thing, which may or may not spark some more
initiative in the communities. I feel sick even recommending it
though, because I'm more interested in going off and building engines
than waxing poetic -- but clearly NASA has failed to capture
imagination to the extent we'd like, etc. etc.
> compromise for its economy. Right now I would guess that, given
> current technology, it would take as long as a full human generation
> of telerobotic development before you get to that point where your
> distant unmanned colony is sophisticated enough that it can build and
> launch a spacecraft to come pick you up and bring you home to your
> personal underground garden palace. Still, that's a lot better than
Well, that calls for simulation. ;-)
> three centuries -and also a lot better than not at all given that just
> one inevitable accidental human death on the space frontier can, in
> the right political climate, kill a whole space program. It would take
I'd be open to ideas on how to promote the concept of "one way trips"
where, from the beginning, everybody knows that it will end in death,
and most certainly _not_ a "good life". Without killing initiatives of
optimism and improvement (i.e., none of the "meh, journeymen will
always die" crap).
> With later technology, for which this sort of project would itself
> accelerate the pace of development, the situation much improves and
> the amount of time shrinks. The more local intelligence you can build
> into your systems the faster everything becomes because that
> intelligence compensates for communications latency and leverages the
> time management of the individual terrestrial technician -he can set a
> lot of tasks on 'automatic pilot' and just wait for the results. First
> generations of telerobotic systems will not be particularly
> intelligent and so the number of technicians controlling them would be
> pretty high -one-to-one for active systems like rovers and excavators-
> and their control pretty direct -which makes latency critical and also
> makes destinations like the Moon preferable to Mars. Barring some
> miraculous breakthrough in AI, this will likely be pared down slowly.
I'm not convinced that it would require ai for sufficient autonomy to
get humans out of the equation. I'm fine with excavator operators
sitting here on earth at first, since this is better than nothing, but
there are tricks to robotic operation, like pouring a slab for a giant
foundation and even out a field on which the robots can do their work,
meanwhile resource harvesting systems are deployed and only at the
"edges" of the "bubble of expansion" do remote human operators work so
as to process irregularities in materials, geography, that sort of
thing. It's not very adaptive to just assume that everything on the
slab is working perfectly, but as a first approximation why not? When
errors do occur, remote humans would log in and take a peek anyway.
> Nanotechnology also potentially aids this tremendously by
> miniaturizing and simplifying the tool kit necessary for building a
> sophisticated infrastructure, especially in the most critical area of
> materials processing, which currently tends to require large,
> specialized, and complex systems. Figuring out how to miniaturize this
> will be the most challenging engineering task for such a colonization
> program. (and one of the key areas yielding great dividends for Post-
> Industrial cultural development, since this too is one of the key
> bottlenecks for open manufacture) With the benefit of nanotechnology
> you get progressively more compact, scalable, and generalized systems.
I haven't yet figured out what you mean when you say nanotech. Are you
talking Drexler and molecular nanotech, Drexlerian MNT? I don't tend
to include MNT in my planning because of how unavailable the
technology is, as well as the methods of bootstraping MNT. But if it
ever begins to work out, sure. For instance, Freitas and Drexler have
recently been working with some computational chemistry tools to
simulate diamondoid tools, such as the movement of hydrogen atoms from
specialized tips (consisting of three or four molecules IIRC). If they
would be willing to release the simulation configuration, because I'm
sure it's just running on CHARMM or something, then sure, I'd take a
look at what they have in mind, but I'm not getting my hopes up.
Over this past summer I was working in a molecular biology lab on,
broadly, "the intersection of computation and manufacturing" as
applicable to biology [and what it was really was me doing lots of
gels, but that's okay]. I was under a DARPA grant for growing a tank
from a seed, or at least that's what the grant said the future
prospects were; the concept is basically similar to amorphous
computing with DNA and such; this technically qualifies as some sort
of nanotech, and Winfree's lab up at Caltech has done some interesting
work on patterned growth, such as the smiley face made out of DNA, or
the cellular automata implemented on strands of self-assembling DNA,
etc. These systems are presently delicate and far from robust (in
terms of contexts of application), and not nearly as abstractly
universal as MNT would be, but it's worth your investigation if you
haven't seen it yet.
> At a certain point you could reduce the entire industrial
> infrastructure for a robust civilization down to about 4 kinds of
> compact and physically simple machines; Nanomining plants as I
> previously described and which could also become the basis of
> structural excavation systems, Nanodigesters which disassemble bulk
> pulverized material in a fluid environment and sort it into respective
> molecular feedstock (used more for recycling than mining), the
> physically very similar Nanofoundries which, again, use a fluid
> environment to fabricate artifacts from feedstock (ultimately used
> more for the very intricate products beyond the resolution of
> mechanical micropositioning), and Nanofabbers which are making
> artifacts using the simpler and larger area approach of stationary
> nanomachines on a NanoChip tool head moved around mechanically by a
> precision translation platform. This sort of technology could reduce
> the establishment of a complete infrastructure for colonization to a
> decade. Long-term, nanotechnology might reduce the tool kit of
> colonization to just one self-delivered piece of self-transforming
> material (what I like to call NanoFoam -basically, that scaffold
> microstructure of a Nanomining root with the ability to make its own
> skin, resulting in a self-transforming structural material that
> doesn't just make what you want, it becomes what you want), at which
> point colonization becomes a project on par with building your own
> vacation home across a couple of years and a means to a systematic
> program of cybernetic exogenesis.
This illustrates some interesting discrepancies between our ideas. I
would grant you all of that if it wasn't for MNT being the basis for
most of it (it is, right?). Paul and I consider a toolkit based on
macroscale technology, mostly made up of things already invented,
designed and probably in use today, as a way to possibly achieve
self-replication, in the style of Freitas' kinematic self-replicating
machines, and recipes for the construction thereof. I genuinely find
it strange that you're focusing on nanotech; isn't this a large
bottleneck in your ideas and plans? What's going on?
Though when I look at your four stages I see MNT going unmentioned, so
I'll get to those in a bit.
Eric Hunting
This notion of reducing the costs of manned exploration of space by
one-way journeys has proven to have some legs in the space advocacy
community but I suspect will always be a rather difficult concept. I
think there will always be a question of whether or not the volunteer
is making a truly reasoned or coerced choice, the society exploiting
disadvantages or situations it helped create. Remember the old
cultural cliche of heartbroken men from around the world joining the
French Foreign Legion to go lose their sadness and memory fighting in
the desert? Is that exploitation? Is a choice made in that situation
an entirely rational choice? Under what criteria can this choice be
considered rational? Is the suicide bomber crazy, brainwashed, or
rational and what would it actually mean for society if he was
rational? Our culture generally lacks the extreme communal identity
that would produce large numbers of volunteers for such missions
without the presence of a threat for this activity to defend others
against. There probably would be a very small but potentially
sufficient portion of the population willing to accept the deal of the
one-way trip for a personal place in history. But, can you trust their
stated reasons for it? This is a very difficult subject area for our
primitive culture. We have a hard time accepting the notion of suicide
as rational, even in situations of extreme clinical suffering, because
we have a hard time accepting the responsibility of society to the
individual and his quality of life. We pretend suicide is always
irrational because if it ever wasn't it demands probing the validity
of the logic behind the choice, which risks exposing society's
potential responsibilities in the situation leading to that decision.
I suspect this idea of the one-way trip into the history books will
always be a minefield. That may be why so many science fiction writers
like to explore it.
I don't mean to imply that this would call for a HAL 9000 sort of AI.
But it may demand more intelligence and communal behavior management
than contemporary robots have so far exhibited. And progress in this
area remains relatively slow. Certainly, a lot of things even with
current automation technology could be 'set and forget' systems. It
doesn't take a lot of intelligence to operate a PV array, to have self-
contained factories processing things consistently and repeatedly, to
have a regolith dredger run a simple pattern by itself, or have a
rover follow a GPS trajectory to get from one place to another along a
pre-surveyed safe route. It's the tasks that have a large variability
in parameters that are the challenge, like repair work, recovery of
supplies from drop zones, exploration, and so on. Early on, these
sorts of tasks will lead in number compared to the more 'by rote'
activities and the latency will become a greater issue in the speed of
performing them. These more directed activities will tend to be
associated with the edge of progress and expansion of the colony;
seeking out new sources of materials, deploying new equipment,
building new structures. I don't think it's an outrageous challenge
but we are still a bit of a ways off from being able to give a team of
robots a general colonization protocol and then just letting them go.
But then, maybe we can get to that level in on-Earth research by the
time a real mission could be deployed. We're still talking a decade or
two out from today. If a project were started this minute, it would
take at least 5 years just to get sufficient support from the general
space advocacy, open source, and academic communities to setup the
first comprehensive field telerobotics labs. Consensus is particularly
difficult for the space advocacy community.
Yes, I was referring to the Drexler/Foresight notion of nanotech, but
I do include the biophysics/biochemistry derived sphere of things as
part of the nanotech lineage. I envision an evolution in applied
nanofabrication that goes something like so...;
Statistical assembly - (where we are today, more-or-less) the meeting
of organic chemistry to biophysics in systems which rely on
statistical modeling to predict the behavior of molecular systems in a
fluid medium and using complex proteins and natural or modified micro-
organisms as 'proto-assemblers'. Basic tool forms: conventional
reactors and Mixer Plants -an evolution of the DNA sequencer combining
complex protein synthesis with large volume long chain processing.
Limited primarily to bulk materials, microelectronics, microphotonics,
and discrete nanocomponents. May produce the first free-moving
specialized assemblers.
Nanomachining or 'crude mechanosynthesis' - assembly based on brute
force manipulation of discrete atoms and molecules in a planar domain
using variants of atomic force microscope technology. Basic tool
forms; NanoLathe -an evolution of ATM based nanomanipulators. Imagine
arrays of modular ATM heads with specialized designs in branched
production lines. Limited mostly to NanoChip -stationary chip-based
nanomachines- production and may evolve to incorporate the first
NanoChip tool heads. May, in combination with statistical assembly
technology, produce active assemblers.
NanoChip processing - a higher-order crude mechanosynthesis based on
specialized NanoChip devices integrated into larger machines. Likely
initial forms would be molecular sorters, which sort molar species in
fluid suspensions, and spinnerets, molecular looms used as an extruder
for bulk simple nanomaterials. Later more advanced tool forms would be
multi-function fabber heads and dissembler heads. Very diverse
potential product range from bulk materials to advanced
microelectronics. Likely to remain a common form of nanofabrication
for long into the future.
Free assembly or fine mechanosynthesis - molecular assembly using
complex mobile assembler systems in a eutactic environment of fluid
hydrocarbons or, possibly, gaseous hydrocarbons in a microgravity
environment. Primary tool form; NanoFoundry. An evolution of Mixer
Plant systems combining in-line NanoChip systems with assembler
'cultures' made of specialized and multi-functional assemblers in
processing and storage container arrays converging on a single primary
process tank. This is comprehensive nanofabrication as envisioned by
Drexler. Would use several modes or fabrication schemes based on the
scale of artifacts being worked on and the order of their fabrication;
free-floating, bottom-up, top-down, inside-out, etc. The most
versatile of nanofabrication approaches, will also likely be much
slower, more complex, and more cumbersome in the form of hardware than
NanoChip based fabbers and early systems could possibly be limited to
operation in microgravity, which would drastically hamper their
application.
NanoFoam - ultimate form of NanoFoundry based on a self-transforming
matrix of scaffold structures with a skin system that hosts colonial
assemblers, modular 'organelles' (volumetric NanoChips with mobility
but which perform mass production or data processing in a stationary
mode), and larger systems in a multiply redundant colonial
architecture. A likely evolution of the strive for NanoFoundries to
match NanoChip based fabbers in convenience and speed while affording
greater versatility thought the use of chrysalis structures providing
on-demand containment. Would also be influenced by the development of
polymorphic nanocompositiees; bulk nanomaterials with integral
mechanisms for changing their physical characteristics. Could produce
artifacts like a NanoFoundry but used primarily to _become_ artifacts
on demand while maintaining a contiguous direct or wireless data
processing, storage, and control network among all like material. Thus
the material can serve as a self-replicating self-recycling matrix for
an entire habitat with the potential to host a distributed computing
environment and simple intelligence. Would likely increasingly borrow
technology from nanomedicine as well as employ biomimicry as a means
of self-organizing and self-optimizing functional design.
Ambient free assembly - the most advanced form of nanofabrication
based on assemblers of very great sophistication and
multifunctionality able to tolerate, operate in, and salvage raw
materials from an ambient fluid environment including air. This is the
form most commonly envisioned by the public as the definitive form of
nanotechnology, thanks to common portrayals in unsophisticated science
fiction, and the form which most inspires all the Grey Goo paranoia.
However, because of the problems of ambient mobility, communication,
navigation, high on-board intelligence overhead, and resistance to
environmental free-radicals and radiation, it is likely to be the most
difficult and thus the last form of nanofabrication realized -assuming
the development of it isn't precluded by NanoFoam which can
effectively do the same jobs with a much simpler technology.
It might seem like I'm focussing on nanotechnology because it's a pet
topic of mine and I like speculating on it, largely because it seems
to me that people have been led into a very narrow picture of the form
it would take and the way it would impact civilization that risks
hampering its development by creating unrealistic expectations. But
actually I'm only suggesting how it might be applied if and when it
appears. At this point that if and when remains speculation, even if
the cultural consensus seems to suggest 'imminent'. We had the same
opinion about AI in the late 60s and 70s and look where that is now.
It's easy to fall into the trap of over-estimation of the near-term
and underestimating the long-term -Greek temple on a golf course
futurism. There was a time when we thought the ultimate future of
personal computer was a HAL 9000 in your pocket and only Ted Nelson
was able to imagine something like the Internet. If you look at the
stages of evolution I anticipate for nanotechnology, for the most part
its impact parallels that of the integrated circuit as it emerged over
the late 20th century. It's not going to be the universal tool
immediately. No instant Singularity. It has this evolution to go
through where, for a long time, it's going to be mostly producing bulk
materials and mass-produced microdevices very much akin to ICs. And
the impact of these products will be subtle -another forestalling of
the plateauing of Moore's Law, another order of magnitude performance
improvement in composites, another boost in basic digital display
resolution, flat panel displays made of unusual new materials, paint-
on lighting and PVs, complex electronics by ink-jet printing,
incremental increases in cost-performance of PVs, higher-performance
fiber optics, larger span concrete structures, new architectural
fabrics and membranes, metamaterials serving as synthetic catalysts or
exhibiting optical properties of negative refraction indexes,
mechanical adhesives, gas-less vacuum-lift dirigibles, etc. There's a
very long gap indeed between this and the cybernetic genie.
I don't think nanotech is strictly a necessity for telerobotic
colonization. We've far from exhausted the potential of design upon
current technology to realize new capability. It can help, certainly,
but it's impact is largely in terms of saving time. Until you get to
that level of things like the NanoFoundry -a relatively late
development marking the maturity of the technology- where you might
actually have the means of fabrication of large complex systems whole
and thus end up with radically different designs for them, the impact
on the architecture of a telerobotic settlement scheme would be
limited to the internal design of production systems and the materials
things are made out of. This would have important accelerating effects
if it happened, but it's still subtle and we're still talking about
the same basic assortment of hardware managed telerobotically in about
the same way. The four phases of colonization I suggest are generic
and derived from historic context, independent of any particular
technology. The chief difference with space is the lack of an export
economic impetus for the middle two phases as was common historically,
which telerobotics is intended to overcome as a radical cost reducer
compared to manned transport and facilities. Past colonization relied
on export income from resource exploitation to sustain and expand it.
Telerobotic colonization can only expect similar income based on the
advance and marketing of technology it employs in space repurposed to
terrestrial applications, which is not insignificant but nowhere near
the scale of income of New World exploitation. NASA could never pay
its own way on commercialization of its spin-offs. Telerobotic
colonization -because of its tremendous savings over manned space
activity and the diverse terrestrial market for of its underlying
technologies- might, just barely.
Eric Hunting
erich...@gmail.com
Eric Hunting
erich...@gmail.com