Re:(the problem with) Steel Buildings
on
Pre-Fab Homes?
·
· Score: 1
I would say the heat issue could be dealt with pretty easily by using lots of spray foam insulation and/or "dead air" gaps between the exterior and interior walls (with plenty of ventilation). You would only cool the interior structure, not the entire steel shell (which would be expensive, as you note).
Wiring is also a concern as well, you wouldn't just run regular cable everywhere - you would want the special grommets as you propose, or PVC/steel conduit piping.
Acoustic insulation should not be a problem - interior walls would be constructed as normal - the steel building is just the shell around the interior construction.
Steel Buildings
on
Pre-Fab Homes?
·
· Score: 2, Interesting
Look into steel building manufacturers and contractors. Many times, these people offer various packages to suit your needs (outside and inside details). Yours may be "funky" (ie, a house instead of a business or meeting hall), but they should be able to work with you to get what you want. Furthermore, such buildings can be *very* cheap (a place here in the Phoenix area is offering a 80 x 150 foot building for $50,000 - sure, that is probably without any "extras" - but still damn cheap for the sq footage).
I hear you there - my wife's car is a Neon, and it needs the timing belt replaced soon - I cringe at the cost (front wheel drive, etc).
On the other hand, I recently purchased a 1979 Ford Bronco with a 400ci V8 - lotsa room to get in, under, around - everywhere - to play/tinker/fix (and yeah, it needs a *lot* of work yet before I will really feel safe driving it - fortunately, most of it is minor).
Not just electric R/C - but *indoor* electric "slow-fly" R/C has taken off, so to speak - something that would have been deemed almost impossible 10 years ago. Sure, tiny motors have helped, but tiny (and lightweight) batteries have also been a huge improvement allowing for these hobbies (in addition to more electric helicopters, which take way more power to run than a plane)...
Well - at least it got two nominations. I nominated it before I read the comments (also nominated SRI's Shakey, and the GE Hardiman suit - though technically not a robot). I have very little information on the ODEX-1.
Last I heard, the prototype(s?) were sitting at the Smithsonian. I have articles from Popular Science, Popular Mechanics, and Omni that showed this robot off nicely. I remember seeing it working on That's Incredible (or the old Ripley's Believe It or Not). I tried to contact the company (at the time, they were working on an "intelligent highway" system), but they had no clue what I was talking about. From the information I have read, Odetics (and the ODEX-1) originally started out as a DARPA funded project for a robot soldier (in fact, that is what the Pop Mech article showed, a swarm of ODEX bots storming a battlefield, guns ablazin'!) - then, somewhere, it dissappeared or lost funding.
Crazy thing was, they had the prototype fully working (at least, it walked rather well - the video I saw of it showed it climbing out of a pickup, and then "scrunching" itself down to navigate through a doorway, then spreading back out for full walking mode)!
Oh, I never thought it did. I didn't even get the thing notarized - even at the time I drew it up, I had the thought that someone, somewhere, had already thought about this, and that likely a patent already floated around in the PTO's database...
I know that there is a model of a Couger that the battery is only accessible from under the vehicle (saw this when my wife and I went shopping for a car once). I also know that on an older Mercedes convertable (hardtop) my brother-in-law had, the battery was in the trunk (!!!). My wife's Neon (97), the battery is up front, easily accessible (however, it will take an act of god to get the camshaft position sensor fixed, if it ever comes to that - fuel lines and vacuum hoses make it impossible to get to without removing them).
I know I have (somewhere) kicking around in my files a drawing I drew up as an "invention" of contact lens VR devices. Now, I know some of you are thinking "yeah, right" - but it is the truth. I was originally imagining the devices to use hard contacts, with a grid on the contacts driving LCD type elements. I also thought that the electronics would be on the edges of the contacts (which would be mainly driver electronics, pickup "coils" for power supply and datacomms, and position sensors to tell which way the eye was pointed in) and the contacts would have to be larger than normal (to allow for changing pupil size). I also gave thought to using that chemical that widens the pupil to the widest size (which they put into your eyes when you go to get contacts). Finally, I thought about "back-light" shades that you could wear when you wanted full immersion (otherwise it would only be augmented reality - AR). I should have gotten that drawing notarized or something, as it is I only signed and dated it. I didn't think much of it, because I figured others have had similar thoughts before me, and I didn't have the money to patent the idea (I could have easily made a mock up of blown-up scale, or a virtual representation in 3D or something - which would be accepted by the PTO as a model) - plus I figured something like it was already patented. I know I have also posted both here and on k5 about the idea (maybe on Fark, too). I am just glad to see others thinking along this line. I have my doubts on how well it would work (or whether the res would be high enough), but the thing I like most about the idea is that it has the ultimate field-of-view (FOV) possible, by covering your entire pupil, so that you can get true *full* immersion, with peripheral vision and all. If such devices are ever built, they will be amazing to use!
Personally, I don't know how to drive a stick - but I can see where a manual would be really useful, such as when you want to apply more engine power and stay in one gear (you can do this a little with an automatic - ie, 1st, 2nd), for certain things like uphill/downhill use, off-road, etc.
Most people drive automatics over here because most vehicles have them (not too many front-wheel-drive manual vehicles are made anymore). Trucks and SUV's (and sports cars) are the exceptions, where you can get one or the other (or sometimes *only* a manual). Those "in the know" go for a manual, others stick with automatics. I don't know why it is different in Britain, but it is something I have noticed before (maybe it has something to do with kilometers per litre - automatics tend to be more fuel hungry, manuals less so, provided you drive them right)...
While your comment on a car wasting energy while sitting idle is a valid one - the clutch issue is only true if dealing with a standard manual tranny - automatics use a "torque converter", which at idle simply spins (wasting energy, though, in the process).
At one time, before automatic trannys, there were "semi-automatic" transmissions (I think one of the marketing names was "hydromatic" - but I may be wrong here), that used much of the same components as a fully auto tranny, but the user switched gears (ie, tightened/loosened the bands on the planetarys) via a pushbutton selector, or a simple stick mechanism - no clutching needed, but the user still had to "listen" for the engine to know when to shift...
Do you have the original copy of the "Manufacturer's Statement of Origin"? This is essentially the "bill of sale" from the manufacturer to the dealership. When they "sell" you the car (and I say this because many people finance the car), they send (the original? a copy?) it to the Motor Vehicle Department in your area (actually, I think a copy goes to them, and the original goes to the lending institution). When you pay up your note, the original goes to the state MVD.
Now, what if you pay cash? Well, the original still goes to the state MVD.
In exchange for this (it is part of registration), you get your "license to drive" - well, actually, to get a license, you have to surrender your MSO to the MVD.
There is a lot of speculation that it may all be bullshit (like all good conspiracy theories), but look into "Right to Travel" on Google.
Basically, as the theory goes, when licensing for automobiles came about, we traded our freedom to travel for the automobile license, and thus have become slaves (not Free Men) to the State...
The reason robot mower companies don't add any geometry tracking is probably because they don't want to "add a penny more" to the cost of producing the robot - that is one more penny in profit, dontcha know!
Anyhow, you likely wouldn't even need a vision system for most of these robots. A few k of RAM would probably suffice - note:
Lawns are mainly two-dimensional surfaces - so all you really have to keep track of is where you have been in the two dimensional plane. And you don't even have to keep the tracking super accurate - down to meter or half-meter square would probably be accurate enough (hell, lets make it nice - to the square foot). Each "square" of lawn gets one byte in an array. Now, instead of a single wire surrounding the property with only one signal, each "edge" of the property gets its own wire with a specific signal or identifier coming from it (maybe it is a different frequency, or maybe it is sending out a different pulse ID or something). So, the robot is placed in this delimited "grid" (which would be marked up in a GUI to show where the wires "are", how big the yard is, what squares in the tracking memory represent the yard, where the robot is placed in the yard at the start, and what direction it points in), and let to do its random walk or pattern to mow. Using "dead reckoning" by tracking the distance the wheels have moved, it will know what squares (bytes in the array) it has visited, and maybe even the coverage (0=no coverage, 255=fully mowed), based on how often it has visited the square, and from what square(s) around it it has come from. Now, the dead-reckoning system isn't perfect (but is cheaper, currently, than GPS), the robot could get turned around and lost. However, eventually it would come to one of the boundry lines, and it would get a signal telling it what side of the lawn it was on, and it would know that it is turned around in some manner - perhaps it could reset itself, and orient from that position, and continue on from there.
Such a system (which I hereby place under the GPL for Hardware license - and my description is prior art) would be very cheap (most of it is software - the rest of the hardware is probably on board most mower robots, and what little isn't is VERY cheap) - but they won't implement it because it would mean a better product that costs them more to make, and hence less profit for them.
Actually, there exists a walking surface that is essentially a conveyor belt that can move in both directions simultaneously - it was created for DARPA's "dismounted soldier" project (think of two belts overlapped at right angles to each other, and the belts consist of multiple rollers running parallel with the direction of travel of the belt - ie, perpendicular to the end rollers for the belts). The system also had hydraulic rams to change the incline of the "terrain", as well as motors and such to provide feedback and terrain feel changes.
Not perfect, but not a bad system for a CAVE, either...
Except you don't see it anymore - dig up some old copies of Byte magazine (pre-1985?), and look for ADDS ViewPoint terminal reviews/adverts/etc - if you have ever seen one of the older models, there does exist on the keyboard a real "ANY" key. I don't know or remember what it was used for (waaay before my time), but such things did, at one time, exist.
I think the thing that saddens me is that with all the computer geeks here on/., nobody seems to remember the history of that which they love (at least, I didn't find a mention about this in the first few top pages of comments)...
If you note, I do write in my second sentence "While a pure nanotube fiber has yet to appear". I realize that pure versions of such fibers don't exist. However, I think a disservice was done by not mentioning such research being done to create such fibers right now. Much of the discussion about "space elevators" seem like so much fantasy and sci-fi to the average person - they don't realize that this technology may not be far in the future, that it may be realized sooner than we think or expect...
They, of course, phrased that wrong - arguably, Philo T. Farnsworth is the inventor of completely electronic television. Until RCA and Sarnoff stole his ideas and ran. This resulted in Farnsworth dying "a pauper". Only recently was he reinstated in that role instead of RCA, ala court action, similar to the Tesla/Marconi debate, ENIAC vs. ABC, among others...
Is that at this conference they seem to think that carbon nanotube fibers of any kind don't exist? While a pure nanotube fiber has yet to appear, why wasn't any mention made in the article of this:
Maybe they did discuss all this and more at the conference - I would like to hope that these scientists and researchers are aware of what is going on in this far-flung field. I only wish they would have made mention of this stuff in the article for the common man, to show that it wasn't all so much "hooey" - that it is something which may be inevitable, and will happen sooner than we all expect.
We (all of mankind) are rapidly moving in a very funky direction, technology-wise. We have carbon-nanotube fibers. We are looking into other advanced fibers and fiber processes. We have found sea-creatures that make insanely great fiber optic fibers (and with the other stuff, we will probably be able to replicate the process very soon). The gains in communications alone will cause a lot of other gains to be made, because of distributed processing amongst far-apart supercomputing centers that need more bandwidth than they already have (and they have a crapload, but not as much as they want or need). Such fibers may help in the optical-computing dept as well. Remember also the stories of "growing diamonds" - that are so pure they are almost impossible to distinguish from real diamonds - and they have DeBeers quaking at the possibilities to their "markets", maybe destroying them. But these companies don't want the diamonds for prettiness or money (well, they want them for the money, true), but to be able to use them for the substrate of computer chips, instead of silicon, for higher speeds and better heat dissipation.
Couple that with all the other "funky" advances we have seen - we are all being dragged in a very wierd direction, speeding up the computing and learning capacity of all involved (and even if you are at the edge of the network, like most of us are here, and not where the action is, you will still be pulled in)...
I don't know where to go with this - except that our current distopia (and if you don't think we are living in a distopia, one every bit as scary, strange, and awe-inspiring as science fiction can come up with - you haven't been paying attention) is going in a new and strange direction, strangely reminiscent of what the "early-years" (which are only touched on) of Neal Stephenson's "Diamond Age" might have been like.
This is all strange shit, yet very few of us are even seeing it or thinking about the real implications, for some reason...
My parents bought me my first mouse for my Color Computer 2 - it has (yes, present tense, because I still have it) a single red button on the top of the black "soap bar" style body, and the mouse ball is a steel ball bearing of some sort (no rubber covering). It plugged into the joystick port of the CoCo, and had abysmal resolution - but for what I used it for mainly (CoCo Max drawing, and CerComp's windowing system), it was pretty cool...
My hometown is Bakersfield, CA (stop laughing) - I-5 bypasses most of Bako, but you can go through it via highway 99. Now, highway 99 is interesting unto itself. It used to run a different route, on what is now called Union Ave (I think that is right) - or also known as "old 99". There is a turnoff off of 99 as you head north, I think just past the turnoff to I-5. Anyhow, old 99 is one of those things you have to see.
It's a 4 to 6 lane highway, with huge eucalyptus trees flanking the shoulders, and in some areas, the median. Runs north for quite a way, until you get into downtown, then it kinda "dead-ends" or changes into a street. The current 99 actually continues north and reconnects with I-5.
Both 99 and Union Ave (old 99), for long sections, use that same grooved concrete slab construction - and it has to be rode on to - ahem - "appreciate" it.
It is like driving on a "washboard" road with the "bumps" being elongated (several car lengths) - basically the concrete slabs "dip" in the middle, and the ends rise "up". I think it is due to a combination of settling, heat, and large trucks. It isn't super quiet, and it feels like a cheap roller coaster. New 99 is better than old 99, but that is only age difference...
So, I hope that stretch stays experimental. We also have a similar type setup (where it isn't paved with asphalt) here in Phoenix on the I-17 - same grooved concrete, same issue. It might be a better thing if the concrete is laid in a very long ribbon, rather than slabs...
I mean, do you realize how much 1000 ft of cat5 weighs? Unless you are planning on using many, many tens of cubic feet of helium, the cat5 is going to weigh it down.
If you really want to do this, you are probably looking into a *very* custom board - I would be thinking something like a camera connected to a microcontroller communicating via 802.11 to the ground. It would have to be completely wireless for what you are wanting to do, unless you are planning on using a very large balloon envelope, or hydrogen (and even then, you would need a large envelope, but not quite as big - probably 12-20 feet in diameter)...
It seems like a lot of people forget that, I know I did until recently. Ink jet printers seem to be a cheap solution - until you realize just how much you are spending on ink.
I own (but no longer use) an Epson Photo Stylus 700, which I bought because I loved the quality of the output when used with the "special" photo paper. I never printer one picture on the paper. I think the greatest thing I *ever* did with the printer was make some nice Thanksgiving party invitations.
It seemed like I was always buying ink - because we rarely used it, but left it turned on. This tended to leave the print heads uncapped (I think they do this on purpose, rather than auto-capping, to sell more ink), and caused the ink to dry out prematurely. But you wanted to leave it on, because it seemed to take forever to "boot" (turn it on, and after minutes of "self-checking" and "cleaning" it would finally be ready. I took a look at how we were printing (rarely, but we wanted good output *now* when we did), what we were printing (most of the time, simple text only stuff, black and white) - and I bought a printer based on that.
I ended up buying a used HP Laserjet 6 (there is a P or something there at the end, too), and a refilled toner cartridge. Total cost: $170.00 - and I have postscript, too. I installed some old 72 pin SIMMs I had lying around to bump the cache up some, and I haven't looked back.
The printer is great - what was really nice was the low page count (25000 pages). I also like the fact that I can use el-cheapo paper in it, and it still looks great (the Epson, on anything under 24lb weight, would "fuzz" - lighter weight paper had more "fur", and the print wouldn't have crisp edges). I also like it that I can leave it on - and then when I want to print out to it, I instantly can - and it just works!
Now, maybe if your job or hobby requires color, an ink jet is what you need to get. But I learned my lesson quick - I don't have *any* need for color. If I want to look at pictures, I look at them on a screen. Just about anyone else can do the same (most people I know have a computer). If I need a print of an image, I will print it in b/w for "checking", then the final can be done at a copy shop or something. I have yet to need to do this, though - but it is the most sensible option, for me.
...told this guy who was really paranoid something that probably still keeps him up late at night.
I told him that even if you shred then burn your documents, that THEY have devices that look at the smoke coming out of your fireplace chimney, and can decode the information that is in the smoke patterns. I told him that as the paper burns, the chemicals that make up the ink vs the paper are different, and this difference changes the information matrix encoded in the smoke, and that a detector with sufficient processing power behind it could reconstruct the document data after it was burned.
One thing I have recently considered, but nobody seems to talk about much, if at all - is if China is really going to use chemical rockets for a push to Mars,or something else.
I tend to think they may be smarter than we give them credit for - and they are going to kick our asses into the ground by getting a manned mission to Mars, and soon. How? By using late 1950's technology - otherwise known as Project Orion.
They have *all* the technology to do this, and it would be a major coup for them to lift 1000 or more tons into orbit and beyond, using nuclear bomb/pusher plate propulsion technology.
It may sound crazy, but we came *this* close to doing it ourselves, but political and eco-activist pressure caused the project to be scrapped.
Much of the project documentation, spearheaded by General Atomics, is still to this day considered "top-secret". Mostly because the Orion Project required very small thermonuclear bombs to run on (small H-Bombs - 1KT to 15KT), plus the documentation on how to mitigate the effects of the pusher plate ablation, and the sheilding substances - are useful information on designing H-Bombs with more destructive effects, while at the same time knowing how to protect yourself from the effects. Also, the design of the bombs basically made them directed energy weapons (so that the energy could mostly go into pushing the craft) - so these small bombs required special radiation channelling systems, which were basically one stage in a staged H-Bomb (ie, fission bomb directs energy to hydrogen/deuterium solid core, to compress and "kick" the fusion process off - a three stage bomb could be made by adding another core/channel system, so that the energy from the smaller H-Bomb could kick start a larger reaction). Plus, there was the design/use of (thorium?) as a means to kick the reaction up higher...
But they surely know about all of this - and if our engineers could work this out in the late 50's using computers (IBM 704?) with less power than a low-end Pentium (heck, even that is overkill compared to the 704) - don't you think they could do the same with today's machines?
I would be willing to bet that they are researching the technology. It is doable. They have the land, they have the bomb technology, and they have the will - I doubt they care about any form of fallout...
At least in the example given, you are using all fairly modern machines. The real hacker challenge comes when you want to interface something a little more esoteric - like that nifty C=64, TRS-80 Color Computer, or $DEITY forbid - a shiny IMSAI 8080 you just picked up off of eBay...
Yeah, that's where the challenge lies. However, even today, kids have it easy: Provided the thing has a serial port and you can code to it, there are small serial-to-ethernet "converters" available (most of them consist of some form of microcontroller acting as a "go-between" from the ethernet interface and the serial port).
I remember one time in the early 1990's when I picked up a Compaq SLT/386 with 6 meg of RAM, running Caldera's OpenDOS (IIRC). No PCMCIA slots - only a serial port and a parallel port. Since network equipment was still fairly expensive (especially those lovely pocket parallel ethernet adaptors), I looked for a solution.
I ended up creating a funky bit-banging parallel port solution using 4 conductor phone line, dual jack adaptors, and custom wired parallel to RJ-11 plugs. I intended to write software to allow all computers on this network to transmit/receive on it - checking for the status of the lines to avoid colisions, random wait times when there was a busy, etc - I was looking to get 9600 baud on this thing. I managed to build enough dongles for three machines, but I never got around to the coding portion. Always wondered how well (if at all) it would have worked...
Slashdot Mirror
Wiring is also a concern as well, you wouldn't just run regular cable everywhere - you would want the special grommets as you propose, or PVC/steel conduit piping.
Acoustic insulation should not be a problem - interior walls would be constructed as normal - the steel building is just the shell around the interior construction.
Look into steel building manufacturers and contractors. Many times, these people offer various packages to suit your needs (outside and inside details). Yours may be "funky" (ie, a house instead of a business or meeting hall), but they should be able to work with you to get what you want. Furthermore, such buildings can be *very* cheap (a place here in the Phoenix area is offering a 80 x 150 foot building for $50,000 - sure, that is probably without any "extras" - but still damn cheap for the sq footage).
On the other hand, I recently purchased a 1979 Ford Bronco with a 400ci V8 - lotsa room to get in, under, around - everywhere - to play/tinker/fix (and yeah, it needs a *lot* of work yet before I will really feel safe driving it - fortunately, most of it is minor).
Not just electric R/C - but *indoor* electric "slow-fly" R/C has taken off, so to speak - something that would have been deemed almost impossible 10 years ago. Sure, tiny motors have helped, but tiny (and lightweight) batteries have also been a huge improvement allowing for these hobbies (in addition to more electric helicopters, which take way more power to run than a plane)...
Last I heard, the prototype(s?) were sitting at the Smithsonian. I have articles from Popular Science, Popular Mechanics, and Omni that showed this robot off nicely. I remember seeing it working on That's Incredible (or the old Ripley's Believe It or Not). I tried to contact the company (at the time, they were working on an "intelligent highway" system), but they had no clue what I was talking about. From the information I have read, Odetics (and the ODEX-1) originally started out as a DARPA funded project for a robot soldier (in fact, that is what the Pop Mech article showed, a swarm of ODEX bots storming a battlefield, guns ablazin'!) - then, somewhere, it dissappeared or lost funding.
Crazy thing was, they had the prototype fully working (at least, it walked rather well - the video I saw of it showed it climbing out of a pickup, and then "scrunching" itself down to navigate through a doorway, then spreading back out for full walking mode)!
Oh, I never thought it did. I didn't even get the thing notarized - even at the time I drew it up, I had the thought that someone, somewhere, had already thought about this, and that likely a patent already floated around in the PTO's database...
I know that there is a model of a Couger that the battery is only accessible from under the vehicle (saw this when my wife and I went shopping for a car once). I also know that on an older Mercedes convertable (hardtop) my brother-in-law had, the battery was in the trunk (!!!). My wife's Neon (97), the battery is up front, easily accessible (however, it will take an act of god to get the camshaft position sensor fixed, if it ever comes to that - fuel lines and vacuum hoses make it impossible to get to without removing them).
I know I have (somewhere) kicking around in my files a drawing I drew up as an "invention" of contact lens VR devices. Now, I know some of you are thinking "yeah, right" - but it is the truth. I was originally imagining the devices to use hard contacts, with a grid on the contacts driving LCD type elements. I also thought that the electronics would be on the edges of the contacts (which would be mainly driver electronics, pickup "coils" for power supply and datacomms, and position sensors to tell which way the eye was pointed in) and the contacts would have to be larger than normal (to allow for changing pupil size). I also gave thought to using that chemical that widens the pupil to the widest size (which they put into your eyes when you go to get contacts). Finally, I thought about "back-light" shades that you could wear when you wanted full immersion (otherwise it would only be augmented reality - AR). I should have gotten that drawing notarized or something, as it is I only signed and dated it. I didn't think much of it, because I figured others have had similar thoughts before me, and I didn't have the money to patent the idea (I could have easily made a mock up of blown-up scale, or a virtual representation in 3D or something - which would be accepted by the PTO as a model) - plus I figured something like it was already patented. I know I have also posted both here and on k5 about the idea (maybe on Fark, too). I am just glad to see others thinking along this line. I have my doubts on how well it would work (or whether the res would be high enough), but the thing I like most about the idea is that it has the ultimate field-of-view (FOV) possible, by covering your entire pupil, so that you can get true *full* immersion, with peripheral vision and all. If such devices are ever built, they will be amazing to use!
Most people drive automatics over here because most vehicles have them (not too many front-wheel-drive manual vehicles are made anymore). Trucks and SUV's (and sports cars) are the exceptions, where you can get one or the other (or sometimes *only* a manual). Those "in the know" go for a manual, others stick with automatics. I don't know why it is different in Britain, but it is something I have noticed before (maybe it has something to do with kilometers per litre - automatics tend to be more fuel hungry, manuals less so, provided you drive them right)...
At one time, before automatic trannys, there were "semi-automatic" transmissions (I think one of the marketing names was "hydromatic" - but I may be wrong here), that used much of the same components as a fully auto tranny, but the user switched gears (ie, tightened/loosened the bands on the planetarys) via a pushbutton selector, or a simple stick mechanism - no clutching needed, but the user still had to "listen" for the engine to know when to shift...
Not sure how true it is, but from the reading I have done on it, it seems pretty valid, even if our so-called "laws" don't support it...
Now, what if you pay cash? Well, the original still goes to the state MVD.
In exchange for this (it is part of registration), you get your "license to drive" - well, actually, to get a license, you have to surrender your MSO to the MVD.
There is a lot of speculation that it may all be bullshit (like all good conspiracy theories), but look into "Right to Travel" on Google.
Basically, as the theory goes, when licensing for automobiles came about, we traded our freedom to travel for the automobile license, and thus have become slaves (not Free Men) to the State...
The reason robot mower companies don't add any geometry tracking is probably because they don't want to "add a penny more" to the cost of producing the robot - that is one more penny in profit, dontcha know!
Anyhow, you likely wouldn't even need a vision system for most of these robots. A few k of RAM would probably suffice - note:
Lawns are mainly two-dimensional surfaces - so all you really have to keep track of is where you have been in the two dimensional plane. And you don't even have to keep the tracking super accurate - down to meter or half-meter square would probably be accurate enough (hell, lets make it nice - to the square foot). Each "square" of lawn gets one byte in an array. Now, instead of a single wire surrounding the property with only one signal, each "edge" of the property gets its own wire with a specific signal or identifier coming from it (maybe it is a different frequency, or maybe it is sending out a different pulse ID or something). So, the robot is placed in this delimited "grid" (which would be marked up in a GUI to show where the wires "are", how big the yard is, what squares in the tracking memory represent the yard, where the robot is placed in the yard at the start, and what direction it points in), and let to do its random walk or pattern to mow. Using "dead reckoning" by tracking the distance the wheels have moved, it will know what squares (bytes in the array) it has visited, and maybe even the coverage (0=no coverage, 255=fully mowed), based on how often it has visited the square, and from what square(s) around it it has come from. Now, the dead-reckoning system isn't perfect (but is cheaper, currently, than GPS), the robot could get turned around and lost. However, eventually it would come to one of the boundry lines, and it would get a signal telling it what side of the lawn it was on, and it would know that it is turned around in some manner - perhaps it could reset itself, and orient from that position, and continue on from there.
Such a system (which I hereby place under the GPL for Hardware license - and my description is prior art) would be very cheap (most of it is software - the rest of the hardware is probably on board most mower robots, and what little isn't is VERY cheap) - but they won't implement it because it would mean a better product that costs them more to make, and hence less profit for them.
Greedy bastards!
Not perfect, but not a bad system for a CAVE, either...
I think the thing that saddens me is that with all the computer geeks here on /., nobody seems to remember the history of that which they love (at least, I didn't find a mention about this in the first few top pages of comments)...
If you note, I do write in my second sentence "While a pure nanotube fiber has yet to appear". I realize that pure versions of such fibers don't exist. However, I think a disservice was done by not mentioning such research being done to create such fibers right now. Much of the discussion about "space elevators" seem like so much fantasy and sci-fi to the average person - they don't realize that this technology may not be far in the future, that it may be realized sooner than we think or expect...
See this site and this site for more details...
Slashdot - Texas Scientists Spin Carbon Nanotube Fiber
Other promising research:
Slashdot - Scientists Crack Silk's Secret
and
Slashdot - Nanotube Applications Grow And Grow
Maybe they did discuss all this and more at the conference - I would like to hope that these scientists and researchers are aware of what is going on in this far-flung field. I only wish they would have made mention of this stuff in the article for the common man, to show that it wasn't all so much "hooey" - that it is something which may be inevitable, and will happen sooner than we all expect.
We (all of mankind) are rapidly moving in a very funky direction, technology-wise. We have carbon-nanotube fibers. We are looking into other advanced fibers and fiber processes. We have found sea-creatures that make insanely great fiber optic fibers (and with the other stuff, we will probably be able to replicate the process very soon). The gains in communications alone will cause a lot of other gains to be made, because of distributed processing amongst far-apart supercomputing centers that need more bandwidth than they already have (and they have a crapload, but not as much as they want or need). Such fibers may help in the optical-computing dept as well. Remember also the stories of "growing diamonds" - that are so pure they are almost impossible to distinguish from real diamonds - and they have DeBeers quaking at the possibilities to their "markets", maybe destroying them. But these companies don't want the diamonds for prettiness or money (well, they want them for the money, true), but to be able to use them for the substrate of computer chips, instead of silicon, for higher speeds and better heat dissipation.
Couple that with all the other "funky" advances we have seen - we are all being dragged in a very wierd direction, speeding up the computing and learning capacity of all involved (and even if you are at the edge of the network, like most of us are here, and not where the action is, you will still be pulled in)...
I don't know where to go with this - except that our current distopia (and if you don't think we are living in a distopia, one every bit as scary, strange, and awe-inspiring as science fiction can come up with - you haven't been paying attention) is going in a new and strange direction, strangely reminiscent of what the "early-years" (which are only touched on) of Neal Stephenson's "Diamond Age" might have been like.
This is all strange shit, yet very few of us are even seeing it or thinking about the real implications, for some reason...
My parents bought me my first mouse for my Color Computer 2 - it has (yes, present tense, because I still have it) a single red button on the top of the black "soap bar" style body, and the mouse ball is a steel ball bearing of some sort (no rubber covering). It plugged into the joystick port of the CoCo, and had abysmal resolution - but for what I used it for mainly (CoCo Max drawing, and CerComp's windowing system), it was pretty cool...
My hometown is Bakersfield, CA (stop laughing) - I-5 bypasses most of Bako, but you can go through it via highway 99. Now, highway 99 is interesting unto itself. It used to run a different route, on what is now called Union Ave (I think that is right) - or also known as "old 99". There is a turnoff off of 99 as you head north, I think just past the turnoff to I-5. Anyhow, old 99 is one of those things you have to see.
It's a 4 to 6 lane highway, with huge eucalyptus trees flanking the shoulders, and in some areas, the median. Runs north for quite a way, until you get into downtown, then it kinda "dead-ends" or changes into a street. The current 99 actually continues north and reconnects with I-5.
Both 99 and Union Ave (old 99), for long sections, use that same grooved concrete slab construction - and it has to be rode on to - ahem - "appreciate" it.
It is like driving on a "washboard" road with the "bumps" being elongated (several car lengths) - basically the concrete slabs "dip" in the middle, and the ends rise "up". I think it is due to a combination of settling, heat, and large trucks. It isn't super quiet, and it feels like a cheap roller coaster. New 99 is better than old 99, but that is only age difference...
So, I hope that stretch stays experimental. We also have a similar type setup (where it isn't paved with asphalt) here in Phoenix on the I-17 - same grooved concrete, same issue. It might be a better thing if the concrete is laid in a very long ribbon, rather than slabs...
If you really want to do this, you are probably looking into a *very* custom board - I would be thinking something like a camera connected to a microcontroller communicating via 802.11 to the ground. It would have to be completely wireless for what you are wanting to do, unless you are planning on using a very large balloon envelope, or hydrogen (and even then, you would need a large envelope, but not quite as big - probably 12-20 feet in diameter)...
It seems like a lot of people forget that, I know I did until recently. Ink jet printers seem to be a cheap solution - until you realize just how much you are spending on ink.
I own (but no longer use) an Epson Photo Stylus 700, which I bought because I loved the quality of the output when used with the "special" photo paper. I never printer one picture on the paper. I think the greatest thing I *ever* did with the printer was make some nice Thanksgiving party invitations.
It seemed like I was always buying ink - because we rarely used it, but left it turned on. This tended to leave the print heads uncapped (I think they do this on purpose, rather than auto-capping, to sell more ink), and caused the ink to dry out prematurely. But you wanted to leave it on, because it seemed to take forever to "boot" (turn it on, and after minutes of "self-checking" and "cleaning" it would finally be ready. I took a look at how we were printing (rarely, but we wanted good output *now* when we did), what we were printing (most of the time, simple text only stuff, black and white) - and I bought a printer based on that.
I ended up buying a used HP Laserjet 6 (there is a P or something there at the end, too), and a refilled toner cartridge. Total cost: $170.00 - and I have postscript, too. I installed some old 72 pin SIMMs I had lying around to bump the cache up some, and I haven't looked back.
The printer is great - what was really nice was the low page count (25000 pages). I also like the fact that I can use el-cheapo paper in it, and it still looks great (the Epson, on anything under 24lb weight, would "fuzz" - lighter weight paper had more "fur", and the print wouldn't have crisp edges). I also like it that I can leave it on - and then when I want to print out to it, I instantly can - and it just works!
Now, maybe if your job or hobby requires color, an ink jet is what you need to get. But I learned my lesson quick - I don't have *any* need for color. If I want to look at pictures, I look at them on a screen. Just about anyone else can do the same (most people I know have a computer). If I need a print of an image, I will print it in b/w for "checking", then the final can be done at a copy shop or something. I have yet to need to do this, though - but it is the most sensible option, for me.
I will never regret buying that laser printer.
I told him that even if you shred then burn your documents, that THEY have devices that look at the smoke coming out of your fireplace chimney, and can decode the information that is in the smoke patterns. I told him that as the paper burns, the chemicals that make up the ink vs the paper are different, and this difference changes the information matrix encoded in the smoke, and that a detector with sufficient processing power behind it could reconstruct the document data after it was burned.
Boy, am I an evil fuck...
I tend to think they may be smarter than we give them credit for - and they are going to kick our asses into the ground by getting a manned mission to Mars, and soon. How? By using late 1950's technology - otherwise known as Project Orion.
They have *all* the technology to do this, and it would be a major coup for them to lift 1000 or more tons into orbit and beyond, using nuclear bomb/pusher plate propulsion technology.
It may sound crazy, but we came *this* close to doing it ourselves, but political and eco-activist pressure caused the project to be scrapped.
Much of the project documentation, spearheaded by General Atomics, is still to this day considered "top-secret". Mostly because the Orion Project required very small thermonuclear bombs to run on (small H-Bombs - 1KT to 15KT), plus the documentation on how to mitigate the effects of the pusher plate ablation, and the sheilding substances - are useful information on designing H-Bombs with more destructive effects, while at the same time knowing how to protect yourself from the effects. Also, the design of the bombs basically made them directed energy weapons (so that the energy could mostly go into pushing the craft) - so these small bombs required special radiation channelling systems, which were basically one stage in a staged H-Bomb (ie, fission bomb directs energy to hydrogen/deuterium solid core, to compress and "kick" the fusion process off - a three stage bomb could be made by adding another core/channel system, so that the energy from the smaller H-Bomb could kick start a larger reaction). Plus, there was the design/use of (thorium?) as a means to kick the reaction up higher...
But they surely know about all of this - and if our engineers could work this out in the late 50's using computers (IBM 704?) with less power than a low-end Pentium (heck, even that is overkill compared to the 704) - don't you think they could do the same with today's machines?
I would be willing to bet that they are researching the technology. It is doable. They have the land, they have the bomb technology, and they have the will - I doubt they care about any form of fallout...
Yeah, that's where the challenge lies. However, even today, kids have it easy: Provided the thing has a serial port and you can code to it, there are small serial-to-ethernet "converters" available (most of them consist of some form of microcontroller acting as a "go-between" from the ethernet interface and the serial port).
I remember one time in the early 1990's when I picked up a Compaq SLT/386 with 6 meg of RAM, running Caldera's OpenDOS (IIRC). No PCMCIA slots - only a serial port and a parallel port. Since network equipment was still fairly expensive (especially those lovely pocket parallel ethernet adaptors), I looked for a solution.
I ended up creating a funky bit-banging parallel port solution using 4 conductor phone line, dual jack adaptors, and custom wired parallel to RJ-11 plugs. I intended to write software to allow all computers on this network to transmit/receive on it - checking for the status of the lines to avoid colisions, random wait times when there was a busy, etc - I was looking to get 9600 baud on this thing. I managed to build enough dongles for three machines, but I never got around to the coding portion. Always wondered how well (if at all) it would have worked...