The problem is that they're trying to get a chunk of a small market (luxury electric sports cars) from a company largely loved by that market, by getting some sort of marginal top-end acceleration improvement. It'd be like if a startup smartphone manufacturer had though that they could convert Apple fanboys during the Steve Jobs era by producing an iPhone copycat that was pretty much the same as the latest iPhone model but with a processor that was 5% faster.
Really, I think bringing up VR is a great point. If you wanted 3D, and were willing to put something on your head.... why wouldn't you go for VR rather than the TV, so you get full peripheral/360 vision that responds to you turning your head? Why get a TV and glasses? Yes, we all know that the lack of difference between what you see in the eyes, and the lack of it responding to head motion, etc - but that's a content problem, not a device problem. VR devices have the full capability to support a true 3D cinematic experience - but you can't purchase mainstream movies in such a format.
If you have a TV that can provide stereoscopy without need for glasses, with viewer position/eye tracking and so forth, then it has a leg up on VR in that you don't have to wear a device (although lacks the peripheral/360 vision experience). But even in such a case, at least at present, you still don't have the content needed to feed such an experience.
3D TV/cinema keeps failing because... it really just doesn't add that much to the experience when all you're providing is static-position stereoscopy. Why don't we enjoy it more? I don't know. But the evidence is pretty conclusive that the general public just doesn't get much more out of it than they do out of a simple 2d screen. We know from using VR in true 3D environments that VR can add a lot to an experience. But just feeding fixed-position views to each eye doesn't do the trick.
Heck, seawater's almost at parity now. Last I checked, conventional carbonate was only about $5/kg and seawater was about $25/kg. There's some serious convergence going on - we may start using seawater sooner rather than later.
Sorry, but right off: "reserves" don't work that way. A reserves figure is based on the current level of exploration, at the current price point, with current production technologies. It's completely disjoint from the actual amount available. At current price points, the current amount of lithium reserves is 7-8 orders of magnitude different from the amount of lithium reserves if you 5x the price of lithium - or correspondingly improve technology to 1/5th the seawater production cost.
Lithium in seawater is 0.17 ppm. The oceans are 1.4e24kg. That's 2,4e17kg of lithium. In what world is that not enough?
And if it's not enough? Hey, the oceans are only 0.17ppm lithium, but the crust is not only far more massive than the oceans, but is also 20ppm lithium - 2 orders of magnitude more concentrated. And that's 20 ppm *average*. Much, much higher in certain environments than others. But we mine things in order from cheapest to most expensive, so utterly overwhelming majority of lithium will remain untapped forever.
Lithium is not an expensive material. It's so cheap that we use it in ceramics and greases. And it never will be expensive, no matter how much we use. 5x higher cost - who gives a rat's arse? It's not a major portion of li-ion battery costs.
The value of things mined from asteroids at present is zero. There is zero market. There is, however, a market for precious metals and gemstones here. Particularly if they're "exotic"
ISRU is something that could be very useful in the future, but first you have to develop the market for it. Meanwhile, in order to develop that market, people are working to undercut launch costs. Which undercuts the value of said resources being in space.
And ISRU is not nearly as simple as people like to think of it. Let's forget about some sort of "spacedocks" welding together spaceships out of asteroid nickel-iron for now, let's stick to the "easy" stuff, like water for a Marsbound spacecraft. Let's say that what you sinter together is only rock on the outside,but mined permafrost on the inside, so the rock can ablate and protect the sandy ice inside for aerocapture at Earth. Let's ignore how this is harder than just sintering regolith alone. How do you make use of what arrives at Earth? First you have to maneuver a spacecraft to dock with each chunk in LEO (after detecting them with radar), then drag it back to where you're assembling your spacecraft. You then need to drill/cut into each one in space. You then need to put it into a boiler to vaporize out the ice. The boiler needs to maintain a high enough internal pressure that water can exist in a liquid state (otherwise it'll just freeze out as ice and cause you difficulty in getting it into your tanks). So you have a condenser, and pumps connected to your tanks, and you fill them that way. If there's any other volatile chemicals in there that were in the rock (organics, ammonia, etc), you'll need to run your water through reverse osmosis or similar. And of course, any humans involved in this process have to be supported during all this time with consumables from Earth.
Or, you could save yourself all of the time, engineering expense, hardware launches, etc, and just simply launch the water to begin with.
When faced with such decisions, people usually choose the latter.
When it comes to sending mined material to the surface of the Earth, however, the situation is a bit different. Your asteroid still needs a regolith/rock gathering rover, a sinterer, a coilgun, and a power source, all delivered to an asteroid's surface. A project that's probably on the order of a few billion dollars when all is said and done. But all of your other costs are normal Earth costs, everything done on the surface of the planet. You have a chunk of land or sea where precious metal-rich rocks rain down from the sky every X months in a storm of fireballs. Radar tells you where they came in (yes, you actually can see meteors on radar!). A first generation mining operation (whether the landing site would be at land or on sea) would probably work with rocks that weigh just a tonne or less each, and would involve taking them back to a processing facility; however more advanced operations might involve rocks massing hundreds or even thousands of tonnes, and mobile processing facilities. It's not clear what the upper bound on survivable sizes would be, or whether there even really is one. Natural (aka, not optimally shaped, not optimally targeted) meteorites found on Earth get up into the dozens of tonnes each. And you can always include more void space into the sintered shape to increase your surface area to mass ratio and thus entry survivability
An object ejected onto an Earth crossing trajectory does end up on Earth. And regardless you're not going to lower an entire asteroid to the surface of the Earth in one chunk. You have to break it into chunks either way.
There could be advantages, mind you, for moving an asteroid to earth orbit to mine. Manned operations become feasible, and maintenance easier (although one could always just land replacement hardware on a distant asteroid, one presumes that the equipment would not be designed to be built as "one-offs"). You can also have your electromagnetic ejection of sintered chunks be the kick that causes reentry in the desired location - basically, same as with the aerocapture option when ejecting from a distant asteroid, but with a much smaller ellipse. But all of this comes at a real cost: relocating an asteroid. Which represents a very great deal of expense relative to how much rock you get. You're not just talking about kicking chunks back to earth with a coilgun that runs on easily-acquirable electricity; you're talking Big Freaking Rockets. Also, the easiest way to slow down when you get to Earth would be aerobraking. But I'm not sure how comfortable most people would be with you aerobraking a giant rock into orbit.
There is no shockwave. Instead, the energy remains predominantly X-rays. These penetrate into the surface and rapidly convert it to plasma, which sends a powerful burst of plasma, gas and debris in one direction and a powerful shockwave through the object, shattering it and imparting momentum to the fragments.
There's a common myth (sometimes even promoted by scientists who haven't worked on the issue of asteroid deflection) that using a nuclear warhead against an asteroid means that you'll just split it into pieces that are even more destructive than it was before. This is not in accordance with the actual peer-reviewed literature on the subject. There've been a number of projects to do supercomputer simulations, and the results of nuclear deflection - both through standoff "pushing" or direct high-intensity surface impact - are better than even hoped. With very reasonable sized nuclear weapons (compared to the size of the body) you can break asteroids into pieces that tend to be very small, moving at velocities well too fast for it to recoalesce. You can also very readily impart an impulse to significantly change the asteroid's trajectory, whether you destroy it or not.
And even if this wasn't the case, researchers in the field are far from in agreement that one large impactor is worse than multiple smaller impactors; smaller impactors aren't as good at excavating material, and the fact that they're spaced out in time reduces the peak heating for the most "survivable" areas. And of course pieces below a certain size suffer a great deal of ablation and/or decleration and excavate nothing. Larger pieces that airburst, while they flash a large area, usually airburst high enough that surface damage can be limited in the damage zone (see, for example, Chelyabinsk - was that really worse than detonating a 500kT bomb at the surface?), and likewise excavate little to no material.
What are you talking about? What "materials"? I certainly hope you don't mean "lithium", because if so it only exposes how little you know about how lithium is produced. Salar lithium, the current preferred source, isn't "mined", it's produced from brine pumped up in salt flats, sun dried, and the individual salts separated from each other. The undesirable salts are left on the surface. Every year, most of the salars flood, taking the salt with them.
There are various potential lithium sources which are mined, and in the future at times they may prove to be more economical than salars or fill in for an abundance in demand that salars cannot meet. But the ultimate lithium source, the effectively inexhaustible one, is the oceans, and that again just goes back to a brine process. Last I checked (which was long ago), oceanic lithium recovery prices were estimated at about 5x as much as typical salar recovery prices. But even a price like that would hardly impact overall lithium battery prices; it's still cheap, and they just don't use that much lithium.
The grid already uses batteries - not extensively, but more and more each year. And they're going li-ion. My brother in law works for one such company, they just bought their first grid-scale li-ion bank. Li-ion is often coming in cheaper than even flow batteries nowadays. But it still has more to fall before competing with general peaking, it's mainly useful for very short surge loads, voltage maintenance on long lines, things of that nature. Within a decade or so, though, it may be giving peaking a run for its money. It depends on how well the pricing trends hold and progress.
That doesn't mean that li-ion is inherently the future. Other techs (some old, some new) are trying to beat li-ion on price, and may well succeed. But li-ion is is used in the grid, today. And the lower its price falls, the more it'll be used.
Nonsense. Li-ion has a vastly higher power density than lead-acid. Where are you finding PbA that can do ~5kW/kg? Because some li-ions go that high. Most these days are over 1kW/kg. Sustained. PbA is, what, ~200W/kg, for brief periods?
The main reason li-ion hadn't taken over the storage market earlier is because of cost. But that's finally changing.
What reminds me of 5 years ago was all of the naysayers back then laughing at the concept that the Model S would be produced at all, let alone in quantities of nearly 100k per year. That they'd fail to produce them, that they'd fail to find customers, that the whole EV thing was a fad, and Tesla was imminently about to go bankrupt.
Indeed. Tesla has some of the most advanced battery packs on the market. It's pretty dang impressive being able to make a car with that much mass of lithium ion batteries with decade-scale lifespans operating in outdoor conditions and has an order of magnitude lower rate of fires per mile traveled than gasoline vehicles.
Also, as for how they're wired up, in case anyone is curious: individual cells are wired up in parallel "bricks" in large numbers, so that if one cell dies, it has little effect on the brick as a whole (contrast with a laptop battery with 18650 cells just in series - if one goes, the battery is dead). The bricks are connected in series into "sheets" to raise the voltage, and the sheets in turn are connected in series to make up a pack. At least that's how they did it with the Roadster; I assume the Model S is individual. Within each brick, each cell is in its own isolated can; the goal is to prevent propagating failures.
The climate control issue took some time to get right. Early Roadsters suffered from fairly high parasitic drain when the vehicle wasn't plugged in, but they refined the climate control algorithm so that they could more properly maintain the pack temperature without wasting energy. Key to maintaining cell longevity are three main factors: charge/discharge rate, depth of discharge (upper and lower), and temperature. Getting temperature right is very important. As for the other two, the cells aren't charged to their full capacity when the vehicle is at "100%", they're at 90-something percent (at least they were with the Roadster). And on your average drive you only use a very small percent of the pack capacity, so in practice it's an extremely shallow depth of discharge. Both normal driving and overnight charging are low current applications per cell; only fast charging and track duty are relatively high current (but even still you're talking at least half an hour to charge or drain most of the pack).
A slow clap for the person who doesn't realize the difference between "selling units at a loss" and "company undergoing a super-rapid scaleup involving building some of the largest buildings on the planet operating at a loss".
Wikipedia lists it as the largest building in the world by area. And from this:
The original plans for the Tesla Gigafactory call for a facility with a footprint of 5.8 million square feet, on two stories for 10 million square feet of floor space. That would already give it the largest footprint of any building in the world. But that could just end up proving the starting point.
The Gigafactory is being built in a modular fashion, and Tesla is reportedly buying up adjacent plots of land in order to expand the facility. Another 1,200 acres have reportedly already been acquired, with an additional 350 or more on top of that also being looked into. Three modular blocks in addition to the original four could take it up to 24 million square feet of floor space.
In normal-people-units that's 539k m^2, 929k m^2, og 2,2m m^2, respectively.
AR is neat, but I'd really love to use the 3d scanning technique as... well, a 3d scanner, to be able to use my phone to to create 3d models of my environment that I can subsequently load into Blender and the like. Does Tango support this, or is the 3d data only available internally? And when it comes to AR, the ability to insert my own models into the environment would be key to me, not just whatever Google or some app developer happened to think of. For example, if I design a greenhouse, I'd like to be able to put that on my phone tied to GPS coordinates to that I could walk around outside/inside it as if it were already built; that would be amazing for getting a sense of it.
So dang much potential for what can come when you have 3d environment mapping embedded in phones. Imagine what photo/video sharing transforms into when it becomes 3d environment sharing and they're seamlessly merged into each other. You map the world as it really is/was, rather than just pictures of it.
Hmm... I really should check to see if my company's phone budget will provide me a new phone this year....;)
I fully expect the surface to be pitted and covered in regolith/debris.
Which IMHO would actually be a good thing. If we're ever to engage in asteroid mining, the last thing you want is to be having to fracture and pulverize everything yourself; you just want to have to load it (which is hard enough in microgravity) into a mold, sinter it into an ideal reentry shape, and electromagnetically eject it onto an ideal Earth-crossing trajectory for controlled aerobrake/aerocapture (either one-stage: aerocapture straight to the surface, with a larger landing ellipse - or two stage: aerobrake to LEO, then do a timed burn to land within a very narrow ellipse). The latter requires a docking or disposable thruster(s) for each return, while the former can be a simple unguided projectile - so there's a tradeoff between how much land you have to allocate for your returns, and how much those returns cost per unit mass.
A landing ellipse on land doesn't need to be empty of people, but it needs to be able to be emptied of people; returns would arrive in waves, not at an even rate, so it'd be fine to continue using the land for farming, grazing, etc in the interim. An alternative would be to sinter enough voids into the reentry bodies so that they float, then land them in open ocean in a place where currents will concentrate them.
Asteroid mining is anything but a short term option. But it's interesting for the long term. You've got some very high concentrations of precious metals (and some gemstones, like peridot), concentrations that rival or exceed the best mines on Earth, with zero overburden. And products made from them would command a premium on the market - a huge premium in the early days.
I was never really feeling NEOCam. It's not looking for earth-killers, just Tunguska-sized impactors. We've got LSST coming online in the early 2020s which will greatly increase our detection rate, and there will be more in the future. If anything, LSST is significantly better than NEOCam (p.38). A 5-10 years setback (counting for the complimentary nature of the two approaches - NEOCam is IR, LSST visible) is extremely unlikely to equate to "losing New York city" or anything of that nature. A Tunguska-scale impactor is a roughly one-in-400 year event, and overwhelmingly likely to impact few to no people. The odds of one hitting a major metropolitan area in that timeframe, which we could have stopped had we known about it, are one in millions. NeoCAM costs $500m. The delay seems acceptable to me, in an area where space budgets are tight.
If there is an imminent earth-killer out there, it's not in the inner solar system. It's a comet. And neither NEOCam nor LSST would likely see it until it was well on its way toward us. Smaller but still devastating comets? Even later. Hence, defense against large impactors has to be nuclear - as large of warhead(s) as possible, mounted to a storable rocket that can achieve significant delta-V. Nothing else but nuclear has the energy density to deflect in such a short time period, and you don't have the time to engineer your deflection craft from scratch and integrate it onto a stack when time is that short. So if we're serious about planetary defense, that's an approach we need to take.
All of this said: I do kind of look forward to the day when we know the orbits of a large chunk of the ~30-40m impactors and a fair minority of the ~20m (Chelyabinsk-sized) impactors. Because those hit often enough somewhere on the planet (generally very remote) that people could actually travel to see them, like people do with eclipses. And that would be a really neat experience:) And I can totally imagine meteorite hunters prepositioning hardware near the likely strewn field.
The "cleared" term is generally understood to have been poorly worded, with most preferring "gravitationally dominant" to be better. The Trojans are trapped in their location specifically because of Jupiter, not in spite of it.
Then again, if we want to get nitpicky, Jupiter fails the planet definition because the point it orbits (the Sun-Jupiter barycentre) is not inside the sun. They corotate an empty point in space rather than Jupiter simply "orbiting the sun" as the definition requires;)
But that's being nitpicky. I have much bigger complaints with the planet definition than that.
VR's an interesting case, if you can incorporate eye tracking. You only need that sort of resolution in the center of view; in the peripheral you could update only one in every X pixels in each direction, and have the display just fill in their neighbors with the same colour. Or you could only send the low frequency components of a compressed data block for peripheral areas, dropping all of the high frequency data (aka fine detail). Eye tracking wouldn't reduce the number of physical pixels that the screen needs, but it would greatly reduce the required bandwidth, by an order of magnitude. The eye tracking and playback latency would need to respond faster than your eye moves.
However, your format would need to plan for an eye-tracking use case. Aka, you wouldn't want data to be stored as scanlines, you'd want it stored as blocks, and those blocks to have the low-frequency components grouped into the beginning of each block.
Obviously what they're talking about is breaking individual sounds down into their own channels based on assignment to 3d spatial coordinates rather than having individual channels for some arbitrary "standard" set of speakers coming from certain "standardized" directions. Which is IMHO kind of neat. You could have as many speakers as you wanted and put them wherever you wanted. Or perhaps combine face recognition and ultrasonic directional speakers to give a very precise sound direction to each person within the area in front of the tv but nobody else in the building. It also lends itself well to alternative display technologies, such as VR. That's nearly taking the concept of sound capturing as far as it can go; about the only thing they could do more than true positioning would be to be able to capture sounds outside the human hearing range (maybe to get your dogs reacting to movies too?;) )
I sometimes think about what would be the "ultimate imaging format". What can really capture everything capturable and renderable? You can view a camera as a viewing frustrum, and could have any number of cameras viewing the same scene. Their frame timings won't necessarily line up unless they were specifically coordinated to - but then again, with rolling shutters, even different parts of the same frame often don't already. And why is that a bad thing? You could get rid of the concept of frames altogether and bundle pixel data into timestamped packets - so long as you have a reasonable way to describe what sort of angles those pixels are corresponding with, ideally a pattern (angle start/stop/increment for each axis, for example). Your display devices can likewise subsequently do away with the concept of frames and just update new data as soon as they get it - a virtually unlimited number of frames per second.
With multiple cameras, or single cameras with depth sensing, you could have a z-buffer corresponding to cameras' pixel data. So your 3d display or VR headset could recalculate your stereoscopy with respect to where the viewer(s) are sitting, rather than just having a generic naive left/right positioning that sometimes causes discomfort in viewers. You can also render different objects into scenes post-facto (if the format also supports embedded 3d data), which could give producers some neat options to showcase their creativity and viewers to customize their experience. In an extreme case, with sufficient camera coverage, one could attempt to backengineer the full 3d environment from the different camera shots (photogrammetry), allowing for realtime free motion within it.
A variety of other data could also be captured - albeit of questionable utility. An interesting, although very high bandwidth, option would be to store light as full spectra rather than just RGB. So you have the potential for perfect color restoration - even accurate enough for spectral analysis, for whatever that's worth (tetrachromats at the very least would appreciate the extra precision). There's also light outside the human visual range - for most users, the only advantage I can think of for storing it (apart from helping assist any 3d environment backcalculation - aka to see things that might be transparent to visible light but not UV or IR) would be that sometimes there's a lot of infrared but not visible light, and that makes you feel warm; a display device being able to radiate heat on demand, and from specific locations could actually be kind of neat, so long as it stays within comfort ranges. Another thing of questionable utility to capture is light polarization; the only utility I can think of for it is to assist in 3d environment backcalculation. Lastly, taking it to absurd extremes, you could capture data completely unrelated to light and sound, such as various forms of ionizing and non-ionizing radiation. IMHO pretty useless except for physics applications and maybe security cameras at a nuclear facility, but.... Likewise things that aren't even directional - magnetic fields, electric
They had an 8K TV setup at my local Best Buy on an 18 wheeler and actually fooled most of us that it was a window to the outside of the trailer, before they told us it was a TV
Apparently you don't have binocular vision. If someone puts a high-resolution print of a tunnel on a wall, do you try to run into it roadrunner-style?
If you can't stand in front of a 1080p and a 4k screen (55" or even smaller) with the same demo showing on both and not see a huge difference then you are blind or have some sort or brain problem.
No, it means you can differentiate between a difference from the number of pixels, vs. a difference from the quality of the source material and accuracy of colour rendering.
Dongles are bad enough with consumer tech. They're worse for backpacking. You know how the key is to have everything light and small? I had a stove that, while heavier than a lot of simple canister stoves, was able to get more fuel out per canister and thus save you weight overall. At least it could, until they discontinued its canisters. But to make up for it they offered a dongle so that you could connect other canisters! A dongle which was heavier than the stove itself. Argh....
Slashdot Mirror
The problem is that they're trying to get a chunk of a small market (luxury electric sports cars) from a company largely loved by that market, by getting some sort of marginal top-end acceleration improvement. It'd be like if a startup smartphone manufacturer had though that they could convert Apple fanboys during the Steve Jobs era by producing an iPhone copycat that was pretty much the same as the latest iPhone model but with a processor that was 5% faster.
Really, I think bringing up VR is a great point. If you wanted 3D, and were willing to put something on your head.... why wouldn't you go for VR rather than the TV, so you get full peripheral/360 vision that responds to you turning your head? Why get a TV and glasses? Yes, we all know that the lack of difference between what you see in the eyes, and the lack of it responding to head motion, etc - but that's a content problem, not a device problem. VR devices have the full capability to support a true 3D cinematic experience - but you can't purchase mainstream movies in such a format.
If you have a TV that can provide stereoscopy without need for glasses, with viewer position/eye tracking and so forth, then it has a leg up on VR in that you don't have to wear a device (although lacks the peripheral/360 vision experience). But even in such a case, at least at present, you still don't have the content needed to feed such an experience.
3D TV/cinema keeps failing because... it really just doesn't add that much to the experience when all you're providing is static-position stereoscopy. Why don't we enjoy it more? I don't know. But the evidence is pretty conclusive that the general public just doesn't get much more out of it than they do out of a simple 2d screen. We know from using VR in true 3D environments that VR can add a lot to an experience. But just feeding fixed-position views to each eye doesn't do the trick.
Huh, forget 5x ;) I just checked up on how much current prices vs. seawater costs have changed since I last looked.
Current price (Li2CO3): $12-14/kg
Seawater price (Li2CO3): $16-22/kg
Heck, seawater's almost at parity now. Last I checked, conventional carbonate was only about $5/kg and seawater was about $25/kg. There's some serious convergence going on - we may start using seawater sooner rather than later.
Sorry, but right off: "reserves" don't work that way. A reserves figure is based on the current level of exploration, at the current price point, with current production technologies. It's completely disjoint from the actual amount available. At current price points, the current amount of lithium reserves is 7-8 orders of magnitude different from the amount of lithium reserves if you 5x the price of lithium - or correspondingly improve technology to 1/5th the seawater production cost.
Lithium in seawater is 0.17 ppm. The oceans are 1.4e24kg. That's 2,4e17kg of lithium. In what world is that not enough?
And if it's not enough? Hey, the oceans are only 0.17ppm lithium, but the crust is not only far more massive than the oceans, but is also 20ppm lithium - 2 orders of magnitude more concentrated. And that's 20 ppm *average*. Much, much higher in certain environments than others. But we mine things in order from cheapest to most expensive, so utterly overwhelming majority of lithium will remain untapped forever.
Lithium is not an expensive material. It's so cheap that we use it in ceramics and greases. And it never will be expensive, no matter how much we use. 5x higher cost - who gives a rat's arse? It's not a major portion of li-ion battery costs.
The value of things mined from asteroids at present is zero. There is zero market. There is, however, a market for precious metals and gemstones here. Particularly if they're "exotic"
ISRU is something that could be very useful in the future, but first you have to develop the market for it. Meanwhile, in order to develop that market, people are working to undercut launch costs. Which undercuts the value of said resources being in space.
And ISRU is not nearly as simple as people like to think of it. Let's forget about some sort of "spacedocks" welding together spaceships out of asteroid nickel-iron for now, let's stick to the "easy" stuff, like water for a Marsbound spacecraft. Let's say that what you sinter together is only rock on the outside,but mined permafrost on the inside, so the rock can ablate and protect the sandy ice inside for aerocapture at Earth. Let's ignore how this is harder than just sintering regolith alone. How do you make use of what arrives at Earth? First you have to maneuver a spacecraft to dock with each chunk in LEO (after detecting them with radar), then drag it back to where you're assembling your spacecraft. You then need to drill/cut into each one in space. You then need to put it into a boiler to vaporize out the ice. The boiler needs to maintain a high enough internal pressure that water can exist in a liquid state (otherwise it'll just freeze out as ice and cause you difficulty in getting it into your tanks). So you have a condenser, and pumps connected to your tanks, and you fill them that way. If there's any other volatile chemicals in there that were in the rock (organics, ammonia, etc), you'll need to run your water through reverse osmosis or similar. And of course, any humans involved in this process have to be supported during all this time with consumables from Earth.
Or, you could save yourself all of the time, engineering expense, hardware launches, etc, and just simply launch the water to begin with.
When faced with such decisions, people usually choose the latter.
When it comes to sending mined material to the surface of the Earth, however, the situation is a bit different. Your asteroid still needs a regolith/rock gathering rover, a sinterer, a coilgun, and a power source, all delivered to an asteroid's surface. A project that's probably on the order of a few billion dollars when all is said and done. But all of your other costs are normal Earth costs, everything done on the surface of the planet. You have a chunk of land or sea where precious metal-rich rocks rain down from the sky every X months in a storm of fireballs. Radar tells you where they came in (yes, you actually can see meteors on radar!). A first generation mining operation (whether the landing site would be at land or on sea) would probably work with rocks that weigh just a tonne or less each, and would involve taking them back to a processing facility; however more advanced operations might involve rocks massing hundreds or even thousands of tonnes, and mobile processing facilities. It's not clear what the upper bound on survivable sizes would be, or whether there even really is one. Natural (aka, not optimally shaped, not optimally targeted) meteorites found on Earth get up into the dozens of tonnes each. And you can always include more void space into the sintered shape to increase your surface area to mass ratio and thus entry survivability
An object ejected onto an Earth crossing trajectory does end up on Earth. And regardless you're not going to lower an entire asteroid to the surface of the Earth in one chunk. You have to break it into chunks either way.
There could be advantages, mind you, for moving an asteroid to earth orbit to mine. Manned operations become feasible, and maintenance easier (although one could always just land replacement hardware on a distant asteroid, one presumes that the equipment would not be designed to be built as "one-offs"). You can also have your electromagnetic ejection of sintered chunks be the kick that causes reentry in the desired location - basically, same as with the aerocapture option when ejecting from a distant asteroid, but with a much smaller ellipse. But all of this comes at a real cost: relocating an asteroid. Which represents a very great deal of expense relative to how much rock you get. You're not just talking about kicking chunks back to earth with a coilgun that runs on easily-acquirable electricity; you're talking Big Freaking Rockets. Also, the easiest way to slow down when you get to Earth would be aerobraking. But I'm not sure how comfortable most people would be with you aerobraking a giant rock into orbit.
That first link was erroneous; I was trying to link the wikipedia article on the world's largest buildings, as you can see from context.
The second link, quoted in my post, is specifically about the gigafactory.
There is no shockwave. Instead, the energy remains predominantly X-rays. These penetrate into the surface and rapidly convert it to plasma, which sends a powerful burst of plasma, gas and debris in one direction and a powerful shockwave through the object, shattering it and imparting momentum to the fragments.
There's a common myth (sometimes even promoted by scientists who haven't worked on the issue of asteroid deflection) that using a nuclear warhead against an asteroid means that you'll just split it into pieces that are even more destructive than it was before. This is not in accordance with the actual peer-reviewed literature on the subject. There've been a number of projects to do supercomputer simulations, and the results of nuclear deflection - both through standoff "pushing" or direct high-intensity surface impact - are better than even hoped. With very reasonable sized nuclear weapons (compared to the size of the body) you can break asteroids into pieces that tend to be very small, moving at velocities well too fast for it to recoalesce. You can also very readily impart an impulse to significantly change the asteroid's trajectory, whether you destroy it or not.
And even if this wasn't the case, researchers in the field are far from in agreement that one large impactor is worse than multiple smaller impactors; smaller impactors aren't as good at excavating material, and the fact that they're spaced out in time reduces the peak heating for the most "survivable" areas. And of course pieces below a certain size suffer a great deal of ablation and/or decleration and excavate nothing. Larger pieces that airburst, while they flash a large area, usually airburst high enough that surface damage can be limited in the damage zone (see, for example, Chelyabinsk - was that really worse than detonating a 500kT bomb at the surface?), and likewise excavate little to no material.
Jeebuz buring all of those dinosaur fossils in the ground and giving them suspiciously old radioisotope ratios was the best prank ever.
What are you talking about? What "materials"? I certainly hope you don't mean "lithium", because if so it only exposes how little you know about how lithium is produced. Salar lithium, the current preferred source, isn't "mined", it's produced from brine pumped up in salt flats, sun dried, and the individual salts separated from each other. The undesirable salts are left on the surface. Every year, most of the salars flood, taking the salt with them.
There are various potential lithium sources which are mined, and in the future at times they may prove to be more economical than salars or fill in for an abundance in demand that salars cannot meet. But the ultimate lithium source, the effectively inexhaustible one, is the oceans, and that again just goes back to a brine process. Last I checked (which was long ago), oceanic lithium recovery prices were estimated at about 5x as much as typical salar recovery prices. But even a price like that would hardly impact overall lithium battery prices; it's still cheap, and they just don't use that much lithium.
No, you simply have no clue about the topic.
The grid already uses batteries - not extensively, but more and more each year. And they're going li-ion. My brother in law works for one such company, they just bought their first grid-scale li-ion bank. Li-ion is often coming in cheaper than even flow batteries nowadays. But it still has more to fall before competing with general peaking, it's mainly useful for very short surge loads, voltage maintenance on long lines, things of that nature. Within a decade or so, though, it may be giving peaking a run for its money. It depends on how well the pricing trends hold and progress.
That doesn't mean that li-ion is inherently the future. Other techs (some old, some new) are trying to beat li-ion on price, and may well succeed. But li-ion is is used in the grid, today. And the lower its price falls, the more it'll be used.
Nonsense. Li-ion has a vastly higher power density than lead-acid. Where are you finding PbA that can do ~5kW/kg? Because some li-ions go that high. Most these days are over 1kW/kg. Sustained. PbA is, what, ~200W/kg, for brief periods?
The main reason li-ion hadn't taken over the storage market earlier is because of cost. But that's finally changing.
What reminds me of 5 years ago was all of the naysayers back then laughing at the concept that the Model S would be produced at all, let alone in quantities of nearly 100k per year. That they'd fail to produce them, that they'd fail to find customers, that the whole EV thing was a fad, and Tesla was imminently about to go bankrupt.
My, how things change. Or, not.
I remember when that happened, after the great wind spill of '11. Those poor sailboats. :(
Indeed. Tesla has some of the most advanced battery packs on the market. It's pretty dang impressive being able to make a car with that much mass of lithium ion batteries with decade-scale lifespans operating in outdoor conditions and has an order of magnitude lower rate of fires per mile traveled than gasoline vehicles.
Also, as for how they're wired up, in case anyone is curious: individual cells are wired up in parallel "bricks" in large numbers, so that if one cell dies, it has little effect on the brick as a whole (contrast with a laptop battery with 18650 cells just in series - if one goes, the battery is dead). The bricks are connected in series into "sheets" to raise the voltage, and the sheets in turn are connected in series to make up a pack. At least that's how they did it with the Roadster; I assume the Model S is individual. Within each brick, each cell is in its own isolated can; the goal is to prevent propagating failures.
The climate control issue took some time to get right. Early Roadsters suffered from fairly high parasitic drain when the vehicle wasn't plugged in, but they refined the climate control algorithm so that they could more properly maintain the pack temperature without wasting energy. Key to maintaining cell longevity are three main factors: charge/discharge rate, depth of discharge (upper and lower), and temperature. Getting temperature right is very important. As for the other two, the cells aren't charged to their full capacity when the vehicle is at "100%", they're at 90-something percent (at least they were with the Roadster). And on your average drive you only use a very small percent of the pack capacity, so in practice it's an extremely shallow depth of discharge. Both normal driving and overnight charging are low current applications per cell; only fast charging and track duty are relatively high current (but even still you're talking at least half an hour to charge or drain most of the pack).
A slow clap for the person who doesn't realize the difference between "selling units at a loss" and "company undergoing a super-rapid scaleup involving building some of the largest buildings on the planet operating at a loss".
Wikipedia lists it as the largest building in the world by area. And from this:
The original plans for the Tesla Gigafactory call for a facility with a footprint of 5.8 million square feet, on two stories for 10 million square feet of floor space. That would already give it the largest footprint of any building in the world. But that could just end up proving the starting point.
The Gigafactory is being built in a modular fashion, and Tesla is reportedly buying up adjacent plots of land in order to expand the facility. Another 1,200 acres have reportedly already been acquired, with an additional 350 or more on top of that also being looked into. Three modular blocks in addition to the original four could take it up to 24 million square feet of floor space.
In normal-people-units that's 539k m^2, 929k m^2, og 2,2m m^2, respectively.
AR is neat, but I'd really love to use the 3d scanning technique as... well, a 3d scanner, to be able to use my phone to to create 3d models of my environment that I can subsequently load into Blender and the like. Does Tango support this, or is the 3d data only available internally? And when it comes to AR, the ability to insert my own models into the environment would be key to me, not just whatever Google or some app developer happened to think of. For example, if I design a greenhouse, I'd like to be able to put that on my phone tied to GPS coordinates to that I could walk around outside/inside it as if it were already built; that would be amazing for getting a sense of it.
So dang much potential for what can come when you have 3d environment mapping embedded in phones. Imagine what photo/video sharing transforms into when it becomes 3d environment sharing and they're seamlessly merged into each other. You map the world as it really is/was, rather than just pictures of it.
Hmm... I really should check to see if my company's phone budget will provide me a new phone this year.... ;)
I fully expect the surface to be pitted and covered in regolith/debris.
Which IMHO would actually be a good thing. If we're ever to engage in asteroid mining, the last thing you want is to be having to fracture and pulverize everything yourself; you just want to have to load it (which is hard enough in microgravity) into a mold, sinter it into an ideal reentry shape, and electromagnetically eject it onto an ideal Earth-crossing trajectory for controlled aerobrake/aerocapture (either one-stage: aerocapture straight to the surface, with a larger landing ellipse - or two stage: aerobrake to LEO, then do a timed burn to land within a very narrow ellipse). The latter requires a docking or disposable thruster(s) for each return, while the former can be a simple unguided projectile - so there's a tradeoff between how much land you have to allocate for your returns, and how much those returns cost per unit mass.
A landing ellipse on land doesn't need to be empty of people, but it needs to be able to be emptied of people; returns would arrive in waves, not at an even rate, so it'd be fine to continue using the land for farming, grazing, etc in the interim. An alternative would be to sinter enough voids into the reentry bodies so that they float, then land them in open ocean in a place where currents will concentrate them.
Asteroid mining is anything but a short term option. But it's interesting for the long term. You've got some very high concentrations of precious metals (and some gemstones, like peridot), concentrations that rival or exceed the best mines on Earth, with zero overburden. And products made from them would command a premium on the market - a huge premium in the early days.
I was never really feeling NEOCam. It's not looking for earth-killers, just Tunguska-sized impactors. We've got LSST coming online in the early 2020s which will greatly increase our detection rate, and there will be more in the future. If anything, LSST is significantly better than NEOCam (p.38). A 5-10 years setback (counting for the complimentary nature of the two approaches - NEOCam is IR, LSST visible) is extremely unlikely to equate to "losing New York city" or anything of that nature. A Tunguska-scale impactor is a roughly one-in-400 year event, and overwhelmingly likely to impact few to no people. The odds of one hitting a major metropolitan area in that timeframe, which we could have stopped had we known about it, are one in millions. NeoCAM costs $500m. The delay seems acceptable to me, in an area where space budgets are tight.
If there is an imminent earth-killer out there, it's not in the inner solar system. It's a comet. And neither NEOCam nor LSST would likely see it until it was well on its way toward us. Smaller but still devastating comets? Even later. Hence, defense against large impactors has to be nuclear - as large of warhead(s) as possible, mounted to a storable rocket that can achieve significant delta-V. Nothing else but nuclear has the energy density to deflect in such a short time period, and you don't have the time to engineer your deflection craft from scratch and integrate it onto a stack when time is that short. So if we're serious about planetary defense, that's an approach we need to take.
All of this said: I do kind of look forward to the day when we know the orbits of a large chunk of the ~30-40m impactors and a fair minority of the ~20m (Chelyabinsk-sized) impactors. Because those hit often enough somewhere on the planet (generally very remote) that people could actually travel to see them, like people do with eclipses. And that would be a really neat experience :) And I can totally imagine meteorite hunters prepositioning hardware near the likely strewn field.
The "cleared" term is generally understood to have been poorly worded, with most preferring "gravitationally dominant" to be better. The Trojans are trapped in their location specifically because of Jupiter, not in spite of it.
Then again, if we want to get nitpicky, Jupiter fails the planet definition because the point it orbits (the Sun-Jupiter barycentre) is not inside the sun. They corotate an empty point in space rather than Jupiter simply "orbiting the sun" as the definition requires ;)
But that's being nitpicky. I have much bigger complaints with the planet definition than that.
VR's an interesting case, if you can incorporate eye tracking. You only need that sort of resolution in the center of view; in the peripheral you could update only one in every X pixels in each direction, and have the display just fill in their neighbors with the same colour. Or you could only send the low frequency components of a compressed data block for peripheral areas, dropping all of the high frequency data (aka fine detail). Eye tracking wouldn't reduce the number of physical pixels that the screen needs, but it would greatly reduce the required bandwidth, by an order of magnitude. The eye tracking and playback latency would need to respond faster than your eye moves.
However, your format would need to plan for an eye-tracking use case. Aka, you wouldn't want data to be stored as scanlines, you'd want it stored as blocks, and those blocks to have the low-frequency components grouped into the beginning of each block.
Obviously what they're talking about is breaking individual sounds down into their own channels based on assignment to 3d spatial coordinates rather than having individual channels for some arbitrary "standard" set of speakers coming from certain "standardized" directions. Which is IMHO kind of neat. You could have as many speakers as you wanted and put them wherever you wanted. Or perhaps combine face recognition and ultrasonic directional speakers to give a very precise sound direction to each person within the area in front of the tv but nobody else in the building. It also lends itself well to alternative display technologies, such as VR. That's nearly taking the concept of sound capturing as far as it can go; about the only thing they could do more than true positioning would be to be able to capture sounds outside the human hearing range (maybe to get your dogs reacting to movies too? ;) )
I sometimes think about what would be the "ultimate imaging format". What can really capture everything capturable and renderable? You can view a camera as a viewing frustrum, and could have any number of cameras viewing the same scene. Their frame timings won't necessarily line up unless they were specifically coordinated to - but then again, with rolling shutters, even different parts of the same frame often don't already. And why is that a bad thing? You could get rid of the concept of frames altogether and bundle pixel data into timestamped packets - so long as you have a reasonable way to describe what sort of angles those pixels are corresponding with, ideally a pattern (angle start/stop/increment for each axis, for example). Your display devices can likewise subsequently do away with the concept of frames and just update new data as soon as they get it - a virtually unlimited number of frames per second.
With multiple cameras, or single cameras with depth sensing, you could have a z-buffer corresponding to cameras' pixel data. So your 3d display or VR headset could recalculate your stereoscopy with respect to where the viewer(s) are sitting, rather than just having a generic naive left/right positioning that sometimes causes discomfort in viewers. You can also render different objects into scenes post-facto (if the format also supports embedded 3d data), which could give producers some neat options to showcase their creativity and viewers to customize their experience. In an extreme case, with sufficient camera coverage, one could attempt to backengineer the full 3d environment from the different camera shots (photogrammetry), allowing for realtime free motion within it.
A variety of other data could also be captured - albeit of questionable utility. An interesting, although very high bandwidth, option would be to store light as full spectra rather than just RGB. So you have the potential for perfect color restoration - even accurate enough for spectral analysis, for whatever that's worth (tetrachromats at the very least would appreciate the extra precision). There's also light outside the human visual range - for most users, the only advantage I can think of for storing it (apart from helping assist any 3d environment backcalculation - aka to see things that might be transparent to visible light but not UV or IR) would be that sometimes there's a lot of infrared but not visible light, and that makes you feel warm; a display device being able to radiate heat on demand, and from specific locations could actually be kind of neat, so long as it stays within comfort ranges. Another thing of questionable utility to capture is light polarization; the only utility I can think of for it is to assist in 3d environment backcalculation. Lastly, taking it to absurd extremes, you could capture data completely unrelated to light and sound, such as various forms of ionizing and non-ionizing radiation. IMHO pretty useless except for physics applications and maybe security cameras at a nuclear facility, but.... Likewise things that aren't even directional - magnetic fields, electric
Apparently you don't have binocular vision. If someone puts a high-resolution print of a tunnel on a wall, do you try to run into it roadrunner-style?
No, it means you can differentiate between a difference from the number of pixels, vs. a difference from the quality of the source material and accuracy of colour rendering.
Dongles are bad enough with consumer tech. They're worse for backpacking. You know how the key is to have everything light and small? I had a stove that, while heavier than a lot of simple canister stoves, was able to get more fuel out per canister and thus save you weight overall. At least it could, until they discontinued its canisters. But to make up for it they offered a dongle so that you could connect other canisters! A dongle which was heavier than the stove itself. Argh....