See, that's the problem with you people. First you insist that 3d printers can only produce low quality plastic junk, and then when faced with the reality that they actually can produce superb products, you treat them as if they're just some sort of incremental change on earlier systems, as if 3d printers are just some form of advanced CNC system.
They are not.
3d printing is a completely different technology to milling, moulding, die forming, etc - your "traditional" manufacturing techniques. You could say that new 3d printers are expanding on the capabilties of 3d printers, but well, that's about the most meaningless truism one could make about technology.
3d printing is not some universal make-all technology. But it has provided for a way to fill a previously poorly met niche: rapid low-volume production of small objects with incredibly complex, nearly arbitrary geometry. For people who need this capability, the rise of 3d printing has been invaluable. Your continual attempts to downplay it as "printing plastic junk" as you'd been doing, or "extrapolation of current capabilities" as you're doing now, are ridiculous.
The current generation of high end 3d printers are very rewarding, and I say this as someone who's used them. The print quality is superb. And no, you don't generally go to existing metal shops which have a CNC miller, you need a shop that has a 3d printer, the two tools are used for different roles. If you have some big part with simpler geometry that you need, you use CNC milling. If you have some small part with complex geometry, you use 3d printing. 3d printing services are already plentiful on the net. Today. They're not at every corner "FedEx Office" today, but they're certainly heading in that direction.
And please, let's stop with the "they're not useful" stuff. I've used them. I know others who have used them. I know multiple companies that use them. Just because you don't has no relevance.
Why is it that this day in age we're still needing to point out that Not All 3D Printing Is Makerbots? If you wanted a professional looking poster printed out would you try to do it on a cheapo home inkjet? You're declaring the term "3d printing" as only applying to "crappy plastic extruders". The "industrial ones that can print in metal", as you put it, are *also* 3d printers. And home users *can* get prints from them, there's lots of online print services. My personal favorite is iMaterialize, as I like their materials and finish selection.
What part of "3D printers can currently only produce mediocre quality plastic representations" do you not understand?
What part of "you're completely and utterly wrong" do you not understand? Makerbots != All 3d printers. There are excellent sintering and spraying metal printers out there, and lots of 3d printing services that 3d-print the initial mold for lost wax casting. People churn out huge numbers of 3d-printed metal parts. Brass, gold, silver, bronze, steel, titanium, you name it. I've purchased such parts myself. I even know of a rocketry startup that's 3d printing inconel aerospikes with a laser sintering system, having it churn out one version after the next in evolutionary sequence so they can maximize the real-world performance.
Welding services do metal just fine (actually, more than "fine", I'd say "superbly"). I designed and printed out a detailed brass medallion as a gift before and it came out just beautiful. Most services these days just use lost wax casting, but there's also metal sintering, and the newest player is thermal spraying, such as laser spraying.
There's also some tremendous software needs in this regard. Sure, the basics are just being able to take scans and print them (although as we all know there's nothing simple about that). But just on the UI side, there's going to be good demand for filters / tools that can repair common types of damage / wear on the part. And many users are going to want to be able to customize their part, so you need everyday-user-friendly 3d modelling that's ideally possible to do straight on the smartphone. Then what about the people who don't have a part in front of them to scan? Some may want it through photos of something they don't have, which means better photogrammetry software that takes fewer images and can properly understand shine/reflection and transparency. And some may even want to be able to sketch it on a piece of paper and scan that in, only having to do tweaks to their sketch, and with the software trying to make a reasonable interpretation of their awful drawing skills.
And of course, regardless of how the models are made, we're going to want a massive collaborative model database built-in to any such apps, designed to encourage people to properly tag and share their models.
Any sort of future where this sort of thing becomes ubiquitous is going to rely on rapid production and delivery - not just of unprintable parts (as above), but also, say, prototype parts. Because let's face it most people aren't going to be sure if what they're printing is really going to fit / meet their needs. It may take a couple iterations, and they may want to make earlier prototype iterations be hollow / out of a cheaper material / with poorer finish / etc until they're satisfied that it meets their needs. If they have to wait a couple weeks between iterations, obviously that's going to suck.
This is of course just a first step. The ultimate would be to be able to get a precise reading *within* an object, not just "what your sensor can read from where you stand". And not just geometry, but composition as well. And not just geometry and composition, but also discontinuities - for example, a screw in a socket may look like a continuous piece of metal even at high resolution, but there's a very important discontinuity there that you need to detect. And of course your 3d print service either needs to be able to print out discontinuities in place (I can envision methods for this), or automatically assemble things from smaller parts.
Note: "gluegun" style printers will never be able to achieve this sort of task flexibility; even basic sinterers would have trouble. The only type I can envision managing it are thermal spraying-based printers; for example, there's at least one laser spraying printer / CNC combo on the market that I'm aware of. With such a printer you're almost material unlimited, being able to custom dial-in a particle impact velocity and temperature over a wide range. The same device could even coat, paint, sandblast, whatever your part when it's done.
What I have trouble envisioning any time soon is a scanner that can read the fine detail on lithographically printed devices, or any normal 3d printer that can print high-end chips. Certain macroscopic products which are produced by complex / sensitive chemical processes might be difficult as well. On the other hand, maybe far enough in the future the scanner might at least recognize the part that can't be printed, and then perhaps the print service could automatically order (say, via Amazon BloodDrone Prime) a delivery of whatever it can't print out, then insert and integrate the unprintable part into the print automatically. But we're looking at pretty far into the future here.
I live in Iceland. It costs a fortune to get specialty parts here. Just ignoring that some parts aren't available at all, and the waits can be huge. Let's look at my needs that I've had to special order in the past month.
1) Welder nozzle: Part of the nozzle at the end of my MIG welding torch was broken, a little metal piece (no, not a consumable like the copper contact tip). Fine, just replace it right? Well, except for the fact that this welder is from the 1960s, the part is simply no longer available, even on used markets; not only did I search, but I contacted the manufacturer who told me as much as well. So instead I had to buy a euro-torch adapter and a whole new torch. Cost, about $180 USD. Plus a couple weeks searching, then the purchase.
2) Refrigerator compartment cover: The clear plastic door on the cheese compartment in my refrigerator came loose, feel on the floor, and cracked beyond repair. Just a little piece of plastic, but the cost for a replacement? About $50 USD and several weeks wait.
3) Torch feed tube: #1 turned out not to be enough! Because due to the differing shapes of the connectors the 120mm feed tube that came with #1 turned out to be insufficient, I need a longerfeed tube to prevent birdsnesting. Except that no store here sells them. I found some online in the UK (180mm) after a day or two of looking around for a reasonable deal, but it's been a pain to get them sent out, I ordered them a few days ago... hopefully they'll get sent out today and arrive in a couple of weeks. After shipping and customs getting their cut? About $50 USD for two little brass tubes a couple millimeters in diameter.
In the case of #3, I was seriously considering designing them in blender and then 3d printing them (I even took measurements for the design). If the ones from the UK don't fit, I may still have to.
And here I am not doing any project that's focused around lots of smaller parts. I can only imagine if I was, say, fixing up an older car or something like that - the ability to scan in a rusty or cracked part, maybe apply a couple filters to clean up the scan (remove any rust growth/pitting, seal cracks, etc) and have a local 3d print service print it out would be incredible. I've tried using photogrammetry for these sorts of purposes but it's just so limited - it has trouble working in the best of conditions, if you face any sort of shine or transparency - which, for example, all three of my above examples had - it's unlikely to work.
Not just the life support, either. Every piece of control hardware in the cockpit is a potential failure source. On ground you can have any number of backups sitting around, ready to switch to a different one at the flip of a switch.
People need to stop picturing all dinosaurs as looking like some kind of leathery reptiles. I mean, we not only know now that velociraptor was feathered, but even how many secondary wing feathers it had (14). Jurassic park would have maybe not been as scary had their "raptors" looked like this.;)
Your analogy was to a ridiculously complex multi-year incredibly expensive infiltration operation by the NSA. Don't blame me if your analogy was ridiculous. If you want a different analogy, go ahead and choose one.
Facts are, UAVs are not commercial airplanes, and are not designed to have the safety features of commercial airplanes. Just ignoring that most of the accidents were early on when they were rushed to the battlefield, you're comparing apples to oranges. You might as well argue that helicopters are deathtraps because kids flying RC toy helicopters crash them a lot.
Air France Flight 296 was crew error. And the body count was "3". The only thing the fly-by-wire system did was prevent their attempt to pull up before they had the speed to actually pull up; the plane was already doomed because they were too close to the trees before they realized their mistake.
Propulsion-only automated flight control systems have already been demonstrated and are expected to be integrated into next-generation commercial aircraft.
Oh wait, you're talking not just about jamming the datalink signal, but also ILS and LPS? Nice jammer you've got. So, how exactly does remote control come into play? If you screw up the signal, you can already crash the plane, pilot or none. There was a pilot on THY726. The plane still crashed. They trust ILS and LPS because the data is usually more accurate than their own senses.
(Plus, it's not even remotely hard to picture additional workarounds that could be added, such as radar, laser, microwave, or ultrasound distance sensors.)
The simple facts are, accident rates on planes have gone way down as the amount of automation has gone way up. Computers are simply more reliable than pilots. The fact that most crashes come down to pilot error is a big problem, despite how you present it (ground error being probably the next most common). A computer will always follow the checklist for the right response to a given problem ("some system failure occurred that should have been handled safely by a well-trained pilot and somehow wasn't."). Pilots screw up.
As for military drones, please, they're not at all comparable to commercial aircraft. Most don't even have a flight radar or anti-collision systems. They were designed to be gotten out to the battlefield as soon as possible, in as little development time and unit cost as possible - and they're not exactly flying a regular London to NYC route either.
Or 100% reliability for starters: it's 2015 and yet still you can easily lose your WiFi connection while sitting only 3 feet away from the access point.
Yeah, too bad there's not something called "autopilot" that could take over if communications are lost. Sadly, as we all know, the instant communications are lost, the plane will explode.
Anyone want to guarantee 100% perfect security for ANY wireless communication? Because if we have remotely piloted airliners (either because there's no pilot, or the pilot is suicidal) someone WILL hack into it.
You have a large high power radio transmission tower (or the ability to jump airgaps into ATC transmission networks), and you also have the ability to break whatever message authentication / validation encryption system that such a remote piloting system would use? Cool! Followup question: are you the NSA, or is there another organization that I need to be worried about? Last question: wouldn't it have been about a hundred times easier for you just to buy a couple MANPADs?
Slashdot Mirror
See, that's the problem with you people. First you insist that 3d printers can only produce low quality plastic junk, and then when faced with the reality that they actually can produce superb products, you treat them as if they're just some sort of incremental change on earlier systems, as if 3d printers are just some form of advanced CNC system.
They are not.
3d printing is a completely different technology to milling, moulding, die forming, etc - your "traditional" manufacturing techniques. You could say that new 3d printers are expanding on the capabilties of 3d printers, but well, that's about the most meaningless truism one could make about technology.
3d printing is not some universal make-all technology. But it has provided for a way to fill a previously poorly met niche: rapid low-volume production of small objects with incredibly complex, nearly arbitrary geometry. For people who need this capability, the rise of 3d printing has been invaluable. Your continual attempts to downplay it as "printing plastic junk" as you'd been doing, or "extrapolation of current capabilities" as you're doing now, are ridiculous.
The current generation of high end 3d printers are very rewarding, and I say this as someone who's used them. The print quality is superb. And no, you don't generally go to existing metal shops which have a CNC miller, you need a shop that has a 3d printer, the two tools are used for different roles. If you have some big part with simpler geometry that you need, you use CNC milling. If you have some small part with complex geometry, you use 3d printing. 3d printing services are already plentiful on the net. Today. They're not at every corner "FedEx Office" today, but they're certainly heading in that direction.
And please, let's stop with the "they're not useful" stuff. I've used them. I know others who have used them. I know multiple companies that use them. Just because you don't has no relevance.
Why is it that this day in age we're still needing to point out that Not All 3D Printing Is Makerbots? If you wanted a professional looking poster printed out would you try to do it on a cheapo home inkjet? You're declaring the term "3d printing" as only applying to "crappy plastic extruders". The "industrial ones that can print in metal", as you put it, are *also* 3d printers. And home users *can* get prints from them, there's lots of online print services. My personal favorite is iMaterialize, as I like their materials and finish selection.
What part of "you're completely and utterly wrong" do you not understand? Makerbots != All 3d printers. There are excellent sintering and spraying metal printers out there, and lots of 3d printing services that 3d-print the initial mold for lost wax casting. People churn out huge numbers of 3d-printed metal parts. Brass, gold, silver, bronze, steel, titanium, you name it. I've purchased such parts myself. I even know of a rocketry startup that's 3d printing inconel aerospikes with a laser sintering system, having it churn out one version after the next in evolutionary sequence so they can maximize the real-world performance.
Welding services do metal just fine (actually, more than "fine", I'd say "superbly"). I designed and printed out a detailed brass medallion as a gift before and it came out just beautiful. Most services these days just use lost wax casting, but there's also metal sintering, and the newest player is thermal spraying, such as laser spraying.
There's also some tremendous software needs in this regard. Sure, the basics are just being able to take scans and print them (although as we all know there's nothing simple about that). But just on the UI side, there's going to be good demand for filters / tools that can repair common types of damage / wear on the part. And many users are going to want to be able to customize their part, so you need everyday-user-friendly 3d modelling that's ideally possible to do straight on the smartphone. Then what about the people who don't have a part in front of them to scan? Some may want it through photos of something they don't have, which means better photogrammetry software that takes fewer images and can properly understand shine/reflection and transparency. And some may even want to be able to sketch it on a piece of paper and scan that in, only having to do tweaks to their sketch, and with the software trying to make a reasonable interpretation of their awful drawing skills.
And of course, regardless of how the models are made, we're going to want a massive collaborative model database built-in to any such apps, designed to encourage people to properly tag and share their models.
Any sort of future where this sort of thing becomes ubiquitous is going to rely on rapid production and delivery - not just of unprintable parts (as above), but also, say, prototype parts. Because let's face it most people aren't going to be sure if what they're printing is really going to fit / meet their needs. It may take a couple iterations, and they may want to make earlier prototype iterations be hollow / out of a cheaper material / with poorer finish / etc until they're satisfied that it meets their needs. If they have to wait a couple weeks between iterations, obviously that's going to suck.
This is of course just a first step. The ultimate would be to be able to get a precise reading *within* an object, not just "what your sensor can read from where you stand". And not just geometry, but composition as well. And not just geometry and composition, but also discontinuities - for example, a screw in a socket may look like a continuous piece of metal even at high resolution, but there's a very important discontinuity there that you need to detect. And of course your 3d print service either needs to be able to print out discontinuities in place (I can envision methods for this), or automatically assemble things from smaller parts.
Note: "gluegun" style printers will never be able to achieve this sort of task flexibility; even basic sinterers would have trouble. The only type I can envision managing it are thermal spraying-based printers; for example, there's at least one laser spraying printer / CNC combo on the market that I'm aware of. With such a printer you're almost material unlimited, being able to custom dial-in a particle impact velocity and temperature over a wide range. The same device could even coat, paint, sandblast, whatever your part when it's done.
What I have trouble envisioning any time soon is a scanner that can read the fine detail on lithographically printed devices, or any normal 3d printer that can print high-end chips. Certain macroscopic products which are produced by complex / sensitive chemical processes might be difficult as well. On the other hand, maybe far enough in the future the scanner might at least recognize the part that can't be printed, and then perhaps the print service could automatically order (say, via Amazon BloodDrone Prime) a delivery of whatever it can't print out, then insert and integrate the unprintable part into the print automatically. But we're looking at pretty far into the future here.
Not everyone has the same needs as you.
I live in Iceland. It costs a fortune to get specialty parts here. Just ignoring that some parts aren't available at all, and the waits can be huge. Let's look at my needs that I've had to special order in the past month.
1) Welder nozzle: Part of the nozzle at the end of my MIG welding torch was broken, a little metal piece (no, not a consumable like the copper contact tip). Fine, just replace it right? Well, except for the fact that this welder is from the 1960s, the part is simply no longer available, even on used markets; not only did I search, but I contacted the manufacturer who told me as much as well. So instead I had to buy a euro-torch adapter and a whole new torch. Cost, about $180 USD. Plus a couple weeks searching, then the purchase.
2) Refrigerator compartment cover: The clear plastic door on the cheese compartment in my refrigerator came loose, feel on the floor, and cracked beyond repair. Just a little piece of plastic, but the cost for a replacement? About $50 USD and several weeks wait.
3) Torch feed tube: #1 turned out not to be enough! Because due to the differing shapes of the connectors the 120mm feed tube that came with #1 turned out to be insufficient, I need a longerfeed tube to prevent birdsnesting. Except that no store here sells them. I found some online in the UK (180mm) after a day or two of looking around for a reasonable deal, but it's been a pain to get them sent out, I ordered them a few days ago... hopefully they'll get sent out today and arrive in a couple of weeks. After shipping and customs getting their cut? About $50 USD for two little brass tubes a couple millimeters in diameter.
In the case of #3, I was seriously considering designing them in blender and then 3d printing them (I even took measurements for the design). If the ones from the UK don't fit, I may still have to.
And here I am not doing any project that's focused around lots of smaller parts. I can only imagine if I was, say, fixing up an older car or something like that - the ability to scan in a rusty or cracked part, maybe apply a couple filters to clean up the scan (remove any rust growth/pitting, seal cracks, etc) and have a local 3d print service print it out would be incredible. I've tried using photogrammetry for these sorts of purposes but it's just so limited - it has trouble working in the best of conditions, if you face any sort of shine or transparency - which, for example, all three of my above examples had - it's unlikely to work.
Or rather...
It is easy to quantify. And the quantification is "4/5ths of commercial airplane accidents are caused by humans doing the wrong thing".
Not just the life support, either. Every piece of control hardware in the cockpit is a potential failure source. On ground you can have any number of backups sitting around, ready to switch to a different one at the flip of a switch.
Probably in this manner.
Yep, XKCD has a comic-form response for everything - even itself. ;)
A thesaurus!
What are you talking about? Every self-respecting nerd should know that they're still here.
People need to stop picturing all dinosaurs as looking like some kind of leathery reptiles. I mean, we not only know now that velociraptor was feathered, but even how many secondary wing feathers it had (14). Jurassic park would have maybe not been as scary had their "raptors" looked like this. ;)
Meanwhile, some of their descendants today look like this and attack like this.
Your analogy was to a ridiculously complex multi-year incredibly expensive infiltration operation by the NSA. Don't blame me if your analogy was ridiculous. If you want a different analogy, go ahead and choose one.
Facts are, UAVs are not commercial airplanes, and are not designed to have the safety features of commercial airplanes. Just ignoring that most of the accidents were early on when they were rushed to the battlefield, you're comparing apples to oranges. You might as well argue that helicopters are deathtraps because kids flying RC toy helicopters crash them a lot.
Air France Flight 296 was crew error. And the body count was "3". The only thing the fly-by-wire system did was prevent their attempt to pull up before they had the speed to actually pull up; the plane was already doomed because they were too close to the trees before they realized their mistake.
Propulsion-only automated flight control systems have already been demonstrated and are expected to be integrated into next-generation commercial aircraft.
Oh wait, you're talking not just about jamming the datalink signal, but also ILS and LPS? Nice jammer you've got. So, how exactly does remote control come into play? If you screw up the signal, you can already crash the plane, pilot or none. There was a pilot on THY726. The plane still crashed. They trust ILS and LPS because the data is usually more accurate than their own senses.
(Plus, it's not even remotely hard to picture additional workarounds that could be added, such as radar, laser, microwave, or ultrasound distance sensors.)
No, system designers are not complete morons.
Seriously, is this your argument? That we can't run a computing device on a battery?
Everybody, throw your cell phones in the trash!
So you're saying that the NSA is going to start taking down planes?
Um, if the US government wants to take down planes, they have whole branches of the government specifically dedicated to that sort of stuff.
The simple facts are, accident rates on planes have gone way down as the amount of automation has gone way up. Computers are simply more reliable than pilots. The fact that most crashes come down to pilot error is a big problem, despite how you present it (ground error being probably the next most common). A computer will always follow the checklist for the right response to a given problem ("some system failure occurred that should have been handled safely by a well-trained pilot and somehow wasn't."). Pilots screw up.
As for military drones, please, they're not at all comparable to commercial aircraft. Most don't even have a flight radar or anti-collision systems. They were designed to be gotten out to the battlefield as soon as possible, in as little development time and unit cost as possible - and they're not exactly flying a regular London to NYC route either.
Yeah, too bad there's not something called "autopilot" that could take over if communications are lost. Sadly, as we all know, the instant communications are lost, the plane will explode.
You have a large high power radio transmission tower (or the ability to jump airgaps into ATC transmission networks), and you also have the ability to break whatever message authentication / validation encryption system that such a remote piloting system would use? Cool! Followup question: are you the NSA, or is there another organization that I need to be worried about? Last question: wouldn't it have been about a hundred times easier for you just to buy a couple MANPADs?