Yeah, that's totally the right idea. I just scanned through the PDF and it seems more geared towards desktop PC use, and it even talks about the gateway running on a commodity PC. Rather, I think we need to think more about cheap, single use computers. Take something small like a gumsitx board, put two ethernet ports on it and load OpenVPN and a key onto it, then plug in the existing smart device into one port and the existing internet connection into the other port...suddenly the device can't be seen (and hacked) by the world at large, and is on a private network for the utility.
Word on the street is that the current grid is already dangerously insecure and extremely vulnerable to digital attacks. You're right that the "smarter" we make it the more vunerable it is, but we've got to do something to fix it already, so we might as well get some of the huge benefits of making it smarter.
I agree that it should be, however it is completely cost prohibitive to get a separate network run to the smart meter in everyone's home. Even running a separate network to all the utility substations would be challenging.
In reality, VPNs run all over the public internet, and can be extremely secure. DOD even allows parts of their classified networks to run over the commercial internet, provided they have the correct encryption gear at each end. The DOD gear is really expensive and tough to get setup, but there is no reason this needs to be the case existing Public Key and Symmetric Key crypto is plenty strong, and could even be implemented in dedicated (cheap) hardware.
Recently I saw that a bunch of stimulus funds were handed out for bringing the nation's electrical grid into the 21st century. A big part of this is using computers to control various parts of the grid, from utility scale substations down into the home with smart meters and smart appliances.
Anytime you take infrastructure and connct it to computers you are opening it up to a whole new set of threats as well as bringing privacy implications.
Here's a couple greatarticles that go into the details better than I can.
I believe that there is are a couple things that really need to be address for grid security:
- Open protocols and specifications
With all the new technology coming down the pike, all sorts of companies will be sprining up with their gadget or software that will solve some problem. They need to work towards making standards of interoperability so that all these entities could work together.
- Network security
Putting millions of new, network connected, devices out there could lead to a field day for hackers. I believe that they sould quickly develop security technologies that manufactures could then cheaply incorporate into their devices.
A lot of this could be easily (and cheaply) addressed with various communities already out there. For instance, SSL technology has already been built into products like OpenVPN that could easily and cheaply secure huge numbers of smart endpoints.
- Privacy
We need to provide software that is built from the ground up to give uses the privacy that they deserve, while still pushing forward great new technologies.
I had two ideas come to mind on ways to use this technology (laptop screens don't really seem like a killer feature).
- Glasses. It would be great to get a screen on the lenses of my sunglasses, most of the time they'd be clear like normal, but info could pop up throughout the day as needed. Or I could sit back and watch a video, and just turn it off if I needed to see something.
- 3D Displays. If one would take many layers of this together, it would be possible to create a 3D display a couple inches deep that would be able to be viewed without any need for glasses. It would be quite limited to the amount of depth available, but even a small amount could be revolutionary.
To expand a little more on my post....This is an excellent idea. In fact, it could probably be done for not a whole lot. The entry-level SpaceX rocket costs $6million per launch, and NASA could probably get a (very) bulk discount, so lets say $5m. If they launch 50 times a year (take a couple weeks off for christmas or something), that's only $250million...just chump change in NASA's budget.
Then it would be an issue of finding payloads for these 50 launches each year. Ideally NASA could have a competitive process that would let individuals or organizations compete for the free launches. Then if the org isn't able to get their payload ready in time, you could have a pool of ready-to-launch payloads that didn't win the competition, but are willing to launch if there is a chance. If all that fails, the rocket should launch anyway, and just do a sub-orbital flight, so that the system stays sharp.
There's a couple issues with this...one if the biggest is range tracking and availability. Having to be available to launch every week would be very tough for the range, because it usually takes them 24hrs+ to get set for a launch, then there are several days booked where the launch can happen any time in that window. To solve this problem, something that launches with this regularity would need to be independent of the current assets, which with cheap GPS and sat data links shouldn't be too big of a deal.
Another thing to think about would be orbital debris. Launch every week, with the competitively determined payloads, which probably won't be as long lived as traditional payloads, could contribute negatively to the amount of debris in orbit. To mitigate this, a big part of the competitive selection process should be how the payload will de-orbit itself, even in the event of a failure.
Even better would be to figure out some items that *you* need for your organization, then you could pay someone on the team to implement them for you. This way you're getting direct benefit for your money, and the project is getting a new feature that will doubtless be useful to others and the developers will get funding to spend time working on the code.
I didn't see it say if it was a CDMA or GSM phone. Does anyone know what network Google has had its internal phones on before? I heard they gave out some G1s, which would have been GSM...did they give out Droid phones (CDMA) internally as well?
One more entry....built around social network tools;-) I'd be up for joining forces with you other guys, I had just set this up quickly before I saw the other ones.
Yeah, I think to get real buy in from the community you'd need to agree to donate the money. Who is going to help you so that you can win $40k?
Also, if you're working as a team, you need to make sure that one of the team members doesn't run off with the answers and submit them on their own at the last second...thus a need for NDAs or something similar
One way to know (better) that you're getting legit data is to have people upload pictures of them looking up at the balloon and taking a picture of the screen on a GPS. I guess it could be photoshopped, but that sounds harder
I posted this further down because I didn't see this one, but I wanted to start a group too...I was thinking google groups might be better than a wiki though. Thoughts?
The idea is to do things in an open manner, similar to an open source project...however there will need to be some secrecy when the results come in, but the idea is to be open.
Hopefully we can use some of the web tools like google maps to help out the team, and general just have fun on Dec 5. Check it out:-)
I agree that the magnetic alignment system will take a long time to null out any perturbations from launch.
I'm not convinced that the platform wouldn't eventually stabilize though. Especially since there's no moving parts. All it has to deal with is some atmospheric drag (which I believe) is pretty constant, and possibly some solar (going from the light side to the dark side) expansion/contraction. Am I missing something?
Even if it was stable, I don't believe there would be ANY pointing capability.
I made a mistake in my mistake correction.
The Falcon 1 only has a 500-700lbs payload...not the low thousands. So you could load two or three of these new satellites on.
There's a sentence in the article that doesn't make sense and I wanted to clarify it for those reading.
These dimensions place FASTSAT squarely in the microsatellite category where it will compete with such as SpaceX's Falcon 1 and Kistler's K-1, NASA said.
Except that SpaceX and Kistler both make ROCKETS not satellites. The competitors for FASTSAT would be companies like Surrey and Orbital Sciences.
However, (what I think the author probably meant to write) is that SpaceX's Falcon 1 (and Kistler's K-1 if it hadn't just been cut by NASA) would be great rockets to launch a small payload like this. Falcon 1 tops out at a few thousand pounds, so you could cheaply load a few of these into a rocket. For $7million for the rocket and a few million more for each satellite you could send up several serious NASA missions.
You can also get it running with a VMware image, if you're not into qemu. (Mentioned in above post, but not in the link).
See: http://wiki.laptop.org/go/VMWare
They warn that the emulation will be slower than it is on the target device.
This might sound like a huge victory for big corporations, but from a legal point of view, decisions by juries don't lead to important precedents.
Even more, it's totally feasible that they appeal this case, and it gets overturned. Then this white house aide will really be eating his words.
For those wondering what the software actually looks like when it's running, wonder no more. You can download a VMware image that will run the XO operating system, with its installed software. There's directions for installing it on their wiki. (http://wiki.laptop.org/go/VMWare)
P.S. For those commenting that XO looks like a skull and crossbones, look at their site icon on the wiki and you'll get what they were going for.
I'm not yet 25 and I still work with DECnet every day at the office. That's not a good thing mind you....we really need to get off these OpenVMS/Alpha machines!
MSNBC says that everyone on the airplane for its maiden flight was wearing parachutes:
http://www.msnbc.msn.com/id/7643311/
Doesn't sound like the kind of thing you'd want to get out in the press....
I think just to be safe, everyone who ever flies on this airplane will need to have their own parachute....just to be on the safe side;-)
The probe that is going to hit the surface will take readings (pictures) until a moment before impact, and send them (very quickly) to the mothership. A moment later the "impactor" will be vaporized and leave a big hole.
The mothership (as well as telescopes on and around earth) will be watching to see the hole and the stuff that gets shot out from the impact.
Slashdot Mirror
Yeah, that's totally the right idea. I just scanned through the PDF and it seems more geared towards desktop PC use, and it even talks about the gateway running on a commodity PC. Rather, I think we need to think more about cheap, single use computers. Take something small like a gumsitx board, put two ethernet ports on it and load OpenVPN and a key onto it, then plug in the existing smart device into one port and the existing internet connection into the other port...suddenly the device can't be seen (and hacked) by the world at large, and is on a private network for the utility.
Word on the street is that the current grid is already dangerously insecure and extremely vulnerable to digital attacks. You're right that the "smarter" we make it the more vunerable it is, but we've got to do something to fix it already, so we might as well get some of the huge benefits of making it smarter.
I agree that it should be, however it is completely cost prohibitive to get a separate network run to the smart meter in everyone's home. Even running a separate network to all the utility substations would be challenging.
In reality, VPNs run all over the public internet, and can be extremely secure. DOD even allows parts of their classified networks to run over the commercial internet, provided they have the correct encryption gear at each end. The DOD gear is really expensive and tough to get setup, but there is no reason this needs to be the case existing Public Key and Symmetric Key crypto is plenty strong, and could even be implemented in dedicated (cheap) hardware.
Recently I saw that a bunch of stimulus funds were handed out for bringing the nation's electrical grid into the 21st century. A big part of this is using computers to control various parts of the grid, from utility scale substations down into the home with smart meters and smart appliances.
Anytime you take infrastructure and connct it to computers you are opening it up to a whole new set of threats as well as bringing privacy implications.
Here's a couple great articles that go into the details better than I can.
I believe that there is are a couple things that really need to be address for grid security:
- Open protocols and specifications
With all the new technology coming down the pike, all sorts of companies will be sprining up with their gadget or software that will solve some problem. They need to work towards making standards of interoperability so that all these entities could work together.
- Network security
Putting millions of new, network connected, devices out there could lead to a field day for hackers. I believe that they sould quickly develop security technologies that manufactures could then cheaply incorporate into their devices.
A lot of this could be easily (and cheaply) addressed with various communities already out there. For instance, SSL technology has already been built into products like OpenVPN that could easily and cheaply secure huge numbers of smart endpoints.
- Privacy
We need to provide software that is built from the ground up to give uses the privacy that they deserve, while still pushing forward great new technologies.
I had two ideas come to mind on ways to use this technology (laptop screens don't really seem like a killer feature).
- Glasses. It would be great to get a screen on the lenses of my sunglasses, most of the time they'd be clear like normal, but info could pop up throughout the day as needed. Or I could sit back and watch a video, and just turn it off if I needed to see something.
- 3D Displays. If one would take many layers of this together, it would be possible to create a 3D display a couple inches deep that would be able to be viewed without any need for glasses. It would be quite limited to the amount of depth available, but even a small amount could be revolutionary.
To expand a little more on my post....This is an excellent idea. In fact, it could probably be done for not a whole lot. The entry-level SpaceX rocket costs $6million per launch, and NASA could probably get a (very) bulk discount, so lets say $5m. If they launch 50 times a year (take a couple weeks off for christmas or something), that's only $250million...just chump change in NASA's budget.
Then it would be an issue of finding payloads for these 50 launches each year. Ideally NASA could have a competitive process that would let individuals or organizations compete for the free launches. Then if the org isn't able to get their payload ready in time, you could have a pool of ready-to-launch payloads that didn't win the competition, but are willing to launch if there is a chance. If all that fails, the rocket should launch anyway, and just do a sub-orbital flight, so that the system stays sharp.
There's a couple issues with this...one if the biggest is range tracking and availability. Having to be available to launch every week would be very tough for the range, because it usually takes them 24hrs+ to get set for a launch, then there are several days booked where the launch can happen any time in that window. To solve this problem, something that launches with this regularity would need to be independent of the current assets, which with cheap GPS and sat data links shouldn't be too big of a deal.
Another thing to think about would be orbital debris. Launch every week, with the competitively determined payloads, which probably won't be as long lived as traditional payloads, could contribute negatively to the amount of debris in orbit. To mitigate this, a big part of the competitive selection process should be how the payload will de-orbit itself, even in the event of a failure.
Even better would be to figure out some items that *you* need for your organization, then you could pay someone on the team to implement them for you. This way you're getting direct benefit for your money, and the project is getting a new feature that will doubtless be useful to others and the developers will get funding to spend time working on the code.
I didn't see it say if it was a CDMA or GSM phone. Does anyone know what network Google has had its internal phones on before? I heard they gave out some G1s, which would have been GSM...did they give out Droid phones (CDMA) internally as well?
Yeah, you'd need some legally binding NDA like aggreement that people with access to the data need to sign
http://groups.google.com/group/open-balloon-finder
One more entry....built around social network tools ;-)
I'd be up for joining forces with you other guys, I had just set this up quickly before I saw the other ones.
http://groups.google.com/group/open-balloon-finder
Yeah, I think to get real buy in from the community you'd need to agree to donate the money. Who is going to help you so that you can win $40k?
Also, if you're working as a team, you need to make sure that one of the team members doesn't run off with the answers and submit them on their own at the last second...thus a need for NDAs or something similar
One way to know (better) that you're getting legit data is to have people upload pictures of them looking up at the balloon and taking a picture of the screen on a GPS. I guess it could be photoshopped, but that sounds harder
http://groups.google.com/group/open-balloon-finder
I posted this further down because I didn't see this one, but I wanted to start a group too...I was thinking google groups might be better than a wiki though. Thoughts?
http://groups.google.com/group/open-balloon-finder
I'm going to try to start a group that can work together to win the prize (and donate it to charity).
http://groups.google.com/group/open-balloon-finder
The idea is to do things in an open manner, similar to an open source project...however there will need to be some secrecy when the results come in, but the idea is to be open.
Hopefully we can use some of the web tools like google maps to help out the team, and general just have fun on Dec 5. Check it out :-)
Torrents for US-English versions: http://www.teeks99.com/OOoTorrents/ (Hosted on a separate server that isn't getting stomped on...yet)
I agree that the magnetic alignment system will take a long time to null out any perturbations from launch.
I'm not convinced that the platform wouldn't eventually stabilize though. Especially since there's no moving parts. All it has to deal with is some atmospheric drag (which I believe) is pretty constant, and possibly some solar (going from the light side to the dark side) expansion/contraction. Am I missing something?
Even if it was stable, I don't believe there would be ANY pointing capability.
I made a mistake in my mistake correction. The Falcon 1 only has a 500-700lbs payload...not the low thousands. So you could load two or three of these new satellites on.
There's a sentence in the article that doesn't make sense and I wanted to clarify it for those reading.
These dimensions place FASTSAT squarely in the microsatellite category where it will compete with such as SpaceX's Falcon 1 and Kistler's K-1, NASA said.
Except that SpaceX and Kistler both make ROCKETS not satellites. The competitors for FASTSAT would be companies like Surrey and Orbital Sciences.
However, (what I think the author probably meant to write) is that SpaceX's Falcon 1 (and Kistler's K-1 if it hadn't just been cut by NASA) would be great rockets to launch a small payload like this. Falcon 1 tops out at a few thousand pounds, so you could cheaply load a few of these into a rocket. For $7million for the rocket and a few million more for each satellite you could send up several serious NASA missions.
You can also get it running with a VMware image, if you're not into qemu. (Mentioned in above post, but not in the link). See: http://wiki.laptop.org/go/VMWare They warn that the emulation will be slower than it is on the target device.
This might sound like a huge victory for big corporations, but from a legal point of view, decisions by juries don't lead to important precedents. Even more, it's totally feasible that they appeal this case, and it gets overturned. Then this white house aide will really be eating his words.
For those wondering what the software actually looks like when it's running, wonder no more. You can download a VMware image that will run the XO operating system, with its installed software. There's directions for installing it on their wiki. (http://wiki.laptop.org/go/VMWare) P.S. For those commenting that XO looks like a skull and crossbones, look at their site icon on the wiki and you'll get what they were going for.
I'm not yet 25 and I still work with DECnet every day at the office. That's not a good thing mind you....we really need to get off these OpenVMS/Alpha machines!
MSNBC says that everyone on the airplane for its maiden flight was wearing parachutes: http://www.msnbc.msn.com/id/7643311/ Doesn't sound like the kind of thing you'd want to get out in the press.... I think just to be safe, everyone who ever flies on this airplane will need to have their own parachute....just to be on the safe side ;-)
The probe that is going to hit the surface will take readings (pictures) until a moment before impact, and send them (very quickly) to the mothership. A moment later the "impactor" will be vaporized and leave a big hole.
The mothership (as well as telescopes on and around earth) will be watching to see the hole and the stuff that gets shot out from the impact.