Hong Kong is one of the most vertical cities in the world. Like many cities, when it ran out of space, it simply built up instead of out.
But Hong Kong has almost reached its upper-limit for going skyward. And as a financial capital of the world with an incredibly dense population, space is at a premium. That means that large dedicated computing areas, like data centers, have to compete for space with everything else, even though they are in great demand.
The solution for Hong Kong might be to stop building skyward and to start looking under its feet. Data Center Knowledge recently reported that Hong Kong may dig out rock caves under the city and build new data centers down there. Apparently putting a data center in a deep cave isn't such a bad idea because the naturally cooler below-ground air could help maintain temperatures as long as the cave is properly ventilated.
The biggest problem with the underground concept is likely to be price. It was estimated in the Hong Kong scenario that digging out a tunnel for a data center would cost up to $600,000 per meter. That makes for an expensive project that all but the most profitable data centers would be hard pressed to ever overcome.
But there may be other advantages to building underground. Apparently other underground data center projects are in the works, or even have been completed in other places, using decommissioned military bunkers as their base of operations. Swedish IPS Bahnhof converted a bunker below central Stockholm into a state-of-the-art data center back in 2008.
The main advantage to using a military bunker, besides the fact that it's already been dug out, is that they were built to survive a nuclear war. Governments looking for the ultimate level of security may want to consider it.
Even with the structure already in place, it will still be expensive to store data in a nuclear-proof bunkers. Probably credit card companies will take advantage of it. So you can rest easy knowing that in the event of a nuclear war, both cockroaches and your MasterCard bill will survive.
Posted on Apr 01, 2013 at 8:41 AM0 comments
After a recent blog post about voting technology, and another where I advocated moving elections online, people were understandably a little wary. But the most poignant comment I got was also the simplest. An anonymous poster said, "Technology is not always the answer…"
Stab me in the heart, why don’t you? Seriously, that's a pretty hard statement to take for those of us who live in a highly technological world. We tend to think that technology might be the solution for everything. And if its not, it just means that a little more work is needed to make it happen.
It was with that in mind that I dove into an interesting report from the Deloitte University Press entitled “Beyond the Bars: A new model of virtual incarceration for low-risk offenders.” The study brings up some interesting ideas about how technology can be used to create a virtual prison for low-risk inmates, allowing them to both be controlled and serve their sentences from their homes.
Home detention, or house arrest, has been around a long time for some offenders, and I'm not completely sure that it is much of a punishment. Many years ago when I was a crime reporter I interviewed a convict who was sentenced to spend a year in his house. I asked him what he did all day and he said that he mostly spent his time watching “Looney Tunes,” an activity that would admittedly drive me crazy, but didn't seem to affect him very much. He was sad that he couldn't "smoke weed" because someone might check up on him.
But debates about punishment aside, the study implies that technologically enhanced home detention — that would combine such things as risk analysis and geospatial tracking — could be effective as a rehabilitation tool. The report sees each offender being issued a cell phone with numbers pre-programmed based on their crimes. So someone with a substance abuse problem could push a button and be connected with a counselor if Bugs Bunny reruns made him want to light up. That connectivity also presents opportunities for training and education.
“Beyond the Bars” even says that gamification could be used to make sure that offenders regularly check in with their parole officers. Detainees would be given points or badges for making it on time to their appointments, which would be virtual video conferences taking place over that same government-issued phone. If they earned enough points, they could be awarded a longer curfew outside the home, a reduction in their sentence or some measure of increased freedom. Before you scoff at that idea, consider that many people released on parole don’t make their appointments, and that often leads to increased risk of participation in new crimes. By encouraging inmates to make it to all their appointments and providing them the support they need to overcome their demons, it makes society safer.
The biggest advantage to a program like this would likely be felt in states such as California, which recently had to release over 33,000 inmates due to a court ruling on overcrowding. The state has no money to build new prisons, but has a robust IT infrastructure that could support a program like the virtual prison. It would allow the state to punish and rehabilitate offenders when the only alternative might be to let them go all together. Obviously the most violent criminals need to be kept away from the rest of society, but many non-violent offenders might successfully be able to work within this program.
So I have mixed feelings. The technologist in me thinks that everything in the report makes sense, assuming the right inmates could be chosen for the program, especially if the alternative is no punishment at all, or serving a couple days of a year-long sentence before having to be released.
But the whole "technology is not always the answer" comment is still ringing in my ears.
So, what do you all think? Would you feel safe having a virtual prison program like this in your community? What if the alternative is simply to have criminals walking your streets with no oversight? It's a sticky issue and not one that will be locked down anytime soon.
Posted on Mar 29, 2013 at 9:12 AM4 comments
I almost had to go back on a remark I made a few months ago, that the United States should push forward with online voting. This was right after the presidential election, where we saw huge lines in some states like Virginia, with polls having to stay open an extra four hours to accommodate people who were still waiting to cast their ballots. Newly reelected President Barrack Obama even pointed it out as an issue that needed to be fixed in his victory speech.
This week CNN reported that there has been an actual cyberattack on a voting system in the United States, or I should say that there was one, in August of 2012, that we are just learning about thanks to a court hearing. The attack took place in Miami-Dade County, Fla., of all places, an area that has somehow become synonymous with election-related problems since the days of the butterfly ballot.
In this case, apparently, no voters were confused by the system, and no votes were changed. Instead a system whereby citizens could obtain absentee ballots was attacked. Hackers apparently created a program to send spam ballot requests into the system. There were 2,500 fake requests made before system administrators discovered that the requests were all coming from a small set of computers (tracked though their IP addresses) and that the requests were coming in too quickly to have been sent by actual humans sitting at a keyboard.
The attack in question was apparently stopped, no real damage was done and nobody really knows what the hacker’s motives were.
But it’s worth noting that there was not very much security on the system, according to an article in the Miami Herald. I suppose nobody thought that a system to simply request an absentee ballot would be that much of a target.
Now that this attack has come to light, it will likely give ammunition to those who say that elections should never be put online. It’s not a bad argument, though it assumes that pen and paper elections can’t be manipulated, which we all know has been done in certain instances in the past.
Instead, this attempted breach should be a warning that almost any part of an election system is a target. That’s not so different from online banking, credit card machines and anything else protecting money or, in the case of an election system, power in society.
GCN’s Cybereye Bill Jackson has pointed out that securing voting rights and moving them to an electronic or online format involves a lot more than just securing the voting machine. I don’t think anyone believes that we should casually begin to elect our leaders by a new method without first securing the process from both ends and the middle. At the same time, it’s a worthwhile effort that could increase turnout, and ultimately make elections more representative of the entire population of the country.
Doing all that will be hard, but when has that stopped the United States? One of my favorite leaders, John F. Kennedy, once famously said that, “We choose to go to the moon in this decade and do the other things. Not because they are easy, but because they are hard.” Surely online voting would qualify as one of those other things.
Posted on Mar 21, 2013 at 12:02 PM1 comments
I’m not sure how many hours I wasted in my youth playing Lunar Lander, or how many times I screamed at the computer that it was cheating. It probably wasn’t, but trying to land a spacecraft on a jagged planetary surface using only a tiny fuel reserve isn’t easy.
Thankfully, I can now claim that I wasn’t wasting time. Instead, I was developing a unique skill set that could one day make robots in space more efficient.
Like many U.S. agencies, the European Space Agency is experimenting with gamification, with an eye toward making processes and devices more efficient or to solve complicated problems with crowdsourcing. In this case, ESA is attempting to improve the way flying robots are controlled in space.
To get a real sense of how robotic flight works, the researchers needed both a virtual and a real-world component. The virtual component is used to test out the controls, but the real world robot was needed to show how real, physical flight controls would actually perform.
Thankfully, they found the AR.Drone, which has sold about half a million units around the world. There is even a large community of flyers who make apps for the device and videos of their flights. Equipped with two cameras, the midget drone flies around on four rotors and can be steered by any iOS device.
ESA created an app that lets users print out strips of paper that can be used as targets for the AR.Drone. Whatever a user attaches the strip to becomes the International Space Station in the game. So a chair or the side of a house work just fine as space station stand-ins. Then, using an iOS device, users attempt to dock the drone with the space station. The app feeds the user positional data, and the screen displays the docking port on the ISS.
So players are juggling both the physical robot in the real world and the virtual landing instructions within the game. If players can keep the two in sync, they will successfully dock the craft.
ESA then collects the scores of the players and their positional data. All that information will be used to help refine the software used by actual engineers trying to dock spacecraft with the ISS. The more people who play the game, the better that subset of data will be in helping to improve robotic operations.
ESA may release other versions of its docking-type games, including one where the Rosetta probe attempts to land on the 67P/Churyumov-Gerasimenko comet, an event scheduled to happen early next year.
And unlike the Lunar Lander game, I’m pretty sure this new app won’t cheat me out of my victory!
Posted on Mar 19, 2013 at 7:04 AM0 comments
Given all that is happening in the world in terms of sequestration and spending cuts, it might not be the best time to advocate government getting onboard with a hugely expensive supercomputing project that has a good chance of delivering no usable data. Then again, it could change everything we know about reality. And it’s not like we don’t currently have the fastest computer in the world with the Energy Department’s Titan.
But first you have to get your head around a concept that initially seems utterly ridiculous. I know that’s what I thought, at first. But then the more I contemplated it, the more intrigued I became.
So here goes: There is actually a chance, logically speaking, that we are all living inside a simulated reality. And one scientist thinks he’s found a way that we can tell.
Now I know everyone is automatically thinking about “The Matrix,” which was a great movie and seemed pretty original, but there actually have been a lot of simulated reality theories, going all the way back to Parmenides, Zeno of Elea and Plato. Wikipedia has an expansive page devoted to all the theories that’s a fascinating read.
To boil it down in my very non-philosophical way, the main theory that puts us all into a simulation is based on three principles.
1) Humans are getting very close to being capable of producing life-like simulations. Look at any modern open-world computer game like Skyrim, and it’s difficult to tell that it’s all simulated and not a real place. How much more would be needed to push it over the top?
2) In the not-too-distant future, humans will not only be capable of making these simulations, but will have the hardware to produce quite a few, if not an infinite number of them, on a large scale, even entire universes. Which leads us to:
3) Either we are the original humans leading up to the creation of all those simulated worlds or we are somehow living inside one of them. In relation to all those simulations, it is, statistically speaking, more likely that we are inside one right now.
Enter University of Bonn physicist Silas Beane, who works simulating nuclear reactions in a grid-like universe created with computers. In an interview with the New Scientist, he said that his best simulations can calculate some of the properties of real things such as the simplest nuclei. But the process also generates artifacts that don't appear in the real world and must be removed. So Bonn believes that mapping a high-energy particle like a cosmic ray and comparing its behavior to an actual cosmic ray could provide evidence of a simulation.
There is another theory that says that a grand-scale computer simulation would be designed to shut down if the residents inside it got to the point where they could make an equal simulation inside of it, in order to prevent computer overload. If that is true, then we're better off with our blissful ignorance than being shut down all together. Or maybe it just means that we are, after all, the real deal.
Posted on Mar 04, 2013 at 1:33 PM6 comments
“You are being watched.” That’s the tag line to my current favorite TV show, “Person of Interest.” The show features a crime-fighting duo who use a network of cameras and computers around New York City and a device called “the Machine” to predict and fight crime. The show is based on some actual technology even it works at a somewhat magical level, but I never thought that NYC would flip the switch on something that is just about as impressive.
In real life, the system is known as “the Dashboard” and was designed by cops working with Microsoft. GCN reported on the development of the system when it was just getting started, and now the Associated Press says it’s ready for prime time. Significantly for NYC, the city will get a cut of the action should Microsoft sell the system to other cities.
The full name of the technology is the Domain Awareness System. It uses a lot of the infrastructure already in place in and around New York and ties it together to create a situational-awareness picture in the command center at police headquarters. Eventually, the plan is to give individual officers full access to the Dashboard via their laptops in cars or on mobile devices for cops walking a beat.
Although the plan is to sell the system to other jurisdictions, I doubt that every city could successfully, or at least as successfully, implement the system. New York has over 3,000 cameras tied into it that watch everything from streets to alleyways to federal buildings to parks. Cameras are mounted on light poles, rooftops and the sides of buildings, and they are a ubiquitous part of the city now -- ever-present but mostly ignored.
The Dashboard takes those camera feeds and ties them together with other data the city collects, such as arrest records, 911call information, data from license plate readers and even sensor information coming from portable radiation detectors. It works by pulling all relevant data together in one place as a situation unfolds.
As an example, during a shooting incident last summer at the Empire State Building, the system was tested, and helped to diffuse the police response, according to a CNN article. With many 911 calls coming in about the ongoing incident, a dispatcher might have thought that multiple shooters were involved. But by using the system, police could instantly view all the cameras in a five-block radius, and rewind footage up to five minutes. They quickly identified a single gunman and could see that nobody else was involved.
Other uses of the system, as pointed out by NYC officials, include instant access to 911-call records from a certain location so responding officers could know the history of the caller, automatically spotting packages that have been left in place for a long period of time and collecting the arrest records of suspects involved in crimes.
The system also, for example, could check its radiation sensors when a bomb threat is made or a suspicious package discovered, to make sure there is no nuclear payload like a dirty bomb. And if radiation is detected, police could take appropriate action and begin evacuation procedures, even before a single officer arrives on the scene.
Given the current capabilities of modern technology, I would suspect that the Dashboard could also be programmed to use facial recognition to find any suspect with an outstanding warrant who strays in front of a camera, which is more like the “Person of Interest” show than reality at the moment. City officials don’t say that the Dashboard is being used that way, but the infrastructure is in place and analytics software is available, so it doesn’t seem that far off. I’ve seen this type of system work in the GCN Lab, and a similar system was used during last fall’s Republican National Convention in Tampa, Fla.
Some would say that this is a huge invasion of privacy; that nobody will be able to walk down the street without police monitoring their every move, especially when this gets pushed all the way out to the beat cops. Perhaps. But what do you think those 3,000 NYC cameras are doing right now? We already are being watched, especially those of us who live someplace like New York City or London with a big security infrastructures.
The Dashboard is just a way to efficiently manage all the data that’s being collected. It takes a somewhat disjointed, blurry picture and focuses all the resources needed, on the fly, to deal with almost any situation. Should it continue to have success in New York, expect to start seeing versions of the Dashboard in other places where the infrastructure could support it, or even as an excuse for some cities to invest in more cameras and other sensors that can suddenly be managed much more effectively.
Posted on Feb 26, 2013 at 1:28 PM3 comments