IBM Says New Software Will Help Predict Natural Disasters 72
coondoggie writes "IBM says it has patented a natural disaster warning system, which uses analytic techniques that accurately and precisely conducts post-event analysis of seismic events, such as earthquakes, as well as provide early warnings for tsunamis, which can follow earthquakes. The invention also provides the ability to rapidly measure and analyze the damage zone of an earthquake to help prioritize emergency response needed following an earthquake."
Oh Really? (Score:2, Insightful)
So they are claiming to be able to predict the unpredictable? I want absolute proof of their claims.
time scale and prior information density (Score:2)
If they can predict a volcano eruption let us say 40 minutes in advance, why not?! Like weather it is probably a matter of time scale and prior information density.
____ ____ ____ ____
"dyslectic people seem to be stupid and now it is a proven fact that they realy are. haha. learn to read losers."
From http://www.knowledgerush.com/kr/encyclopedia/Dyslectic/ [knowledgerush.com]
Re:time scale and prior information density (Score:5, Insightful)
The thing is, they can't. As others have pointed out from actually reading TFA, the software is to help prioritize disaster relief efforts, not predict the unpredictable.
Let me know how your fortune telling pans out (even so-called weather prediction is actually just a forecast based on statistical models that generally do not hold up, hence why such "predictions" can only be made with a degree of certainty leading up to a probability of 0 (for a non-event) or 1 (for an event) at the time predicted.
Re: (Score:2)
The thing is, they can't. As others have pointed out from actually reading TFA, the software is to help prioritize disaster relief efforts, not predict the unpredictable.
Let me know how your fortune telling pans out (even so-called weather prediction is actually just a forecast based on statistical models that generally do not hold up, hence why such "predictions" can only be made with a degree of certainty leading up to a probability of 0 (for a non-event) or 1 (for an event) at the time predicted.
Actually they are not predicting the event but are predicting the damage radius from the event. There is a damage zone where buildings look OK but really are dangerous to occupy, being able to predict how big this zone is would be very useful, you don't want to use a building that will fall down in the next after-shock as a shelter. The existing network has some serious latency, the last Earthquake I felt here in Michigan took an hour to get posted to IRIS [iris.edu].
Re: (Score:2)
Now stuff like that IS predictable. Forecasting volcanic eruptions with any certainty? Or the weather? Have fun with that. I'll focus on what's important (damage control) and not worry about the rest.
Re: (Score:2)
Usually the idea is to get the (kitchen sink) patent then wait for somebody to come up with a clever idea and sue their ass off.
Re: (Score:1)
There is no such thing as 'unpredictable'. There is just 'too hard for us to predict with our current level of technology and available resources'
Re: (Score:2)
You can never account for every last variable. You can never 100% say that the "weather is going to do such and such on such and such date" because something might come along and change those conditions. The same with volcanoes and earthquakes.
Predict it? I can! (Score:1)
Re: (Score:2)
No, disasters don't go away because of better building standards. They are only mitigated. What you are talking about is the human toll of natural disasters. Earthquakes, volcanic eruptions, hurricanes, tornadoes, and the like won't go away simply because we have built better buildings. They'll still eventually destroy anything and everything we can construct.
Re:Oh Really? (Score:5, Informative)
No, that's just the all to familiar sensationalist headline.
Re: (Score:3, Insightful)
I should have realized this. /., once site that really had news for nerds on stuff that mattered. Now it's yellow-dog journalism at its best/worst.
Re: (Score:2)
Re: (Score:2)
Statistical analysis software at its best. (:
URL is 404 Not Found
Re: (Score:3, Informative)
Re: (Score:2)
Thanks, I think I'll some fun with this. (:
Re: (Score:1, Interesting)
Earthquakes, tsunamis everything is perfectly predictable if you have enough data and processing power. Look at how weather is being predicted all over the world for years.
Re: (Score:2)
Not predicted, statistically modeled with a great degree of accuracy (that is a hair worth splitting).
Re: (Score:2)
Not predicted, statistically modeled with a great degree of accuracy (that is a hair worth splitting).
As someone who is comfortable with statistics and machine learning, I have to wonder what you're driving at here? You started by making a sort of "Guffaw!" statement about predicting the unpredictable. And now you're trying to be pretty precise about separating prediction from accurate modelling. What bone are you picking? This is an honest question: what is the egregious misuse of terminology that you are arguing against?
And before you answer (if you will), I'm asking independently of the actual
Re: (Score:2)
Re: (Score:2)
Not predicted, statistically modeled with a great degree of accuracy (that is a hair worth splitting).
As someone who is comfortable with statistics and machine learning, I have to wonder what you're driving at here? You started by making a sort of "Guffaw!" statement about predicting the unpredictable. And now you're trying to be pretty precise about separating prediction from accurate modelling. What bone are you picking? This is an honest question: what is the egregious misuse of terminology that you are arguing against? And before you answer (if you will), I'm asking independently of the actual summary or TFA. I don't care much about that. Let say this started out with someone saying they had invented a machine that can predict earthquakes (with some important provisos). It's my belief that the rest of your comments still naturally follow. So lets say the claim is that someone can use a bunch of sensor readings integrated with a computer model to model the relevant forces with sufficient fidelity that the model allows actions to be taken (like evacuations or "get in the basement with some bottled water" warnings, or whatever). Is this fundamentally impossible for some reason, or does it just not match a definition of prediction that I'm missing?
I remember someone saying; "There are lies, damned lies and statistics"
Re: (Score:2)
Samuel L. Clemens (with a degree of uncertainty (; )
http://en.wikiquote.org/wiki/Statistics [wikiquote.org]
Re: (Score:2)
Prediction is the 100% accurate forecasting of future events with out any error. Modeling, like is done with the weather, is done with a margin of error (I'm not sure how much so I won't even bother trying to guess). They are fairly accurate but never 100% over time.
So, to summarize what I just said:
Prediction: Accurate 100% of the time over any length of time
Forecast/Modeling: Accurate to within a certain period of time (sometimes its really short, sometimes it isn't), at which point inaccuracies start hap
Re: (Score:2)
I would say that a prediction is simply a statement about a future event. Systems that make predictions (predictors) can have varying degrees of accuracy. Few predictors are 100% accurate, considering the large number of low probability (and usually unmodeled) events. Even a perfect weather predictor may be in error if the power goes out or an asteroid hits the earth.
An
Re: (Score:2)
Most of what is unpredictable is forecastable.
Re: (Score:2)
Only the Sith and mathematicians deal in absolutes.
Re: (Score:3, Informative)
Erm.. they're claiming to be able to predict tsunamis in the aftermath of earthquakes, not the earthquakes themselves.
Although earthquakes, too, should be predictable if you can get the right measuring equipment in the right places: "just" measure the strain, and when it gets close to the slip point, earthquake is likely.
Anyway, the software IBM is talking about doesn't do that. It takes the earthquake as data input, and spits out the likely damage resulting as information output. Something that is both u
Re: (Score:2)
See, I thought we already had software that could do that.
Re: (Score:2)
So they are claiming to be able to predict the unpredictable? I want absolute proof of their claims.
I will believe it when they predict an intelligent decision from the government, how's that for unpredictability?
Re: (Score:2)
No kidding.
Re: (Score:2)
Predicting is easy. It's being right that is hard
As Shakespeare would have it:
GLENDOWER. I can call spirits from the vasty deep.
HOTSPUR. Why, so can I, or so can any man; But will they come when you do call for them?
The Tempest
Re: (Score:2)
No, they're claiming that they can do post-event (that means after the fact, if it's not obvious) analysis of an earthquake or other unpredictable event, and then predict the entirely predictable tsunami caused by same.
Re: (Score:2)
The submission stinks in accuracy. Granted this is /. and is almost a 100% given.
Concentric circles? (Score:4, Funny)
Can they predict series of concentric circles emanating from a red dot?
Once the red dot appears, it's too late for the little rows of human figures.
News (Score:2)
In related news, IBM today filed a patent that uses random number generators to predict real world events. Study claims the accuracy of this new approach is already twice as accurate as weather forecasting.
PowerBall (Score:1)
Re: (Score:3, Interesting)
If you're talking about psychics, the reason more of them don't win the lottery is that they can also take a peek at what happens after they win the lottery. Which they might not like.
no use that power for finding roulette wheel bias (Score:2)
no use that power for finding roulette wheel bias.
Re: (Score:2)
roulette wheel bias? Before or after they push the button under the table?
Actual article (Score:4, Informative)
"Patented", eh? (Score:3, Interesting)
Re: (Score:2)
Why not? Large part of medicine is patented, too...
Bad summary (Score:3, Informative)
Two keywords:
POST EVENT
ie there is no "prediction" of the event - only of its impact on nearby humans. Sensationalism at its best.
Just a thought. (Score:1)
If properly employed -- and connected to a large network of similar computers -- it could be used to map out the aftermath of an earthquake quickly, speeding up the work of emergency responders and potentially saving lives.
Maybe I haven't thought this though, but if there was an earthquake might it not bring the local ISP down? Power, infrastructure damage, etc. would really make this unreliable. It's like when the power goes out in your town and everyone waits for the first person to say something stupid like "The air conditioning isn't working and it's going to be hot, well I guess we should put some box fans in the windows. Hey, can someone che
And what now? (Score:2)
A better patent would be ... (Score:2)
IBM says it has patented a natural disaster. (I just truncated the summary text).
Any countries experiencing natural disasters would have to pay patent fees.
"Yo! You want emergency relief help? You pay your patent fees first!"
Re: (Score:3, Funny)
How hard? (Score:2)
RTFA (Score:3, Informative)
According to Big Blue, the invention would require a piece of software running on each machine in a data center that would gather data generated by vibration sensors, known as MEMS accelerometers, within computer hard disk drives to analyze information generated by seismic events. This technique is enabled by collecting hard drive sensor data and transmitting it via high speed networking to a data processing center, which can analyze the data, classify the events, and enrich the data -- in real time, IBM says.
From there, it can be determined exactly when a seismic event started, how long a seismic event lasted, the intensity of a seismic event, the frequency of motion of a seismic event, direction of motion of a seismic event, IBM says. This invention is able to crowd-source important earthquake IBM stated.
Combine this data with the WTF was that?" posts on Twitter [xkcd.com] and you're all set. (Also, don't many new laptops also have accelerometers? Quake@Home!)
.
Re: (Score:2)
Re: (Score:2)
I'm not sure that the MEMS sensors used in hardrives would tell you anything useful, in seismology 60 Hz is high frequency and 0.01 Hz is low frequency and it seems that this would be well out of the MEMS band of sensitivity, the amplitudes aren't always that high either. Additionally a data center seems like it would have massive amounts of audio frequency vibrations from cooling fans, lots of 60Hz noise from AC power and lower frequency rumbles from the HVAC systems operation. This just seems like a comp
Re: (Score:2)
If you have a whole array of disks and precise understanding of when they fired (a clock with more resolution than the sensor) then you might be able to compare a whole bunch of sensors' data to get useful information.
Waiting for the lottery prediction software (Score:1)
Similar suggestion on Slashdot in 2008 (Score:2)
Didn't I suggest something similar to this on Slashdot back in 2008?
http://hardware.slashdot.org/comments.pl?sid=440258&cid=22283136 [slashdot.org]
There's no real "Launch" (Score:1)
Who needs software? (Score:3, Funny)
Richter scale...? (Score:1)
Just ask Sir Bedevir (Score:1)
"Explain again how sheep's bladders may be employed to prevent earthquakes."