Schneier on Security
A blog covering security and security technology.
« Friday Squid Blogging: Squidsoup |
| Wi-fi Blocking Paint »
October 12, 2009
Using Wi-fi to "See" Through Walls
Posted on October 12, 2009 at 6:14 AM
• 31 Comments
To receive these entries once a month by e-mail, sign up for the Crypto-Gram Newsletter.
That is impressive. I wonder who will be the first to abuse it, the government in some form since there is no law against it, or a criminal becuase there is no law against it. Either way, should be interesting where this leads.
Pretty cool, but not really practical.
It's 802.15.4, not 802.11, so wi-fi is not the appropriate term for it. Although you could make the same demo with 802.11, just that nodes would be more expensive, and less battery friendly
That looks like the DEW line! If you have to plant radio receivers all around that's plain old RADAR and Its been known for many years that RADAR could track things; if you are going to the trouble of installing your own very large array of receivers [which you will probably need an armored robot to install since its right up near the house] there's probably even better designs you could use - bring your own transmitters as well and pick optimized frequencies and pulse rates to customize your design based on what you are tracking [like MRI].
You would probably have true 3D VR vision then and could just drive a robot inside and physically grab the perp; or at least the innocent homeowner since you're at the wrong house.
I noticed that it doesn't seem to be picking up the dog that's walking around (unless they removed the dog from the area and I didn't notice). I wonder what exactly it's tolerances are, how big does the moving object have to be and how still does someone have to sit to not be picked up?
Anyways, I think that something like this could actually have uses, despite it's seemingly impractical nature. For example, the radios could be build into walls and supplement security cameras to aid in tracking. Nobody would be the wiser of it either, aside from it pumping out radio signals.
Interesting; I know it's only a general-interest publication, but I'd like to see more info on issues like precision and accuracy; calibration (what do they need to know about the positions of the sensors, or if the sensors move (by accident or design)?); dynamic response: what happens if the people remain still for a while; how well would it separate several individuals? and so on (and on ..). But interesting.
I'd like to see them come up and deploy it with multiple people already in the building and see how it does. I bet it needs to start with a baseline, and their software probably can't handle multiple people. Both of those issues would make it fairly useless for most proposed uses (multiple bad guys are probably already in the house by the time you get there).
I doubt this radar scheme would work through the aluminum sided walls on my house but the roof is asphalt shingles and wood.
They could mount the array on helicopters or better yet use sattelites in space to track me and my neighbors movements in our houses!!!
now im goinng to have to create jammers and sell them for a huge profit.
Hmmm... You could use something like this for those twelve-hour police/shooter standoffs.
Embed the antennae in 'Police - do not cross' yellow tape, encircle the house and then monitor - so you can figure out much earlier that the house is empty...
I did something similar with cell phones when the Joker was terrorizing Gotham City.
One of the potential applications it lists is sensing if someone's still alive, in the case of a scene judged too hazardous for first responders to enter.
It's something of a moot point-- I know from my own EMT training that "there are no heroes" in EMS. If the scene is too hot, you wait outside the hot zone until things cool down. There are a lot of other people in the 911 chain who are trained to deal with that particular flavour of hot-- cops, firefighters, haz-mat teams, confined space rescue, etc.
Rules number one, two, and three all state that the EMT should avoid becoming a patient at all costs, even if it means a patient may die. It's a simple calculus: if you go down, you've just strained the system by requiring another ambulance at the scene, and you've rendered yourself unable to help anyone else. Every ambulance is staffed by at least 2 EMTs, so the problem expands exponentially.
To this end, many air-ambulance services in the US (most notably helicopter services) will not reveal the nature of the call, apart from the location, to the flight crew prior to acceptance of the mission. The idea is that the crew should evaluate flying conditions (expected for all phases of the mission-- from base to scene, scene to hospital, and hospital back to base) with no outside influences. Knowing that there's a baby clinging onto life by a frozen cobweb might make the crew want to take a call in marginal flying conditions, and that's where you hear about flight crews crashing, with loss of life of all on board. At the least, you've lost the crew, at worst, the aircraft crashed into people on the ground, causing further casualties.
Even with the distribution of knowledge, sometimes less is more.
Notice that you can just look inside the windows that are all along that wall and get the same information. Let's see them do it in a house with the typical number of windows, and I will be impressed.
The dog wasn't carrying a cell phone, or maybe you have to crawl around to avoid being detected.
This seems like an ideal experiment designed for success. Who is going to bombard a house with 34 sensors at 6 feet away? To be workable you need to use two or three and at much longer distances.
This capability is out there on the military side already.
New Device Will Sense Through Concrete Walls (April 2006)
"Troops conducting urban operations soon may have a superman-like ability to “see” through concrete walls up to 12 inches thick to determine if someone is inside a building."
"According to Edward Baranoski, Defense Advanced Research Projects Agency's (DARPA's) Special Projects Office, the new radar scope will give warfighters searching a building the ability to tell within seconds if someone is in the next room. Users will be able to detect movements as slight as breathing by simply holding the portable handheld device up to a wall."
"One of the potential applications it lists is sensing if someone's still alive, in the case of a scene judged too hazardous for first responders to enter. It's something of a moot point"
1) Think triage on a building-by-building basis after a natural or man-made disaster. Instead of having to make an entry with a full team to determine if people are alive of dead, rescue efforts can be focused much more efficiently, Also, Google "tactical EMS" for another perspective.
2) While I understand this point of view and somewhat agree, I am increasingly disgusted by the "Safety first, victims last" attitude, especially when it is combined with closing the scene so that no help is coming, nor can victims self-rescue. Police in the town of Gretna firing on people trying to leave New Orleans after Hurricane Katrina is one extreme example.
I agree that there is no point to rescuers becoming hurt or killed. However, this is a reconnaissance tool and even if you have to walk away from trapped victims, you can report their existence so that appropriate resources can be (eventually?) dispatched.
@"...it doesn't seem to be picking up the dog that's walking around..."
The red "x" mark does not appear on the screen as a direct consequence of a physical law; it's drawn by software. I'm pretty sure that for the demo they are just looking for the biggest blob in the data, and from the available information we cannot know whether the method allows to see the dog as well as a man, or not.
>or by first responders looking for signs
>of life in a building that is too
>dangerous to enter.
Doesn't matter. One bit.
If the building is too dangerous to enter, it's too dangerous to enter.
There may be a case with technical rescue situations such as a building collapse when you can have a victim wait an extended time for specialized resources, but victims in such a situation tend not to be mobile.
For a situation like a fire, the decision on whether the building is safe to search and you have sufficient resources to carry out such a search must be made independently of speculation of whether someone is inside.
When you let emotions of speculation become involved, you make bad decisions.
If the situation is deemed safe, you search.
If the situation is deemed unsafe, you don't.
When a neighbor tells you someone is home, you don't know how accurate that information is.
Likewise a father telling you all the family members who were home are safely outside he may not know that their 14 y/o daughter had a spat at the sleepover she was at, walked home, and went to bed.
You let the fire, building, and resources dictate your actions; not rumor.
I'm surprised at the lack of creativity by the commenters. Most people assumed that the most applicable use of this type of technology is surveillance; calling it impractical. Surveillance might be the last place this is implemented, try to be more imaginative.
@Matt from CT
I think you are thinking of a far too restricted use case. As Andrew pointed out it could be used in natural disaster incidences to quickly search for trapped people in cases where the rescue teams are undermanned.
I could also certainly see this as aiding in manned searches of buildings as well, in a similar manner to how rescue teams will sometimes use search dogs to attempt to locate victims in buildings. Clearly you don't rely on this technology to make important decisions like whether or not to send in men, but that doesn't mean it's useless.
This will quickly scale to a usable form factor. That is was so quickly built by a student and prof shows how accessible this has become. Oh and my first question after watching the video..
Can he still have kids? ;)
I'm pretty sure any military application would use devices more powerful than breadboarded usb keys, since they can probably be assumed to have a larger budget than a two-man Utahn graduate grant team.
>now im goinng to have to create jammers and sell them for a huge profit.
Tin foil hats for houses?
This is pretty obviously an outgrowth of the old body capacitance sensors used in monitoring secure areas. It would be pretty easy to spoof since it would also detect moving pieces of any conductors or electronic devices. For SAR they would only detect a person (or electronic device) that was moving; if you are pinned down, tough luck.
"Tin foil hats for houses?"
In "Old England" we've had them for centuries on churches.
The trouble is that it was lead not aluminium "baco foil", and lead became worth enough that people started to steal it...
So with EU hazardous chemical requirments lead can nolonger be practicaly used for roofing.
So sadly some old churches are indeed getting a building grade aluminium "tin foil hat" replacment....
"now im goinng to have to create jammers and sell them for a huge profit"
Hmmm what do you mean by a "jammer" all you are likley to do is create a nice RF field to exploit as an "offset doplar" motion detector...
Now if you could build "party poppers" with strips of aluminium foil of the right length then you could sell them as low cost "chaf systems".
But in a building it's going to get a little anoying very very quickly ;)
guess it's NOT a crazy idea to cover your walls and windows with aluminum foil after all ...
no the jamming would be to "noisy" for offest doppler.
I love how on the tv shows the good guys can often see the infrared signatures of every person on every floor of a building via satellite.
metal screens and some blinds would probably be sufficient for the windows.
I have a workaround solution. Put cats in the building. Or dogs if you need something bigger. It should mess up the readings just enough.
Yes, we got the same spiel at first aid training. When coming on an accident scene, the second thing you do is evaluate the scene for safety, because the last thing you want to do is add another person who needs to be rescued by the professionals later. If you see a large number of people lying dazed in a small valley, for example, think about whether or not there might be a poison gas in there before rushing in.
That's the second thing you do, because the first thing you do is make sure that the professionals HAVE been called in and are on their way.
Yawn, what is new?
This stuff has been discussed in the defense sphere for decades and more recently in other "remote" sensing areas.
The delay in application is the availability of cheap, dispersible sensor devices.
Schneier.com is a personal website. Opinions expressed are not necessarily those of BT.