Posted by: movieotaku | August 7, 2008

Dark Knight: Sonar Phones?

OK, I know it’s only a comic book movie, but still: sonar cell phones?

Work out the details.  Let’s assume it is possible to force anyone’s cell phone to become an echo transponder (it maybe possible for all I know).  The ultrasonic beep would have to be generated on the phone—cell phone transmissions have a notoriously low range: less than 10 kiloherz (kHz).  The human ear can usually hear up to 20 kHz, and some people beyond that.

If the cell phone speaker can manage to generate the ultrasonic beep because Batman’s software completely took over the codec, then you still got problems.  The sound goes out, and we’ll be nice and pretend it actually forms a nice sweeping cone, but when it comes back, it’s still coming through a crappy microphone and digitized by a cheap ADC.  I’ll play nice and assume Batman was really smart and came up with a nifty software hack to perform the real-time compression of ultrasound.  How does he know which direction the echo came from?

Polar diagram for a hypothetical cell phone microphone

Polar diagram for a hypothetical cell phone microphone

Sonar in animals depend quite heavily on being able to tell exactly where a sound came from.  For bats and some birds, it’s the pair of ears on their head.  For whales, they use their lower jaw and ears.  These mechanisms allow them to precisely figure out direction as well as distance.  A cell phone has a very limited ability to detect direction, not enough to give the detailed pictures in the movie.  The microphones on cell phones are designed to be very responsive in a narrow band in front of it and almost deaf to anything outside of that band.  So at best you’ve got a spotlight, but all you get back are a bunch of echoes that tell you nothing of the shape or direction. The microphone still can’t tell what angle the echo came in at.

The one part about that plot device that came close to reality was monitoring all that cell phone traffic for someone’s voice: there are rumors that the NSA’s ECHELON program can pick out voices of “parties of interest” from thousands of calls.  Of course, I’m sure that’s just tin-foil hat thinking. :-)

About these ads

Responses

  1. I think that is your tinfoil hat thinking Re: picking out parties of interest. I don’t have to be a scientist or technological wizard to figure out that the cell phone sonar is not possible. But on a practical side, the logistics of the monitoring system and cell phone hack was ridiculous. It was on par with the idea that the Joker would have been able to plant explosives in every room of the hospital undetected.

    Nice post, though.

  2. fascinating. and about time someone pointed it out.
    thanks!

  3. […] obviously there is always the potential for this type of technology, today the reality is that there are a few kinks in Batman’s plan, the main issue being that you would not know the direction that the echo came […]

  4. Theoretically speaking….

    if Batman had the layout of the building (knew where all the walls were, etc) and then he triangulated the positions of each cell phone(see quote from different site below) then he would know at least where everyone was (since he probably had all the records and knew who each phone belonged to).
    I don’t know exactly how accurate the triangulation is, but I guess somewhere within the range of 3-6ft (aprox accuracy of good GPS systems)
    now say (based on the quote below) that by knowing the density of different objects (drywall, steel, wooden furniture, flesh) he could get a rough sense of what was in the room by how long the signals took. and I don’t think that the issue in that case is the cell phone receiving the signals back, rather the towers.

    I assume your question relates to a real-time and automatic calculation of a handset’s location within a network rather than a post hoc calculation that can be done manually for a specific case (eg to help find a missing person).

    Where a handset is within range of four base stations, the time it takes for a signal to travel from the handset to each base station (or from each base station to the handset) is measured. As the speed at which a radio signals travels and the distance between each base station are known, it is possible to use the difference between those measurements to calculate three hyperbolas that share a unique intersection point, which represents the location of the handset. Two hyperbolas would be sufficient to calculate the general whereabouts of a handset, but three are necessary to calculate an accurate position.

    There are a couple of ways to calculate the time differential. Time Difference of Arrival (TDOA), also known as Time of Arrival (TOA), involves the network measuring how long it takes a signal sent from a handset to reach three or more base stations. In contrast, Enhance Observed Time Difference (E-OTD) involves the handset measuring the time differential between simultaneous signals sent to it from several base stations. In both cases, it is essential that all the base stations within the network are synchronised.

    GPS uses essentially the same technique, but uses satellites instead of base stations.

    It’s been a while since I’ve looked at a triangulation so there may be some new or updated techniques. The ethical aspect is usually addressed through the user’s prior consent, for example, as part of their initial subscription to a location-based service.

    http://www.linkedin.com/answers/technology/wireless/TCH_WIR/241579-22205819?browseCategory=PRO_PET&goback=.maj

  5. Jack:

    I didn’t have a problem with finding the location of the cell-phone: I’m well aware of using a variety of techniques to locate a cell phone (in fact, a physicist recently used them to track a number of people in the UK ).

    It was using sonar to create a 3-D picture of the room. :-)

  6. yeah, that’s a little far fetched IMO, but say batman had all the building layouts (city hall building inspection records?) then he could reconstruct a 3D outline

  7. What good would a 3D outline do for reconstructing the contents and people in the room?

  8. hi, im an electronic and communications engineering student! and ive been studying the movie for my report on sonar navigation, i think batman doesnt need to have all the building layouts to reconstruct a 3D outline because sonar can differentiate the density, proximity, etc, thru the strength and the time it took for the sound waves to bounce back. And considering the fact that most, if not all, have cellphones, you can actually reconstruct the environment around the phones. So, imagine you’re surrounded by phones, then each phone can make up a picture of you in different views, which can be put up into a 3D image. The only glitch was if cellphones could actually be effective as a high frequency audio transmitter/receiver with air as medium.

    • You still can’t explain that without multiple microphones, you can’t actually determine the angle at which the echoes return, therefore you will never be able to calculate the distance from the phones to whatever object that reflects the echo back… right?

      • That’s why you look at the differences in the time of arrival of the signal to a bunch of receivers. You can use that to triangulate the position of the source, simmilar to how GPS does it, only in reverse.

  9. Ok. If detecting the location of cellphones is actually possible by analyzing the content of what is being spoken, can’t we use this as an application to detect the location of criminals and catch them before they create further chaos ? Or is it already being used by countries secretly ?

  10. Yeah!. I dont think that is impossible. Something that i think it could be a problem it would be the strength of the phone signal.

    when you’re talking about one or a few regular cellphones that could be a problem, like the scene of the china skycrapper. Wait!… if you remember: Wayne’s company was developing this technology for the army, so they would’nt use a regular cellphone to create this powerfull device.

    I think when you have a lot of regular phones signal stregnth wouldn’t be a problem. It would work as internet, each cellphone would send a little pack of info. so when you gather all the info of the cellphones you can create an image.

  11. I found out today that the iPhone has a dog whistle application meaning that it can produce ultrasonic frequencies. So at least one phone is half way there.

  12. What if you used four nearby cellphone together for better sound localization? For each phone, you use the echo delay to determine a hollow sphere of possible positions. Every point where four sphere surfaces intersect is an object. This doesn’t quite fit with the movie, but I think it could produce a low res 3D image.

    • What if you used four nearby cellphone together for better sound localization?

      Not bad, not bad. But then for each phone, you’ll receive mutliple echoes at different times, so then you’ll have to figure out when echo matches up with the other echos from other phones. If each echo has a unique signature, quite possible, then it’s a tractable problem, but there’s still the problem of directionality: cell phones are still very directional. But good idea!

  13. Naturally, when using “regular” cell-phones, any such design would make sure to not depend on (audible) pressure waves (with their limited range and detection capabilities), but instead use the more powerful and accurate EM-transceiver present; that of the GSM module.

    The result would be an image that would show something comparable to an X-ray, and would allow for much more coverage per transmitter.

    Another big advantage is that you do not need to modify the phones themselves, since they emit GSM frames on regular intervals. Those frames identify the originating device.

    Then, all you need to be able to scan are 3 (for 2d-mapping) or 4 (for full 3d-mapping) impossibly sensitive receivers capable of detecting (and isolating) the echos of individual GSM frames. And of course some other piece of impossibly complex signal analysis software that manages to make the correct assumptions on positioning of objects.

    Any signal theory grad students listening? :) Seems like an extremely fun project.

    Naturally this concept could be expanded to consider all EM-emitting devices. No need for phones, anywhere there’s electronics, that thing would be able to see the surroundings.

    • @Roy,
      When I saw the movie, I thought they were using EM signals from the cell phones, as opposed to sound/ultrasound. Everybody seems to be thinking of acoustic signals. I thought I had missed that piece of info…

      This example got me particularly upset about how Hollywood movies seem to multiply one particular capability by infinity. Here’s my opinion:

      On SONAR imaging:
      Though theoretically possible, it is impossible to obtain a REAL TIME VIDEO with the resolution they show. The main issue, I think, is the signal wavelength. For a frequency of 1 kHz, the wavelength would be about 34 cm. Now you have to take into consideration that the resolution is inversely proportional to the wavelength and that your spot size would be around that order of magnitude. This means that at best you could “see” blobs moving around.
      You could then say: “Hey, let’s go with ultrasound, instead of 1 kHz pitches”. That’s good and dandy, but then the signal would be greatly attenuated by pretty much any obstacle, let’s say somebody puts their cell phone in their pockets or purse, or leaves it lying facing down on a table.
      In other words, if you want a larger range, go with lower frequencies, at the expense of image resolution, and viceversa.
      This is only from an image quality perspective. If we take into consideration real world parameters that affect the signal strength then we have to compensate for inhomogeneous air properties, barriers between the cell phone and your target; if there is more than one cell phone in the vecinity then you’ll have interference from that other device; if the cell phone is moving (say the person is walking or in a car) you may have doppler shifts, directionality of the receivers, etc.
      Assuming that you can perform the calculations on the cell phone itself (this is easier said than done), you still have to transmit all that information to a central processing unit. Again, at VIDEO RATE speeds. Then you also have to put into the equation other more specific factors like:
      1. Gotham city was under attack from multiple fronts. The lines will probably be congested from people trying to talk to their loved ones. This may make the immensely large amount of data you need to transmit a lot slower.
      2. Pray that there is not a power outage that may affect any of the systems involved (remember there were explosions, car crashes, riots all over the place).
      3. Hope the desired cell phone has coverage.
      4. The user doesn’t turn off his/her cell phone or the battery dies.
      5. Nobody notices the sudden spike in data transmission and tries to correct that in the telephone company, etc.

      Ok. Now that you have solved, or kept under control all those issues, you have to work on a display unit. Here are some of the requirements for the display unit:
      1. Has to fit the Dark Knight’s mask.
      2. Water proof (what if it rains?).
      3. Shock proof (you don’t want the display to break and the pieces hitting Batman’s eyes).
      4. Provide binocular vision. If they don’t, then Batman may lose sense of depth.
      5. Be retractable (what if he wants to use his own eyes?).

      Another one I was almost forgetting. The processing central needs a transmitter powerful enough so Batman has coverage wherever he goes. This is the easiest one, but even so, it adds another step to the process.

      And all that was accomplished in the span of a few weeks.

      On EM imaging:
      If you use EM signals, you pretty much have the same challenges, except that the EM antenna is omni-directional, eliminating the directionality problem mentioned before. However, the problem is that you are stuck with a single frequency (I think around 2 GHz from most phones). This is a wavelength of ~15 cm. Pretty crappy resolution if you don’t have a bunch of detectors to make an image. And you’ll also have to make the detectors do something they were not supposed to do, that is detect coherent signals, record them, calculate an image, and transmit it, all without the user noticing it.

      In short, though the idea is plausible under extremely controlled conditions, taking this experiment outside the lab will be impossible.

      • Juan, I agree that it is practically infeasible to use the existing network of cellphones, not even minding the legality issues. For audio this might actually work in a lab set-up, and the bandwidth required from such a cell-phone is at most the detector’s bandwidth, so a lossless 44KHz audio signal would be all you need to transmit (not even close to video bandwidth), and this can be externally processed when location information of that cellphone is available and accurate (this is hardly possible as the most accurate positioning is easily off by a few meters). Yet this will never work for EM since the hardware in those cellphones simply is not designed to do such capture for EM signals.

        However, you could listen in on all the EM traffic that is generated by such devices without controlling those devices. What I suggested was that you carry your own set of detectors (4 so in a tetrahedron configuration so you can do 3D positioning), and use that to identify the position of neighbouring devices. Since most devices emit EMR which can be identified back to that device (carrying some sort of signature, e.g. GSM frames), it might be possible to identify the echo (and the origin of such echo) of such a signal. With that, you have the location of your reflecting surfaces, and essentially an X-Ray of the surroundings.

        But the bandwidth and resolution needed for the detectors, and the complexity of signal isolation seems to be the real problem here. Higher resolution requires the detectors to be further apart, which makes the detector configuration nonportable. Then again, it might be useful to embed such detectors structurally in a building.

  14. […] Esto significa que el concepto de Visión Sonar 3D de Batman:The Dark Knight no estaba del todo descabellado. […]

  15. […] significa que el concepto de Visión Sonar 3D de Batman:The Dark Knight no estaba del todo descabellado. Share and […]

  16. Well to my recollection x rays have been around for years now and uses a frequency strong enough to view into your body to view bone particles and also view other items such as paper products and woods using a microwave technology the same microwave type of technology found in our wireless phones just in much much lesser frequency. So what makes you believe this does not or can not exist in our phone products? it is just a different frequency of radiation used with the same technology. I would never say you can take an x-ray with a phone but who’s to say relay imagery is not or well never possible? I say this propaganda of either information found in the imagination of a script writer for a movie or information found by having a pet peive for foundary in radioligy or simply hears news on technology presently being researched. maybe this technology is so old and un needed it is not important enough to have posted? who knows? theres very smart people out there you just have to join the group and believe it can happen.

  17. Google microwave technology and look to see what you get.

  18. yea but lets not forget morgan freeman is the brains of the outfit.. shawshank redemption people.. i think hes perfectly capable of creating or acquiring such wonderful toys..and i dont know about you but i got electrical tape over my laptop camera..

  19. If he had access to all the cell phones in the city and he was really smart could he use some form of advance triangulation. (Advance because he has access to more than three cellphones)

    So say in a room he has three cellphones, he could determine their positions by having a unique frequency from each phone (assuming each phone is able to transmit and receive said frequencies) and then with time etc… calculate the location of each phone relative to each other, and thus be able to map out that whole room. (Of course the movie had him map out the whole city that’s a little less realistic but hey, I bought it when I watched it Haha :D)

  20. oh I see someone beat me to the punch haha

  21. […] a sonar network unbeknownst to those carrying them.  There is a rather interesting discussion here on the subject.  Another variant is to use cell phones to spot high levels of radiation. […]

  22. all you need is a round microphone and a round speaker, its basic geometry applied to the Doppler effect. the Doppler effect is pinpointed in the latency of the transmission signal. in air there will be a set speed of sound so there for the results will be consistent if the energy source is of a relatively medium power. the problem occurs when gravity is a variable. as long as you are on the same planet the batman phone sonar system should have been achievable since the 80’s, that or I’m paying way too much in taxes.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

Follow

Get every new post delivered to your Inbox.

%d bloggers like this: