"Smart" Billboards Debut in Sacramento

k0osh.CEOofCLIT writes "Remember the billboards in "Minority Report" that scanned your eyes and changed the advertisement based on your shopping preferences? The Sacramento Bee reports: "Soon, this sign along the Capital City Freeway will be able to change its message based on what radio stations motorists have tuned in.""Yeah, Chris can't spell. He and Rob should form a club. *grin*

Good idea. :\

[gt25500]

And watch the number of accidents increase 10 fold because drivers are too busy looking at these billboards. I'm avoiding Sacramento (I know... spelling is badass).

Re:So how . . .

[spectecjr]

So how exactly do these billboards figure out what radio stations people are listening to? Do radios emit EM signals that can be used to determine what they're tuned to (it's been a long time since I took a physics class, somebody help me out here)?

Yep - as do television sets.

It's called heterodyning, and is used to decode FM (frequency modulated) signals. Basically, you mix the signal coming in with the frequency you want to listen to, and the signal at that frequency gets amplified (due to the interference), and the outcome of that is rectified, amplified, and is ultimately what you listen to.

So the billboard picks up the frequency you're mixing the incoming signal with (because you need a frequency generator to create that frequency, and they will emit it -- there's not much you can do to stop it short of burying it in a completely metal box -- which kind of stops the incoming radio signal).

Simon

Re:How is that possible

[Samir Gupta]

I presume that they work on the superheterodyne principle [attbi.com] which 99% of commercial radios out there use. Basically, when picking up radio waves off the air, a radio will remodulate the radio waves to an intermediate frequency inbetween the carrier frequency, and the final output which is sent to the amplifier. This intermediate frequency is emitted by the oscillators (ie, your radio receiver is also a transmitter, which is why radio "receivers" are banned on commercial airliners like cell phones are) and can be picked up and detected just like any other radio wave.

Based on the frequency of the IF wave, the billboard can presumably tell what the majority of the radios in the near vicinity are tuned into.

Re:Privacy?

[John Hasler]

> there's no passive way to do this at all.

Wrong. All they have to do is monitor the radiation from the local oscillator in your radio. The British government uses this to detect unlicensed radios and TVs. To stop them modify your radio to use a non-standard IF.

At one time...

[djupedal]

...Sac-o-tomato was a hotbed for consumer testing. We used to get all the new softdrink flavors and designer cookies, chips, etc. before many other regions around the U.S. Remember, as go California, so (eventually) goes the rest of the U.S. Nothing to brag about, however.

Re:How does this work?

[Anonymous Coward]

>Is there a physics major in the house?

How about an electrical engineer?

This works because nearly all receivers use the superheterodyne principle. The receiver converts the incoming signals to another frequency where most of the amplification and filtering is performed. To do this conversion there is an oscillator, called the local oscillator, in the receiver and this is what can be detected by the billboards. In AM radios the local oscillator is 455 kHz higher in frequency than the station the radio is tuned to and you can hear it by putting two radios very close to each other, tuning one to a station near the high end of the AM band, then tuning the other radio to 455 kHz below that station's frequency so that you hear a tone.

The same thing can be done with FM radio, TV and most other receivers. The reason it works is that all receivers are built using the same basic design so the difference between the local oscillator and station is known.

Re:Privacy?

[John Miles]

Sure, there's a "way." Most FM radios in the US, presumably including the ones in most cars, do their analog signal processing work at an intermediate frequency (IF) of 10.7 MHz. To convert the station's frequency to the IF, the radio uses a local oscillator tuned to either Fincoming+10.7 MHz or Fincoming-10.7 MHz -- usually the former, since it means the range of the oscillator is smaller as a percentage of its output frequency. So if you're listening to a station at 95.5 MHz, your radio is emitting a very weak local-oscillator signal at 106.2 MHz. A receiver at the billboard's location only has to watch for the LO signals corresponding to the stations that are paying to advertise on it at the moment. Often you can demonstrate this yourself by putting two FM radios next to each other, tuning one to a blank spot on the dial near the high (or low) end of the band and sweeping the other one back and forth across the band until it appears to interfere with the first radio.

This is also how UK residents who operate their TV sets without the proper government license are ferreted out. A van cruises around the neighborhood listening for radiated TV local-oscillator signals from unlicensed households.

Re:How does this work?

[Istealmymusic]

Okay, time for a clue. As I'm sure you know, your radio antenna receives all wavelengths simultaneously. The receiver has to filter out all but your tuned-in frequency. To do this, a so-called resistor-capacitor (the cap being your tuning knob) "RC tank circuit" is utilized to provide an oscillation to beat against the mish-mash of the received environmental waves. Local oscillators of this kind are powered by a solid-state Gunn oscillator [harvard.edu] in a Phase-Locked Loop [york.ac.uk] (PLL).

The output is fed through a low-power Schottkey diode [semiconwell.com] to clamp the waveform and lock onto the desired frequency. I'm sure you can tell what I'm getting at: in order to receive frequency RF, one must generate frequency IF [bldrdoc.gov] via local oscillations (LO), and IF directly corresponds to RF. Stephen Wolfram points out [wolfram.com] the relationship V[IF] = V[RF] + V[LO] for increasing and V[IF] = V[RF] - V[LO] for decreasing. Armed with this formula and decent knowledge of the radio's tank circuit, it is trivial to pick up the LO and IF frequencies your car radio transmits, albiet inadvertedly, and customize the billboard contents accordingly. Quite simple really.

Hmm...

[oGMo]

So the billboard picks up the frequency you're mixing the incoming signal with (because you need a frequency generator to create that frequency, and they will emit it -- there's not much you can do to stop it short of burying it in a completely metal box -- which kind of stops the incoming radio signal).

OK, I know very little on the subject, so I want to know if it would work to shield the radio, but not the antenna. Would the internal frequency it still leak "back up" the antenna? Could you extend this in some way so that it wouldn't? (Second, unshielded receiver box, sending a "shielded" signal to the receiver/decoder/whatever.) I mean (given you're paranoid enough) you could probably make a box to encode the whole signal digitally and send it encrypted to a shielded box for digital processing. If you were desperate.

(And for those who say "who cares, why be so silly over such a small thing"... well, it might not matter now, when your radio station of preference is being monitored, but at some point, it will. That's when this knowledge becomes useful.)

Re:How does this work?

[Anonymous Coward]

This explanation is really messed up. RC tank circuits don't exist but LC (inductor and capacitor) tank circuits, or resonant circuits, do. Gunn diodes are a good component for building a microwave local oscillator; phase locked loops are a system for building a stable local oscillator. Finally, schottky diodes don't "clamp" the waveform but they are often a good choice for use in a mixer.

Why don't you go to an explantion of superheterodyne receivers [attbi.com] and learn how they really work?

Re:What wll happen in the following circumstances:

[Myco]

RTFA. This is a huge billboard, not an individual-targeting device. It polls the majority of cars passing by and uses the resulting demographic data to select targeted ads.

Re:What about gridlock?

[LostCluster]

It's a majority rules kind of thing... the strongest LO frequency it finds is converted back to the station it corresponds to, and that is what determines which of the four ads show. If there's no way to make any sense of the signals, then the board just remains in place showing whatever ad the last group of cars that it could make sense of indicated.

Re:Hmm...

[libre lover]

That's precisely how the local oscillator is detected - it leaks back up the antenna and is radiated.

If you're paranoid you can build your own receiver and have it use a non-standard IF frequency. It would really jack up the price of the receiver as you would not be able to take advantage of cheap off-the-shelf components - you'd have to design something akin to the transistor radios of the '60s and '70s which were packed full of individual transisors as opposed to today's designs which use one or two ICs.

The reason this works is because 10.7 MHz is such a common IF, meaning that the internal oscillator runs at either (FM station frequency)+10.7 MHz or (FM station frequency)-10.7 MHz

RTFA!

[rmohr02]

So what happens to this thing when you've got six lanes of traffic inching by, and they're all listening to different things?

The billboard sees all the signals coming from radios and bases its ad on the most common signal.

Distraction = Accidents

[Ratso Baggins]

Here in Oz "roadside TV" is a no-no, you cannot change a roadside billboard automatically, or with any frequency manually. This is a result (apparently) of a study on driver concentration.

The sugestion is that if a billboard changes in a drivers immediate or peripheral vision they will be distracted/alarmed by it.

This much different from seeing a billboard off in the distance and reading it at your leisure as you approach.

Re:Privacy?

[Lumpy]

have you actually tried to detect that signal outside a vehicle? ok now do it to a MOVING TARGET.

and what you are talking about is not the case.. the BBC transmits a subcarrier with a tone on it that is easily detectable. It's the same detection scheme used by american cable TV companies to snif out people stealing cable tv. It's a simple device and putting the subcarrier there makes it air tight in court.. trying to say that "we detected what channel your tv is tuned to doesnt work in court... saying we detected our special signal we transmit to catch them.... does.

Re:Privacy?

[cornjchob]

Touche.

But those signals are very week, especially when going 70 mph. It doesn't make sense to me; seems as though it wouldn't be worth the trouble. These oscillators can't be putting out much power, especially considering the interference produced by the amplifier also in your car's stereo. Output needs to be insanely low, and on top of that, depending on the frequency, the billboard would have to filter out actual radio stations.

But hey, if they want to waste their time, so beit.

Get your unlicensed 100mW tx ready!

[StandardCell]

If you have a transmitter that's not more than 100mW ERP transmitting a nice sine wave around the mixer frequency of FM radio, anyone living nearby can just point it at the billboard and voila! The billboard will likely read only your frequency. Make a directional antenna and make it even better. Best of all, as long as ERP is not higher than 100mW, there's absolutely nothing they can do - no more than you can do anything about them. Fight fire WITH fire!

Confused billboards? NOT!

[Anonymous Coward]

I'm quite surprised at all the comments about people building their own transmitters to fool the billboards, or asking "what happens if there's m different people listening to n different radio stations? won't the billboard go crazy?".

The solution to this is quite obvious with only minimal thought. First of all, any insight into what people are listening to is better than just sticking up random messages. Ultimately, the billboard company won't care if somebody successfully spoofs the radio station detector -- they'll still get their advertising dollars no matter what.

Second, with the thousands of cars that pass by, all they need to do is implement something like a 20 minute moving average. Find the top 3 radio stations listened to on average in the last 20 minutes. Broadcast advertisements suitable to these top 3 stations in the proportions detected. If the ads run an average of 10 seconds each, and 10% listen to station A, 30% listen to station B, and 60% listen to station C, then for every station A ad shown, 3 station B ads are shown and 6 station C ads are shown.

So if station C is a teeny-bopper station, then you'll get 60% zit cream and tampon ads. If station B is an oldies, then you'll get 30% Viagra and Depends ads. And if station A is country, then you'll get 10% gun racks and pickup trucks and piss-poor american beer ads.

If signal spoofing becomes a problem, they can try and weed out the "too powerful" signals and just focus on the second highest. But who cares... they're still charging primo dollars to the advertisers who want to believe that they're getting more effective and targeted advertising.

Now if somebody finds a way to tap into the signal that powers what the billboard displays, then THAT would be an excellent worthwhile hack.

Re:Two way street?

[EmagGeek]

The government goes to great lengths to shield computers they use. The CIA has equipment that can indeed snoop what's being displayed on your monitor, and also pick up your keystrokes from emissions from your keyboard cable. None of it is rocket science. It's all based on the simple principle that a current on the surface of a wire has a very predictable radiated field.

They are able to tell what radio station you're listening to by picking up local oscillator to RF leakage in the mixer stage of your receiver. A radio receiver has a variable local oscillator that is mixed with the incoming RF. That LO is mixed with the RF to produce a signal at both the sum and difference of the LO and RF. The sum is discarded (filtered) and the difference continues down through an IF filter (at 455kHz). Depending on the frequency of the LO, a certain station will end up at 455kHz in the IF stage.

In any mixer, there is leakage from the LO input to both the RF and the IF ports (this is, incidentally, how cops can tell you have a radar detector, they listen for the LO frequency leaking out to the antenna port). So, the billboard has a receiver that can tell what your local oscillator is tuned to and decide what to display based on that.

In a field of 100 cars, the billboard receives the most spectral power on the LO frequency that most of the cars are tuned to (since it all simply adds), so the billboard can also know which radio station most of the cars in the field of view are tuned to, and make a decision based on that.

Re:Privacy?

[Lumpy]

this is going completely by memory of the equipment I know we use and what is installed and active in the headend.

There is a signal injector that creates a subcarrier for an analog TV channel, it is inserted before the modulator so that it becomes part of that channel... I.E. put it into HBO, Showtime, Skinamax, etc... the pay channels that are supposed to be premium channels.... what is most stolen here in the states.

this subcarrier is decoded by the Tv's reciever... it has to because the TV is trying to get the video + audio + SAP (or what is now descriptive audio channel) + the stereo indicator carrier + everything else.

because Televisions dont do anything with this signal the TV ignores it. From what I remember It rides Near the SAP audio subchannel so that it get's decoded completely by the television... I.E. the carrier signal rides with the audio signal all the way up to the final stages before it's converted to audio. this gives it the best path for propagation. It's a low frequency carrier inside the carrier around 90-150Khz so it makes it past all the IF stages. I cant remember the exact operating frequency... and I do believe it is agile in that regard also... It's a nice unit.. completely controlled by PC with the ability to watch/detect the resulting signal after the modulator.

I'll see if the head end techs will let me grab the manual from them.