I have a thought. There are already services that attempt to do this, with the exception that they do not automate the reporting: http://www.trapster.com/
Radar detectors are very finicky, and my understanding is that it is difficult to build a reliable one.
Now, my experience is with a specific model (and I think a very very popular one, the Passport 9500), it has an RJ45 connection which plugs into the unit and then the other end plugs into your car cigarette lighter for power. The end which plugs into your car cigarette lighter has a dongle and LED which flashes when radar has been detected (along with a mute button). There is two-way state information transmitted over the line in addition to power.
It would be awesome to build a "man in the middle" device to intercept and interpret the signals from popular device models, then send it's own signal to the iPhone/Android/WP7 phone to share the information.
This has the large benefit of being extremely cheaper to manufacture. Additionally, you're leaving the challenge of building the actual Radar Detection to well established companies who know what they are doing.
Additionally the service could look for patterns in the data. False positives are a huge problem with detectors. When every car is reporting radar 24/7 at this one location, you know it's most likely a false positive. Newer Passport models have this logic built in with GPS.
> Additionally, you're leaving the challenge of building the actual Radar Detection to well established companies who know what they are doing.
This is probably going to become more and more important in future. Both RF and processing electronics are becoming cheaper and more capable - and this will lead to 'smarter' radar technology being affordable in speed-gun type devices. Old-school dump pulse-doppler systems are trivially easy to detect, but there are many radar technologies out there that will present a much bigger challenge. It wouldn't surprise if we soon see PCL (passive coherent location) based speed traps soon. Passive radars like these can use radiators of opportunity (cell towers, TV, FM radio) and don't need to emit any energy of their own. They aren't undetectable, but are much harder to detect.
The point is that radar detection is going to become much harder in future, and while the equipment will probably still get cheaper, the technical sophistication required for detection will increase.
> You cannot passively detect another passive receiver.
In general, there isn't really such a thing as a passive receiver. The typical radar receiver design will downmix the incoming signal to some lower frequency (IF) using its 'local oscillator'. Some of the energy from this oscillator will leak out of the receiver and be detectable. This is only one of several possibilities for detecting receivers.
So no, not undetectable, but much harder than detecting an active transmitter.
Are you joking ? How do you distinguish this "leaked oscillation" coming from radar-detector from all other pieces of electronics your car is stuffed up with ? Say, from your car stereo ?
Thank you for all the great feedback! I am Ari Krupnik, the guy in the KickStarter video.
My idea is to make and sell devices, and make the software free--as in freedom and as in beer. The business model is that stuff that costs money to make (hardware) costs money to buy. Stuff that users collect (data) I distribute for free. I want to make sure that data that individual users collect is available to users--for any use that they want for it. There are several closed systems on the market. These companies get people to collect data for them--and then lock it up. They create an artificial scarcity of data and try to make money from that scarcity.
It attaches to the Valentine 1 RD, which is already known as one of the best out there. I hope this guy doesn't underestimate the engineering that goes into a quality radar detector these days. Belscort, Valentine, and Whistler spend millions of dollars and have decades of experience in the engineering. The ability to detect quick trigger, but also having the necessary DSP to filter out false signals being so important, I don't know if this will be a viable product anytime soon.
Edit: It appears he's using a Cobra radar detector. They are a complete joke. This is a non starter if that's what he's going with.
If you have an interest in radar detectors in general, I suggest you check out http://radardetector.net. Read some of the comments on there about Cobras.
I have the V1 and RadarActive transceiver and iPhone app. The hardware works well, but I kind of wish the software was open source, or they at least provided a linkable library or plugin system of some kind (might be limited by iOS) since I keep thinking of neat things I could do with it. For example it would be nice to integrate it with the "Augmented Driving" app.
A public API for the speed trap database would be useful too. I'm not sure what their business model is, right now they just sell the hardware interface.
What I'd really love to see is existing proven radar detector manufacturers incorporate Bluetooth into their detectors, allowing apps to interface with them. Valentine will never do it, but Escort might.
FYI SignalActive (the company behind RadarActive) is based in the SF bay area (San Mateo I think?)
There are limitations on what I can open-source on the microcontroller side. My NDA with Apple explicitly prohibits disclosing communications protocols connection details. There are no restrictions on open-sourcing the iOS side of the software, including drivers for the device. My intention is to GPL the iOS code and let people build whatever they want with it.
My idea is further to have an open RESTful API to the data that anyone can use. I want to make the data genuinely useful to other people. I want to have a GPL-style license for data as well.
SignalActive are in San Mateo. I've been to their office, they are real friendly people.
Looking through their site - that's $90 for the iphone-detector adapter alone. Plus the cost of the radar detector itself, plus (it seems) a service subscription fee in the future. Perhaps I am missing something, but with the setup like this I don't see them building large enough user base to bootstrap the service. It's a good idea, but the adapter cost needs to go down significantly (to under $20) and it should be usable with existing trap detection networks, e.g. Trapster.
That's my biggest concern as well. I own a Passport Escort 8500, a really high-quality radar detector, and even with that, I've been caught off guard. I've used other cheaper units and they're useless from a practical standpoint. They constantly chatter about radar that doesn't exist, or they detect so late that you're already busted. Absolutely pointless.
Any info as to why the V1 is the best out there? I have a Passport 9500ix and my friend owns a V1. I'm constantly noticing that his goes off for many, many more false positives than my 9500ix does. Not trying to start a war or anything, btw. Just curious.
Your ix has a GPS that automatically locks out falses after driving past them a few times. The V1 doesn't. It is also commonly accepted the V1 is more sensitive, but has more falses. This is why people will buy a 9500ix for in town, and a V1 for the highway.
I'm pretty much fascinated with the idea. Every radar-detector has one or several LED indicators. What you need is just to hook up to these LEDs and you will support virtually any radar-detector available on the market. The cost of componets hardly exceeds $10. For extra $10 you can easily add bluetooth chip to get rid of wires coming to your iPhone and thus expending range of supported mobile devices. All this can be done by any EE student as a homework. Basic software can be written in J2ME to cover almost any cheap-phone. And this is what that guy is most likely doing.
The detector in my video is a Cobra XRS9345. I am doing what mikeknoop and ruslan are suggesting: hacking into the LEDs and reading the device's idea of what it detects. The actual signaling is non-obvious and I intend to publish the code that interprets them once it's stable. One idea that I'm kicking around is building a kit that you can solder to your detector's LEDs.
I wonder how much market there is for a kit that requires the user to open up his detector.
Ari, aside from legal issues, I think it mostly depends on how hard the installation process is going to be. If you manage to design a very small PCB which will use bluetooth, can be fit inside radar and needs only three wires to solder (DGND, VCC and LED's anode), then you can easily get some users, especially in the Valley where ppl love gadgets which they can create or expand themselves :-).
BTW. For those who are afraid of openning up their detectors you can easily desing a special version which will hook up to the LED using photo-resistor. I.e. a kind of device you stick up to detector and put its sensor over detector's LED.
> Wonder if there is a monetization method beyond hardware there...
Doesn't need to be.
If you want to be a Fortune 500 compan, go public and make your VCs rich, you want to control the entire stack form hardware through to data, "monetize" (i.e. hog) everything and hide your algorithms from your competitors and users. This is what movie studios are trying to do.
If you want to have a living and product that is useful to people, you can make enough money from hardware sales. You don't get to ring the bell on NYSE, but I can live without that.
* Trapster is the most popular one with manual reporting.
* RadarActive hooks up to an actual detector (Valentine One, considered by many to be the best) to automatically report radar/laser detections.
* Some of the high end Passport detectors and possibly others have GPS and trap databases, but not in real time AFAIK. You have to hook it up to a computer to update the database, so it's only useful for fixed traps like red light / speed cameras.
I hope the networks don't get too fragmented, because the real power lies in the network effects. You need a critical mass of users/detectors reporting for it to be useful.
There is some moral ambiguity here, SHOULD this sort of thing be done?
Can one in good conscience build and maintain a service that at it's core is about circumventing laws designed to save lives?
I am not condemning this and there are a lot of arguments you can make justifying it. But in all I would not choose to be involved it its production, just as I wouldn't work for a weapons manufacturer but I don't condemn my friends that do. (I can't say I don't get a little jealous when they talk about the drones they work on).
In the city I grew up in (Edmonton, Canada), the police used to tell the radio stations where the radar traps were. The goal of a radar trap is to get people to slow down.
That was in a gentler time; nowadays I recognize that radar traps are often used for revenue generation, so forces might be more reticent about revealing location. But 20 years ago even the police themselves would have had no issue with this, and might even have endorsed it.
I have met this guy several times and can say a few things about him: he is the real deal; he is very good, technically savvy, has been working hard and is developing strong expertise in connecting hardware devices to mobile devices, beginning with iPhone. He has a very good reputation at the Hacker Dojo. I'm betting you are going to hear a lot more about his stuff in the future.
The bigger problem with this strategy is that almost everyone is moving to Laser based detection systems. Detecting at this point simply becomes a notification post-mortem.
Additionally as mentioned in previous comments, a solid radar based detector (X, K, Ka, and Lidar) is very difficult to make a good one. Valentine1 being one of the best available on the market.
For the laser problem, definitely look at laser interceptor USA. These guys showed up to Radar Roy's competition with a winnebago they couldnt get a reading on.
Laser is exactly why a radar/laser detector network is useful, but yes, it must include laser detection capability. RadarActive (http://radaractive.com/) integrates with the V1 to do exactly that.
Might want to add a bit of a delay before announcing that you've just detected radar. Otherwise, police following site site will know who's reporting them...
Have you seen http://www.trapster.com? Same idea, without the automatic reporting though. And if they know who's reporting them, so what? It's completely legal.
I am a gringo that lived in Australia for 6 years. In the state where I lived the police had two kinds of cameras.
Large fixed cameras were well marked a couple of kilometers ahead. You are only at risk for a short time, But you would be amazed how the Aussies drove right smack on the speed limit as soon as they saw the signs. Speeding under the camera netted you big points, double on holidays.
The other kind were unmanned mobile units in white vans. Typically placed in urban areas in plain sight (at the bottom of a hill--bastards!). A fine, but no points however.
Another Aussie state actually had a law that the cops could not hide while operating a radar gun.
I found the Aussies to be extremely aggressive drivers, but only at slow speeds and in parking lots. The only time I saw Mad Max, he was driving with 'P' plates or at Summer Nats.
Bottom line: visible enforcement is strong prevention. Most U.S. departments are really after the revenue.
It more than undermines it, it encourages speeding in areas that don't show radar. Turn on an inverse of the map and it gives you speeding zones rather than radar zones.
The marketing attempt at "... RadarLoc encourages law-abiding behavior." is patently false. It encourages the opposite, speeding outside radar enforced zones.
Make radar detectors emit radar when a trap is confirmed.
Use case: Radar detector alerts you to what it thinks is a trap. If it is a trap, you hit a button and it will emit radar for 5 or 10 seconds to warns approaching motorists? The bonus is all existing radar detectors can pick up the signal.
I think by "emmitting" Tocomment means "sending update to the server", thus notifying everyone who is connected and is in some certain range from the point.
No I domean emitting radar. Now a radar gun obvoisly doesn't need a license and that emits detectible radar.
To make my idea legal how about building a combination radar detector/radar gun. Officially the purspose of the radar gun is to get a second estimate on your speed when you may be in proximaty of a trap.
Radar detectors are very finicky, and my understanding is that it is difficult to build a reliable one.
Now, my experience is with a specific model (and I think a very very popular one, the Passport 9500), it has an RJ45 connection which plugs into the unit and then the other end plugs into your car cigarette lighter for power. The end which plugs into your car cigarette lighter has a dongle and LED which flashes when radar has been detected (along with a mute button). There is two-way state information transmitted over the line in addition to power.
It would be awesome to build a "man in the middle" device to intercept and interpret the signals from popular device models, then send it's own signal to the iPhone/Android/WP7 phone to share the information.
This has the large benefit of being extremely cheaper to manufacture. Additionally, you're leaving the challenge of building the actual Radar Detection to well established companies who know what they are doing.
Additionally the service could look for patterns in the data. False positives are a huge problem with detectors. When every car is reporting radar 24/7 at this one location, you know it's most likely a false positive. Newer Passport models have this logic built in with GPS.