That's why I suggested embedding the RFID chips in the road. Don't have to worry about the vehicle finding them in that case!
Is this the future of technology?
- Thread starter farmkid
- Start date
Not sure if this has been mentioned, but there is a lot of spare capacity on the roads as they exist today. Imagine if cars could run almost nose to tail on them. They run at the speed limit etc..
We just need to use a more Intelligent Transport System.
You need sessors imbedded into the road and/or signage. Sensors in the car, and we let the computers drive us and we become more passangers than drivers.
You type in your destination and you get driven there. Can't find a parking spot, the computer will take you to the nearest one to your destination.
We just need to use a more Intelligent Transport System.
You need sessors imbedded into the road and/or signage. Sensors in the car, and we let the computers drive us and we become more passangers than drivers.
You type in your destination and you get driven there. Can't find a parking spot, the computer will take you to the nearest one to your destination.
QR codes have a built-in error correction for redundancy. If a part of the QR code is invisible, you can still read it, and it doesn't matter which part. The ones used in signs could potentially use more redundancy if that's deemed necessary.
sojourner said:
Not necessarily. The city sign for the Austrian village Fucking often gets stolen, and so did the road signs for Butt Hole Road, and I've got a feeling that this might also happen in places like Gropecunt Lane, Tickle Cock Bridge, Paint Lick, Pussy, Dead Women Crossing and even Å. QR codes can't be read by a human, so the incentive to steal those signs will drop.
On the other hand I expect the opposite might happen with the Llanfairpwllgwyngyllgogerychwyrndrobwllllantysiliogogogoch's sign, because the original does not fit in the car, while the QR code one will.
The signs would still have the names of streets on them as well. No point creating a separate sign for the QR code to clutter up the landscape. The incidence of theft might actually go up with the introduction of the QR version. At least till the novelty wears off.
Meh, there are QR codes plastered all over New York City on posters and flyers and stuff. They really aren't that novel anymore, at least in places where there's poor GPS signal (downtown urban areas.)
At the end of the day, an automated car won't rely on one particular location technology--it could be required to use at least two methods to pinpoint its location (or just GPS as long as it has a reliable GPS lock.) So, some places would have QR signs, some wouldn't, some might use RFID chips, or wi-fi locations, etc. Given how cheap RFID scanners and GPS chips and such are, it's not adding much to the price of a car to include multiple location methods.
At the end of the day, an automated car won't rely on one particular location technology--it could be required to use at least two methods to pinpoint its location (or just GPS as long as it has a reliable GPS lock.) So, some places would have QR signs, some wouldn't, some might use RFID chips, or wi-fi locations, etc. Given how cheap RFID scanners and GPS chips and such are, it's not adding much to the price of a car to include multiple location methods.
I was just being silly with the theft incentives.
The real way automatic location recognition deals with sign theft is what was just mentioned – relying on multiple sources for the information.
Another thing is that the computer can remember informational hints it has seen recently while the QR codes can encode much more information about a location, so the computer could be aware where the stolen signs were and what was written on them without encountering them. Signs are not hard to deduce – if a QR sign is saying that London is in 9716 metres, after going for 9716 metres you'd know that you've entered London even if the sign is missing. A human can neither count the metres in their head nor remember all signs they have seen. They can also miss signs.
With a computer, the number of signs required to signal the same information is decreased, so theft is less of an issue. The only requirement is that the computer can read the signs, e.g. all of them are using QR codes. If the computer-readable version provides more information (e.g. street IDs, street names and distances for the next couple of intersections, digital signature from the municipality, etc.), they are even more theft- and prank-resistant. They would even be rename-resistant.
The RFID option offers the same advantages, although it has two small issues – if the RFID information isn't signed, I could simply drop some RFIDs on the street to confuse the cars behind me; and if there are missing, broken or phony RFID tags they are more difficult to spot.
Last edited:
You already have the infrastructure to rely on – GPS, the easily recognizable road signs, the street themselves, the measured distance travelled and the calculated relative position, the WiFi and mobile phone tower databases, etc. But for reading street names, QR code makes more sense than text recognition, and it's possible that the first autonomous cars won't attempt to read street names at all. I don't know if they are even that useful. What the car really needs are coordinates, possibly street IDs for referencing the map, location IDs and information about the road ahead, including directional information for destinations of interest.
So any machine-readable information sources will be just an additional aid, and they will probably contain information that's different from what's available right now, or something that we can't think of at all.
So any machine-readable information sources will be just an additional aid, and they will probably contain information that's different from what's available right now, or something that we can't think of at all.
Last edited:
^Highlighting parts of my previous post as they relate to your post.
This was a really interesting read. I wish the 22nd century onwards was more fleshed out, but of course that can't be reasonably predicted.
There are a lot of things I suspect may well happen that are described, particularly:
There are a lot of things I suspect may well happen that are described, particularly:
- Automation putting stress on the economy as robots increasingly replace human workers.
- Continued environmental degradation
- Computing power continuing to grow exponentially (though I'm not sure it will reach the "billions of earths of humans" levels described in the timeframe specified). Strong AI is probably only a few decades away.
- Rise of the BRIC countries as the expense of the US
- Medicine advances to the point that by the end of the 21st century people essentially have indefinite lifespans.
- Nanotechnology really takes off and all sorts of crazy stuff becomes possible.
BTW, the Mega Engineering episode that introduced me to PRT is called "Personal Pods". I don't know how often Discovery reruns it, though.
Cooperative navigation is doable, but not so simple as it might seem at first. Consider: Car A informs Car B about its navigation solution (including uncertainty). Car B incorporates this into its own navigation solution, along with the measured distance (and possibly bearing) to Car A.
Now, Car B informs Car A about its navigation solution. The trouble is, Car A cannot treat this as a completely independent measurement, because it is already incorporating what Car A believes. It can be handled, but it's a bit tricky mathematically.
Yeah, I'm not sure cars telling each other where they are is that useful in terms of establishing a fixed location reference. Certainly useful for managing traffic flow and accident avoidance, though!
What about messing up with all cars in the traffic? If they cooperate with each other science fiction and crime film writers will have one more plot device at their disposal. If they don't, writers would have to make it up.
What do you mean? Doing something like spoofing signals so a car is confused as to what cars are nearby, thereby causing an accident? This is where I think the sensors become most important. Physical sensing data should always supersede informative radio signals. If the camera says there's a car directly ahead, that's more reliable than a signal saying the car is actually behind you.
But I'm sure some enterprising individual will figure out how to fuck it up completely, regardless.
I doubt there will ever be true flying cars as we understand it from sci-fi. Unless we find some magic unobtanium that would allow cars to simply hover, it won't work. I don't even know how SUPERMAN does it.
I still say flying cars are completely not an issue and not really that interesting...info tech is way more exciting. Flying cars=industrial/linear view of future. No longer very relevant.
RAMA
Similar threads
