Slashdot Mirror
Traffic Studied Using Computer-Linked Cars
mprindle writes "Yahoo News has an AP article about a system that links individual cars to analyze traffic patterns, which allows the drivers to avoid traffic jams and accidents. This system is part of the 'smart highway' initiatives. The data from the car is sent to a central server and from that data traffic patterns in a 40 mile radius. According to the article this technology is less expensive than using poll mounted antennas or ground sensors."
From the summary:
Another fine proofreading job, Zonk.
--
Go ahead...mod me down...you know you want to.
From FTA: Acura's 2005 RL features a navigation system that provides real-time traffic updates for 20 major cities; information is transmitted to the cars via XM radio satellites. Traffic data is aggregated from local police, transportation departments and other sources. The big question: How much are people willing to spend to avoid sitting in traffic? List figures 10 to 15 percent of drivers in a given area would need to participate to make the system effective. The devices bought separately cost about $1,000.
lmao. the poster has been online too long... I believe what he was looking for was a pole mounted sensor... Its funny how what you do everyday becomes evidenced in the misspellings you cause. Go with that where you will.
Gravity Sucks
will this alert me when my wife's car in the vicinity, when my ... um ... "colleague" ... is with me in the backseat ...
If EVERYONE has a computer in their car to help them avoid traffic jams, then it would be absolutely pointless. The traffic would become more widely distributed, sure, but it'd shift away from highways that are designed to hold traffic, and into residential areas that aren't. You're going to have traffic somewhere, so whether it's on the highway or on another road is immaterial. Thus, these computers are pointless for anything more than data-gathering.
They'll tout the lower the cost of the 'system' so they can easier monitor our location, driving habits and speed. When in reality, they are artificially lowering the cost of the system just for those benifits.
*takes off tinfoil cap*
Doubt it'll ever happen in my lifetime (with all the whisle blowers and such out there) but still.
But how long until we can get some level of computer-controlled vehicles? Once the technology has matured a bit, I'd MUCH rather trust a reasonably engineered computerized system than the thousands of other drivers around me on my way about town. Not that I shouldn't be able to turn it off, but I think the concept would really grow once we switched the carpool lane to the auto-drive lane, and manual drivers learn to stay clear of the 80+mph traffic that flows on it.
Ryan Fenton
There are way too many people who get on a road even when traffic alerts suggest they take other routes. People who are already stuck in traffic have no real choice. But it doesn't end there. There is always people flooding out of the exits onto the highway even when it's jammed even when another route would have made things easy for everyone.
Then again, there is the problem of people just not paying attention to these traffic alerts. In which case, this study is totally pointless.
My two cents.
I find that in the cities where i've lived (San Diego, Atlanta), that even when the highways are gridlocked, there really aren't viable alternatives on surface streets. They're either too far off the route or they're also crowded. So even with a system like this, I don't know that the alternate routes would be that much better a solution, you're still spending close to the same amount of time on the road. It's either gridlocked on the highway or you're gridlocked on the city streets. Maybe better mass transit is the answer.
"Plans are for fools! Oglethorpe, the plutonian (Aqua Teen Hunger Force)
You have to remember that there is such a thing as a traffic engineer. Most potental accident locations have alternate routes pre-plotted. The detours may take you on lower capacity roads, but if you have the ability to filter over a wider area you could overcome this problem by directing people to different alternates. This could really help in some cases.
How long until insurance companies synchronize some of those devices they are testing that memorize your driving habits to some kind of wireless network? Although big brother would be watching, insurance monies would go to your pocket via safe driver discounts and analysis of vehicles' behavior right before accidents.
All your Sybase are belong to us.
But what if your car is stopped in the breakdown lane because of a flat tire or something? If you are the only wired car (from the relatively small pool) on that road at that time, will the system simply think that the road is at a standstill and tell everybody else to avoid that particular road?
Stupid Drivers
It's funny how we spend so much time on alleviating traffic concerns, when it would be simpler to just abandon the car. It's to the point where it's often twice as fast and cheap to use public transport. When I'm in a large city, I park my car at a terminal, hop the train, and go. Not only do I not have to worry about traffic and the associated stress, I also buy back all the time I'd waste behind the wheel to catch up on reading and paperwork. And while using public transport can sometimes mean walking a block or two, it's no worse than finding a parking spot. Really, why, in North America, are we so fixated on the automobile for personal transport?
Be relentless!
No mention of the cell tracking method someone demo'ed a couple years ago? It used data from cell towers to monitor anonymized speed data for cell phones for a certain service, as measured by 2d direction finding the various towers could perform on a phone based on signal strength.
The method, while it generated controversy on slashdot for the possible privacy implications, was a viable and cheap method to get this same data without adding specific new hardware.
Some of the previous post write this type of tech off without considering it could be used as one part of the solution. The only way to improve traffic is a system that effectively intergrates several options. One option, let's say mass transit, cannot do it alone.
Imagin if you could tie this system into traffic signals. The combination of routing a certain set of vehicles to alternate routes along with changing the timing of lights on several routes could ease congestion in many cases. Most of the gridlock I see is not caused my a major accident but small incidents. Add an effecient system that deals with moving hazards off the road quickly,something like what they have on the autobahn we probably see huge back ups reduced. There will always be some gridlock but that does not mean a system has failed.
What you just described is optimal utilization of all available routes. That means no traffic jams at all, anywhere (unless there just is no more capacity anywhere at all, in which case you have gridlock).
Yeah, it sounds like a good idea now, but think about the privacy issues. If they can track where I am, where I'm going, and how fast, what's to stop a ticket for showing up in my mailbox everytime I go 1 mph over the speed limit? How long until some creep hacks the system and has access to everything he knows to stalk whoever he wants and do all kinds of no good? I don't want radio transmitters in my clothes, I don't want my cell phone to track me, and I don't want my car disclosing my personal information.
Click here to read it. Same story.
Ant(Dude) @ Quality Foraged Links (AQFL.net) & The Ant Farm (antfarm.ma.cx / antfarm.home.dhs.org).
It's an alternative to a select antenna.
What I want is a website that sneaks GPS units onto police cars so we can find out where all the best donut shops are. At least then when I am sitting in traffic I have a dozen artery-clogging donuts to keep me busy.
I Am My Own Worst Enemy
According to the article this technology is less expensive than using poll mounted antennas or ground sensors
Poll: Mounted antennas or ground sensors. You decide!
[ ] Mounted Antennas
[ ] Ground Sensors
General Motors has been doing all sorts of experiments with cars that are driven by computer. They've shown some experiments on television where about eight or ten cars are driving eighty miles per hour on a road at "tailgating" distance from one another.
The idea is not that computers are better at driving than humans, but is a solution to the problem that the driver of each vehicle sees only those cars that are immediately around him on the road. This means that if the vehicle in front of him is stopped, he must stop, too. Imagine a stoplight at an intersection. The light turns green, but you're behind ten cars, so by the time you start going, the light turns yellow again. Why? This is happening because you can't go until the car immediately in front of you goes, and the driver of that car suffers from the same problem. What if all the drivers communicated, so that when the light turns green, everybody would push the gas at exactly the same time? And more specifically, if everybody pushed the gas exactly the right amount so as to accelerate at exactly the same rate? Many more cars would make it through the intersection before the light turns red. Also, we'd all get where we're going a lot faster. That is currently impossible because there is no "central command", no way to create an overall driving strategy for everyone on a given road. Everybody does what he believes is best, and this causes all sorts of bottlenecks that shouldn't otherwise exist.
A system that would essentially control all the vehicles on a road would do exactly that, and more. Now, I imagine that at first, this will only be available on a select few roads as an "experiment", and only people whose cars have the internal components to steer and control themselves at the instruction of external computers will be able to participate. I think the system would work by providing central control locations on a sort of grid, where each section of road has its own control system, and as cars leave one section of road and enter another, their information would be passed on to the next computer down the grid. Also, each vehicle would have to contain the additional sensors to "close the loop", essentially by providing an internal control inside the vehicle that would allow it to slow down or come to a stop in case there is something in the road that the central computer doesn't know about, or some other condition arises.
This system would have tremendous benefits:
I'd say that's four steps, with turn signals and not cutting people off being separate. However I guess they really could be lumped into one group of "learn how to use lanes" which I think is my biggest pet peeve. I guess this rant is more directed at freeway traffic, so stop signs and lights aren't that big of an issue.
The major points are:
-Find a lane and stick with it. Weaving in and out of lanes to get a car ahead almost never actually moves you ahead in traffic, and is a big part of why there is congestion in the first place. The major exception is in using the left, or passing lane. Use it when the person in front of you is going slower than what you are comfortable with. Get out of it when someone is coming up on your tail.
-Get into the lane you need to be in as soon as is reasonably safe. Don't swerve over four lanes of traffic to barely make your exit. It's annoying, dangerous, stressful and just plain dumb.
-Learn how to merge and switch lanes: if one car goes at a time from each lane/ramp merging, traffic fits together like a zipper and can move smoothly. If people keep nosing in, traffic comes to a halt and accidents ensue. Using a turn signal and actually looking is a definate prerequisite. And if someone wants to merge into the opening in front of you (You do have a big enough gap, right? more on this later) let them. There's a good chance that they'll be switching over to the next lane or exit soon anyways.
Notes to traffic engineers (I bet there's a couple of slashdot):
-Left lanes are PASSING LANES. They are NOT on/off ramps. That's what the right lane is for. If there isn't enough room to fit the ramp on the right, maybe a ramp isn't needed there.
-Merging traffic needs time to actually merge. Two or three car lengths is NOT enough space to effectively merge into.
Notes on tailgating:
-Stop it already. Creeping up on the person in front of you will not get them to go any faster. I repeat, it will not get them to go any faster. I see the person being tailgated slow down more often than speed up or get out of the way. Tailgating also actually gets you through SLOWER than not tailgating. If the person in front of you makes a minor speed adjustment, you need time to compensate. If you are tailgating, a minor slowdown on a curve or from being cut of means you end up stomping on the breaks. That means the person behind you has to step on the breaks harder... eventually someone can't stop in time. You aren't getting yourself where you want to go any faster, you're just tying up traffic, being a hazard, and stressing yourself out.
-If you are the one being tailgated, ask yourself, are you in the passing lane? If yes, get out of the way. I'd rather have that asshole in front of me where I can see him and react (Because I leave enough room that I KNOW I can react) than behind me where I have no control over the situation and what he'll do. Chances are he'll end up passing you on the right and cutting you off anyways.
And wherever you are: give right of way to emergency vehicles. It should be common sense and common decency, but it doesn't seem to be a common act. If the lights and siren are going, that means there is AN EMERGENCY
But, I guess I really didn't say much new from what you said, just kinda expounded on basically the same things.
I'll never make that mistake again, reading the experts' opinions. - Feynman
Hmm... maybe we could eventually implement a system to allow for timeshifting traffic rather than spatial shifting. Reward people for driving in off peak hours (or rather penalize for driving in peak hours) such as higher tolls during rush hour in places that already have tolls. Reward companies that offer more flexible scheduling. Schedule semis/other large trucks to avoid rush hour traffic. Dynamic planning of delivery routes which incorporates traffic flow information. This would, however, take a lot more sociological engineering than a box that says "I recommend that you take a right at the next exit."
I'll never make that mistake again, reading the experts' opinions. - Feynman
We still have massive congestion and the occasional 4 hour gridlock, but its better that it would be otherwise. Of course, if there were alternative routes to take when you had a warning of congestion, it would help!
The only real solution is for everyone to cycle, and so long as there is diesel oil or rape seed (Cannola) oil on the planet, I for one, won't get on a bike.
Sent from my ASR33 using ASCII
I call shenanigans! Everyone knows that slashdotters don't have wives.
What are you doing with your colleague in the back seat? Board games?
If Mr. Edison had thought smarter he wouldn't sweat as much. --Nikola Tesla
Gridlock is a very special condition that comes from assholes not obeying the law.
Let me explain. (Long, I apologise)
Say, you are coming up on an intersection in heavy traffic. The street is bumper to bumper traffic. You get to the intersection, but notice that the lane is backed up all the way up to the crosswalk across from you. You decide "Well, it's likely that the traffic will move in a few seconds" and you enter the intersection, stopping behind the car in front of you, but you are still blocking (at least partially) the intersection. Then that green light turns yellow. Then red. You still haven't moved. To make things more realistic, lets say the driver behind you did the same thing.
The intersection is now blocked, but the driver that you have blocked thinks to himself "Oh HELL NO. I'm getting out into the intersection before the light turns yellow, because I KNOW that this idiot in front of me will move in a second" and he pulls up into the intersection.
What you the driver behind you and the driver that is in the intersection facing you don't know is the fact that up ahead there is the same thing happening, becaome some other driver is blocking the intersection as well going the other way.
It dosen't take too many drivers that are assholes to create a situation that impacts hundreds of cars over many city blocks. Bad gridlock spreads as quickly as it takes for cars to line up.
The only way to unravel gridlock is the same method that is used to avoid it. Don't be an asshole and enter an intersection if you cannot clear it immediatly. You are taking the risk of entering a condition that you cannot get out of.
Bad gridlock can only be untangled from the fringes of the locked up area one light at a time working inward.
Computer navigation suggestions cannot help with gridlock, because those that are in the situation are either the idiots that are responsible, or innocent and stuck. What it may help with is allowing you to possibly avoid a current gridlocked area.
Being a labrat for this study I can describe exactly how the program works. Theres three pieces, your pocket pc, your modem, and your gps receiver. To get it to work you must turn on the products in this order, -Turn on GPS -Put Modem in Pocket Pc -Turn on pocket pc -Connect to the network -Start the nav program If you don't do this rain dance in front of your palm pc it will not work. It is very picky about how it wants to work, and is NOT open to interpretation. Furthermore this program is obviously in beta testing there are a lot of issues that occur, such as randomly losing all the data on the flash card that holds the programs for the study, or not being able to connect to the network. Personally one thing I've learned about this program is how unstable Windows CE really is. Now to the day to day operations of the program It has voice directions that tells you where to go to avoid traffic, although since its going though small speakers its not very loud, and if people listen to music the way I do you'll never hear it. Also since its using mapquest type directions sometimes it gives you directions that make no sense, like to get to my house to the highway it gives me a dozen turns, when I know that i can get to the highway in one turn and 10 minutes faster. I have seen it work however, on my way to school it will direct me around the huge wall of traffic that occurs everyday at one stoplight. Overall I'm not a huge fan of any technology that uses gps to know exactly where you are. I'm partaking since it is a fun study and its free. Once the study's over I'll most likely never use it again and put my unit up on ebay. But for people on a long trip or for people who live in an area that has traffic problems (Albany NY DOES NOT have traffic problems despite what the article said) I could see this being useful.
the solution to that is have someone go around smack ing the brain dead PHB's in the head that think that employees MUST be at work at 8:00 am and leave at 5:00pm.
let workers swing their shift 1 hour or so from the max point. let me come in at 8:30 and leave at 5:30. or let me come in at 7:00 and leave at 4:00
too many managers think that being there exactly at 8:00am is important, in reality it is not and has not been that way for decades. Also giving employees the ability to telecommute 1 day a week will also help. many MANY people can effectively work at home one day a week.
until some sanity can be pounded into management we will continue the rat race that causes problems twice a day.
Do not look at laser with remaining good eye.