Where I drive the HOV lanes are often more crowded than the regular lanes, though this is due in part to letting electric vehicles in there. There is an HOV interchange between two major roads. You can pay to use it even if you are not in a carpool, but at least once a week it is so crowded that they reserve it for real carpools and electric vehicles only. The charge is made with congestion pricing - many if not most days it is $7 for a mile or two, more than the bridges over the Bay.
Yes, but I also just got BOI approval for my company so no more runs out to Lad Phrao ![]()
That’s good. And you know you can do your 90-day reporting online now? There are still some glitches in the system, it wouldn’t work the first time I tried it, but it works now.
I don’t like motorcycle taxis myself. I’d take a regular taxi from the subway stop anyway.
Immigration used to be in Soi Suan Phlu, and that was more convenient for most people. I think they still use that office for Burmese, Cambodians and Laotians. That’s where they go, I think.
I tried to do the 90 day reporting online and it just refused to submit and didn’t do anything and I was using internet explorer correct version and everything.
Yes I don’t like motorbike taxi’s, I got a broken thumb by falling off one, but sometimes they are the only way to get somewhere in a reasonable time. Anyway now I can use the one stop shop so much easier.
Again, this will only work when virtually all the drivers on the road are computers, and that probably won’t happen in a while.
And it’s an overstatement to say most traffic is due to human error in driving. Cars, even driven optimally by a computer, need braking distance to avoid collisions, so there will always be a certain distance between them that limits how many more you can pack into a road. In fact, a computer might be smart enough not to tailgate, making that distance even larger than humans and therefore decreasing the number of cars per mile of road.
Yeah, HOT lanes. We have them too.
Try the 90-day reporting online again next time. Like I said, it eventually did work for me. Others I know too.
I simply won’t take motorcycle taxis, period, regardless of the time. It’s not motorcycles I have a problem with but rather the motorcycle taxi drivers themselves. If I ever need someone bumped off, I’ll look one up, otherwise forget it.
I wonder if there are studies about this. Because I see a lot of congestion caused by human error. When I get into an exit lane, I zoom past lots of cars, and am surprised to find the blockage I just escaped being caused by several cars in parallel leaving 10 or 15 car lengths in front of them, though the traffic is moving quite slowly. That would be gone. One of reasons that traffic near major exits is slowed up is that drivers in the HOV lane slow to try to move over, and do this across four lanes of heavy traffic. Smart cars communicating with each other would do this much better. Then there are people crossing double white lines to merge from an exit only lane at the last minute, and causing cars behind them to slow and cars the get in front of to slow.
An 8086 could drive better than a lot of people I see.
But tailgating causes slowdowns, because any time anybody brakes, people behind them have to brake hard, and that gets passed back.
The big factor that causes congestion is, counterintuitively, traffic speed.
You might reckon that the faster anyone completes a give journey, the less time they are on the road taking up space, and therefore the less contribution to congestion that they make. But, not so.
The reason is that, the road space a vehicle takes up the made up of (a) the space actually covered by the vehicle, plus (b) the gap between that vehicle and the vehicle in front. And of course the faster the traffic travels, the larger that gap is. In fact the gap increases in proportion to the square of the speed increase (and this is true whether we are looking at the theoretical safe gap, or the somewhat smaller gap that on average we tend to leave). The result is that any advantage arising from shorter journey times is more than offset by the disadvantage of greater road space required. The faster the traffic speeds, the more congestion problems you will have. Human driverss tend to drive as fast as they road conditions allow (subject only to speed limits); that tends to cause avoidable congestion.
Since googlecars will have much better information about other cars on the road, including those not in sight, and about road conditions generally, they’ll be able to moderate their speeds to minimise traffic congestion. Once there’s a critical mass of them on the roads, the moderated speeds they adopt will compel driven cars to travel at more or less the same speed (or, as already discussed, you can have roads or lanes available only to googlecars). The result is that, in urban systems dominated by googlecars, you can expect much less congestion. And that’s before you take advantage of all the additional road space freed up by not having to provide on-street parking.
I seriously doubt that traffic congestion is caused by too many cars leaving too much room between them. When a road is congested, it’s more typical to see people driving too close, not too far away.
Wow, I’m old enough to understand that reference.
I suppose the short answer is that you are closer to everything you want to do outside of your small crappy apartment.
I would disagree about the “too dense” part though. I live in Hoboken, NJ which is the 4th densest city in the US. In fact, 10 out of 11 of the densest cities in the country are part of the New York City metro area. What makes these areas very livable (presuming you like city living in the first place) is that you don’t need a car (although I actually do have one). For the most part, anything I need is in walking distance. Rather than have to deal with car seats and whatnot, I can just toss the baby in a stroller and walk down the street to the nearest store or restaurant. I can grab a few (or many) drinks after work and not have to worry about how I’m going to get home. My commute is an 8 minute ferry ride across the Hudson River, then a 15 minute walk to my office. It’s a hell of a lot better than dealing with road traffic. Even in the winter time, I don’t have to deal with an icy car or plowing out a spot.
Besides the ferry, Hoboken also has NJTransit bus service and PATH subway service to Manhattan. There is also a light rail and PATH service connecting Jersey City, Weehawken and a few other towns along the Hudson along with a major commuter rail/PATH/Ferry/Bus transit hub. Plus you have cabs, Uber, Zipcar, bike-shares or just plain walking.
The only reason I have a car is because my condo came with a spot in a nearby garage for $60 a month. Otherwise parking is the main challenge to having a car in NYC.
Where I think the problem is in the “middle density” area of pretty much everywhere else in the country. If you live way out in the suburbs, exerbs or rural areas and drive to some remote office park, that’s fine because you don’t have a million people converging on that same office park. Although collectively, everyone heading everywhere at once can clog up the roads.
It’s when you get into the dense urban sprawl surrounding the cities where there isn’t enough public transportation to make it practical to ditch the car. Specifically, I’m thinking about Boston. Unless you work downtown near South Station and commute in by rail, or live and work within a part of Boston or Cambridge serviced by the T system, public transportation is completely impractical. In fact, in the 5 years I lived in the Boston area, I rarely rode public transportation and only found it marginally more convenient than actually driving in a few instances. And driving in Boston sucks.
I believe New York City has spent a lot of money on the Port Authority Bus Terminal and Penn Station preserving their historical feel as circa 1970s transit hubs for hobos, drunks and petty criminals.
A computer controlled car needs much less braking distance because it should have a much faster reaction time than a human driver. If it takes a human 1 second to see and react to something in front of them at 70pmh, that is 100 feet of braking distance the computer doesn’t need.
Also, if all, or most of the cars on the road are computerized, they can theoretically be networked. That would allow more coordinated movement in anticipation to road conditions and destinations.
Interestingly, I read that one of the problems that Google was having with their automated cars was that they drive “too perfectly”. That is to say, they follow every single rule to the letter. You would think that’s a good thing, but what happens is that they sometimes get stuck at 4-way stops because the more aggressive human drivers never allow for optimal conditions for the Google car to enter the intersection.
Another problem is that because the Goggle cars have a much faster reaction time and can safely operate the vehicle within a much tighter performance envelope than a human driver, they often make extreme maneuvers to avoid collisions. It’s perfectly safe from an engineering standpoint, but extremely unpleasant for the driver to have the car suddenly brake or maneuver wildly to avoid another car.
But my point is that it still needs MOST of the braking distance as with a human driver because most of it isn’t reaction time, but simply the physics involved. A car needs a certain distance to slow down or stop even if it instantly reacts to the need to slow down or stop.
Yeah, those are interesting and they also point to the problem of mixing human and computer drivers.
Your point is incorrect. The standard advice of keeping “at least 3 seconds” behind the car in front of you is mostly due to the driver’s ability to assess that the car in front is breaking and then reacting to it. Remember the car in front doesn’t stop instantly either. Theoretically if the cars have the same handling characteristics, you could program the robotic cars “keep 6” behind the car in front of you". The millisecond the car in front starts to slow down, the car behind would also apply breaks to keep the distant constant. By the time a human driver sees the car in front is slowing, it has already reduced it’s speed considerably.
IOW, in human timescales, all the cars slam their breaks on simultaneously rather than the “rubber band” effect we currently have with car A slow, then B, then C, then D…
Now in practice, you wouldn’t do that since a truck needs more stopping distance than a Ferrari and other variations on handling characteristics, road conditions and whatnot.
Sometimes it does (sort of) - when it is in a collision. Or trying to avoid one.
But I see your point about coordinating.
I have no doubt that in a world where all - or most - cars are automated and coordinating with each other that we’d be able to pack in more cars per lane-hour of available highway. I also have no doubt that, without policy changes, all that extra capacity will be immediately consumed by cars driving around with no passengers (to go pickup their owners, for example) or by demand induced by the extra capacity. And we’ll be back to where we started.
Actually, strike that. It’ll be worse traffic-wise, since there will be less disincentive to driving in heavy traffic. Today, it’s a miserable experience for the driver, which causes some to time-shift their driving out of peak hours or forego car trips entirely. In this hypothetical future, it’ll be a mild inconvenience, as the human occupants f the vehicle while away their time sleeping, working, watching TV, etc. If you want to get anywhere in a reasonable time, your only choice will be fully grade-separate mass transit, or walking/biking.
Traffic congestion is caused by too damn many cars. Cars close to each other are a function of reduced speed because of this and volume.
However, there are pockets of extra slowness - and I’ve noted that these are often caused by slow cars and trucks. Especially in middle lanes. Our traffic columnist has a name for them - road boulders. Cars stuck behind them wind up changing lanes which causes braking. I especially noticed then when driving over a bridge several miles long, with no entrances or exits to confuse things.
I haven’t driven in LA enough to be sure, but when I do I often am moving at 50 in very heavy traffic with little space around any car. I notice few road boulders there. Maybe the limited entrances and exits to HOV lanes help.
Remember, I’m talking about our traffic moving at 10 mph here, not tailgating at 50 in light traffic. That’s dumb.
Human error is definitely the proximate cause of traffic jams in most cases. People fail to maintain speed around curves and up hills, they screw up merging, or they do something dumb like spill coffee or cut someone off. That causes a ripple effect and eventually it gets to bumper to bumper traffic. There’s no doubt computer drivers will be better and will increase the capacity of the roads.
They can only drive that close because the road is congested, and therefore speeds are slow, and therefore they don’t need to leave gaps. But the road is congested at a particular point because so many cars are arriving at that point at the same time. If their arrivals were even a little more spread out, there’d be no congestion at that point and they could drive right through it without slowing.
Typically what happens is that people drive as fast as they feel is safe (or legal) into the zone where congestion occurs (which is where various feeder road are feeding into the highway), thereby maximising congestion in that zone. We know that if you can get people to drive at a more moderate pace on the open stretches of the road we avoid congestion later on, and overall journey times are shorter for the average road user.
The thing is, even leaving aside the understandable instinct to drive as fast as is reasonable in the moment, drivers don’t actually have the information to know what the optimal speed for avoiding congestion is, since this varies according to the time of the day, the day of the week, the weather and lighting conditions, etc, and on what’s happening up ahead, as to which they have little or no information. The other thing is the “prisoner’s dilemma”; if all drivers drive at the optimal speed, all benefit, but if most drivers drive at the optimal speed but one drives as fast as he can, he benefits even more. Thus a driver making a decision about how fast he will drive will make a decision which will benefit him if all other drivers could and did act in the communal interest. But if all drivers act individualistically, congestion results and all lose out.
But google cars can have the information needed to identify the optimal congestion-avoiding speed, and they can be programmed to choice in the communal interest. Result; dramatically less congestion, and shorter journey times.
Okay, but wouldn’t that mean some cars would have to go slower so they didn’t arrive at that point? How is that any different from going slower due to congestion?