|
Arafa posted:\And here is my $10,000 question. When stopped, why do some drivers leave gigantic spaces between themselves and the next car, even in tight left turn lanes and other very congested spaces? This has to be one of the worst and most avoidable forms of congestion. Yet I can never find any consistent reason why people do it. I've asked a lot of people and no one really seems to know why. Because of that, I think it must be some unconscious force, like an excessive desire for control or safety or something. Any thoughts? I do it so I can do a quick U-turn in case of some sort of life-threatening emergency. What. I'm paranoid.  Plus it saved me from getting hit the other night in the Taco Bell drive-through when some idiot put her car in reverse.
|
#
?
Jul 30, 2009 14:49
|
|
|
|
| # ? Apr 20, 2024 06:17 |
|
|
Cichlidae posted:Due to some fun cognitive biases, it's always going to seem like your lane is moving the slowest. In theory, the leftmost lane should always drive fastest (It's the law here). If there are left entrances or exits, lane drops, narrow shoulders, or heavy congestion, then the middle lanes can be faster. This makes sense in theory, but I've never known it to be true - I consistently notice that in some types of traffic jams and in some traffic jams in specific places the right lane is noticeably faster than any of the others, and this is regardless of what lane I'm in. My impression is that it's because people simply do assume the left is fastest, and tend to gravitate towards it because of those cognitive biases, freeing up the right lane.
|
|
#
?
Jul 30, 2009 16:29
|
|
|
Tomfoolery posted:Great thread! I have always been miffed by the misalignment of traffic light timings and I'm glad that I'm not the only one thinking about it. It's a combination of lots of factors. Contractors and engineers alike have a saying that everything will take twice as long as we think, so always estimate double the amount of time needed. Each project has its own unique set of problems, and, even when a contractor is working, they might not be visible from the road, leading to the appearance that nobody's doing work. That aside, nearly 50% of work zones are "vacant" at any given time. Some work just takes a long time to do because of its nature. Pouring concrete? You have to wait up to 28 days for it to cure and take samples before you can expose it to live traffic or build the next stage. Putting up a new pole? Well, you have to call the 5 utilities who were using it, and get them to shift wires over at their convenience. I'm also partially to blame. I set hours when the contractor can and can't work, in order to prevent congestion. The busier the road, the fewer the available hours. Then we run into problems, like the geotech issues in the Big Dig, contractors going out of business, materials out of supply, inferior metal, etc. quote:2. I'm not sure if this is just a Chicago thing, but we have annoying diagonal streets in areas with otherwise only north/south and east/west roads. These diagonal streets create many slow three-way intersections which greatly increase traffic. Chicago had a clean slate to design the roads however they'd like after a fire destroyed downtown in 1871 - why would they introduce these diagonal streets? We do some ridiculous things in the name of "retaining urban character." Maybe they put them back in just because they were there before. Keep in mind, the nature of traffic in 1871 was completely different than it is today. What things work well for horse-drawn buggies in roads without traffic signals wouldn't work at all for cars. quote:3. Could we get some idea of road costs? For example, how much does a mile of road cost? A mile of highway? Are bridges really expensive? How much does cost affect your designs? Let me answer the last question first: costs are our biggest consideration after safety. I can design a perfect road that would solve congestion problems for 50 years, but if it costs 3X as much as one that will handle traffic for 10 years, we'll build the smaller road. As for individual costs, keep in mind this varies depending on where you are. CT has one of the highest costs of living of any state. 1 overhead sign support: $250,000 1 mile of bike path along a railroad line: $1,000,000 1 replacement bridge, 30' wide by 30' long: $2,000,000 1 mile of freeway in an urban area: $500,000,000 1 mile of freeway through the forest: $100,000,000 1 large bridge replacement, 100' wide by 500' long: $50,000,000 1 relocated utility pole: $10,000 1 traffic signal, small: $50,000 1 24" by 24" sign: $120 1 foot of granite curbing: $30 1 policeman directing traffic, per hour: $75 1 ditch digger, per hour, benefits included: $60 1 traffic engineer's salary, per hour: $30 quote:4. If you could design all of a city's transportation (roads, highways, and public transport) from scratch, how would you do it? If it helps, let's pretend our hypothetical city is symmetrical and on flat land, with population density decreasing from the center. That's a pretty vague question, since it really depends on the local conditions and what sorts of networks are available in nearby cities. I'll have to think about it more; it's not something I can answer on my lunch break. Edit: forgot to close a tag! Cichlidae fucked around with this message at 23:09 on Jul 30, 2009 |
|
#
?
Jul 30, 2009 17:19
|
|
|
Undeclared Eggplant posted:This was very interesting. I've seen two kinds of timings used in intersections like this. What you mention in the second scenario is called split phasing, and it's not very efficient. Whenever you have a mixed-use lane, for example, a combination left/through lane, you have a bit of a dilemma. If you have a left-turn phase, then the first through vehicle won't be able to move and will block the whole lane. Likewise for a through phase and a left-turning vehicle. In cases like this, it's most efficient to have everyone on that approach move at once, hence split phasing. Another time it'd be used is if the intersection is either very small, so opposing left turns (1 and 5, for example) would collide if they tried to move at the same time. Split phasing ensures that only one side goes at a time. And, paradoxically, if the intersection is very large, we may put in split phasing for (admittedly nebulous) safety reasons, usually just on the side streets. Hope that answers your question.
|
|
#
?
Jul 30, 2009 17:25
|
|
|
Cichlidae posted:Rt 6 And how about that 6-10 connector? I've yet to hear it NOT mentioned on a traffic report. That's never a good sign... Cichlidae posted:We do some ridiculous things in the name of "retaining urban character." Maybe they put them back in just because they were there before. Keep in mind, the nature of traffic in 1871 was completely different than it is today. What things work well for horse-drawn buggies in roads without traffic signals wouldn't work at all for cars. This immediately makes me think of the cobblestone streets in Newport. Not. Car. Friendly. BadSamaritan fucked around with this message at 17:45 on Jul 30, 2009 |
|
#
?
Jul 30, 2009 17:37
|
|
|
niteice posted:http://maps.google.com/maps?f=q&sou...006968&t=h&z=18 I'd be willing to bet that it's been 5-way for a very long time. All of our policy documents say that intersections are only 4-way and 3-way, and never to intentionally create something more complex. What will eventually happen, as we've already done in hundreds of places around the state, is "tee up" the fifth leg and make a new intersection away from the existing one, transforming the old intersection into a 4-leg. Observe:  It's a very simple improvement to fix a complex problem. quote:Moving down that road a couple of miles: That's a characteristic of any cloverleaf. Sometimes it's done intentionally, such as at 138 and US 1 in Rhode Island, to allow U-turns. Overall, though, the drawbacks of cloverleafs far outweigh the benefits. quote:I see your move and raise you I-91 in New Haven. You're thinking of CT 34, not I-91. As with many mistakes, 34 is more of a political catastrophe than an engineering catastrophe.  New Haven didn't really need an expressway, but the mayor wanted to remove some 'slums' (read: minorities) from the city, and they were conveniently located in a straight line. Under the guise of "urban renewal," he had a bunch of city blocks leveled. Why does route 34 stop where it does? It would have required a bridge over Yale's rowing grounds (on the left side of the picture), and Yale basically blocked the whole thing... well, the second half of the whole thing, since the first mile was already built. Yay politics!
|
|
#
?
Jul 30, 2009 17:44
|
|
|
nm posted:Meet the most dangerious interchange in Sacramento Looking at it from this level, that interchange certainly looks up to spec. It doesn't seem to have a technical issue, other than being too close to other interchanges and thus causing weaving, though it could be I'm just not zoomed in enough. Little problems like bad pavement, shoddy guardrails, or low-quality pavement markings and signing can really doom a whole stretch of highway. And ring roads are great, as they effectively separate through traffic (going around the city) from local traffic (going in and out of the CBD). If you look at the road network 100 years ago, there were no ways to go around a city; that just doesn't fly in the age of automobiles, where cities are the most congested areas and can fit the fewest lanes.
|
|
#
?
Jul 30, 2009 17:49
|
|
|
One thing I've actually enjoyed (the High Five is fun to drive on but hell on wheels during rush hour) about Dallas roads is the extra U-turn lane at every frontage road stoplight. There is a left-only lane that becomes the right lane of the road it turns onto, and a left+U-turn lane where the U-turn is protected and unregulated, so you can just flip a U without worrying about more than "oncoming" traffic from your right when you get headed the other way. It's handy! I actually have another question and it's more like a defensive driving one: I am usually pissed off because people can't merge well (say onto a not-so-busy highway where the speed limit is 65), and do so slowly. I was under the impression that it's the merger's responsibility to essentially be moving At Speed when they attempt to merge. With nobody in front of me, an onramp is just about the only time where I really get to use all of my torque/horsepower  , because I want to be screaming along by the time I have to get in front of someone potentially doing 70+ miles per hour. Is this right or I guess "appropriate"?
Gunshow Poophole fucked around with this message at 18:21 on Jul 30, 2009 |
|
#
?
Jul 30, 2009 18:18
|
|
|
Cichlidae posted:Looking at it from this level, that interchange certainly looks up to spec. It doesn't seem to have a technical issue, other than being too close to other interchanges and thus causing weaving, though it could be I'm just not zoomed in enough. Little problems like bad pavement, shoddy guardrails, or low-quality pavement markings and signing can really doom a whole stretch of highway. I drive this section everyday and these are my observations. This interchange has all of those things. As 50 turns into the Capitol City Freeway and merges with 80 there is a lot of weaving as people try to move to the left to head to San Francisco and others are heading to the right to merge onto I-5. The entrance to I-5 can't handle the volume during rush hour (not to mention the traffic on I-5 itself that backs up the entrance further) so people will either drive really show to find a hole they can get into or will slam on their brakes when they find one. People also tend to wait till they are very close to the entrance forcing hundreds of cars to have to brake and let them in. This area also has very poor lane markings with several lanes shifting or ending entirely. It also doesn't help that you are driving directly into the setting sun reducing visibility. I don't think it was ever designed to handle the volume it currently does. Roseville, Folsom, and Elk Grove were all sparsely populated farmland and not sprawling suburbs all of this was built. I wish there could be some sort of beltway that linked all the suburbs so everyone wasn't forced to into the city center, but I highly doubt this will ever happen due the enormous costs especially now that CA is broke.
|
|
#
?
Jul 30, 2009 18:24
|
|
|
Great thread OP! This prompted me to write an email to my county DOT about some traffic lights that are a huge pain in the rear end near my house. To my surprise, they wrote back and are going to change the timing on them. Hot drat.
|
|
#
?
Jul 30, 2009 21:53
|
|
|
evilweasel posted:This makes sense in theory, but I've never known it to be true - I consistently notice that in some types of traffic jams and in some traffic jams in specific places the right lane is noticeably faster than any of the others, and this is regardless of what lane I'm in. My impression is that it's because people simply do assume the left is fastest, and tend to gravitate towards it because of those cognitive biases, freeing up the right lane. There's a big stigma here against driving in the right lane. It has almost the same power as using the urinal right next to someone else. Additionally, there is the issue of conflicting traffic from entrance and exit ramps, which leads in well to my next lesson! Lane balance. You're designing a freeway, and you want to put in an off-ramp and an on-ramp. The volumes are roughly as shown below. Do you choose option A or option B?  The off-ramp in option A is what we call a lane drop. This means that the rightmost lane is an exit only lane. Well, you probably already know what I'm going to say: you should have picked option B. "But Cichlidae!" you lament, "option B has a merge area! Merging is lame and will cause more accidents!" Yes, this is true! However, it's safer than the last-minute lane changes that will happen once distracted motorists realize they're in an exit only lane. Option A also results in weaving if there's another exit downstream, while option B does not. "But Cichlidae!" you retort, "option B has three lanes in between the ramps, when it only needs two! That's wasted capacity!" Ah, but what happens if the ramp is under construction? Those 3000 cars would all be forced through the 2-lane section in option A, whereas with option B, they'd have the same number of lanes as before. And that introduces the concept of lane balance: the number of "basic" lanes on a stretch of a freeway is equal to the lowest number of through lanes anywhere on it. You can stick in extra lanes between interchanges from time to time, or to provide for a 2-lane exit or entrance, but those are auxiliary lanes and don't count. The number of basic lanes is what determines the freeway's capacity. Makes sense, right? Weakest link and all that. If you wonder why I-84 is so messed up in Hartford, despite having 5 lanes in some places, note that there are only 2 lanes through the I-84/I-91 interchange. Same thing goes for I-91, in fact.
|
|
#
?
Jul 30, 2009 22:03
|
|
|
BadSamaritan posted:And how about that 6-10 connector? I've yet to hear it NOT mentioned on a traffic report. That's never a good sign... The junction of 6 and 10, called the Olneyville Bypass, is the oldest freeway in Rhode Island. Logically, that means it's the least compliant with modern standards. It had something like 8 exits in 1/2 mile back when it was brand new, and only a couple of those have been removed. I know I hate driving through there, as there's no lane consistency and the curves are designed for such low speeds. Luckily, RIDOT plans to overhaul it before too long. You'll be able to get from 10 North to 6 East, too!  Yes, this is the improved version. On the eastern part of the 6-10 connector, it's always jammed because of the numerous on-ramps and weaving areas on I-95 in front of the mall. With the volumes there, I-95 should really have a couple more lanes, but there's nowhere to put them.
|
|
#
?
Jul 30, 2009 22:08
|
|
|
Here's a question: Why the gently caress is Massachussets so retarded with their placement of FastLane (our version of EZPass) tollbooths? Every other state does it in a way that doesn't cause problems, while I want to kill someone every time I hit the tollbooths on the Mass Pike. It seems the problem is all the Fast Lane lanes are on the right side, and only the right side, while every other state puts them on both sides.
|
|
#
?
Jul 30, 2009 22:12
|
|
|
Stew Man Chew posted:One thing I've actually enjoyed (the High Five is fun to drive on but hell on wheels during rush hour) about Dallas roads is the extra U-turn lane at every frontage road stoplight. There is a left-only lane that becomes the right lane of the road it turns onto, and a left+U-turn lane where the U-turn is protected and unregulated, so you can just flip a U without worrying about more than "oncoming" traffic from your right when you get headed the other way. It's handy! We have a few of those here, too. Only problem is when you get an idiot who tries to U-turn from the rightmost left-turn lane and smashes into someone going left from the second turn lane. quote:I actually have another question and it's more like a defensive driving one: I am usually pissed off because people can't merge well (say onto a not-so-busy highway where the speed limit is 65), and do so slowly. I was under the impression that it's the merger's responsibility to essentially be moving At Speed when they attempt to merge. With nobody in front of me, an onramp is just about the only time where I really get to use all of my torque/horsepower It's absolutely appropriate. The Green Book (our other transportation engineering Bible) says that traffic entering a freeway should accelerate to just a few mph short of the mainline speed before merging in. We were actually talking about that today at work; I-84 east of Hartford, which now is 15 lanes wide in some places (8 eastbound, 7 westbound), used to be 2 lanes each direction with stop-controlled on-ramps until the 80s. Cars back then had much slower acceleration than cars today, generally speaking. Can you imagine trying to accelerate from 0 to 65 in 20 feet to fit into a 100-foot gap in traffic? Driving in a K-car, no less. Holy poo poo... so glad we fixed that.
|
|
#
?
Jul 30, 2009 22:14
|
|
|
the7yearplan posted:I drive this section everyday and these are my observations. This interchange has all of those things. As 50 turns into the Capitol City Freeway and merges with 80 there is a lot of weaving as people try to move to the left to head to San Francisco and others are heading to the right to merge onto I-5. The entrance to I-5 can't handle the volume during rush hour (not to mention the traffic on I-5 itself that backs up the entrance further) so people will either drive really show to find a hole they can get into or will slam on their brakes when they find one. People also tend to wait till they are very close to the entrance forcing hundreds of cars to have to brake and let them in. Don't feel too bad, because not even a beltway would be a permanent fix. As long as there's excess capacity, people will move to the area, only stopping once the roads are jam-packed. It's a little fatalistic, but unless something big happens, any freeway we build will end up congested. Now comes the tricky question: do we build freeways to handle growing cities, or do cities grow because we build freeways?
|
|
#
?
Jul 30, 2009 22:18
|
|
|
mintskoal posted:Great thread OP! This prompted me to write an email to my county DOT about some traffic lights that are a huge pain in the rear end near my house. To my surprise, they wrote back and are going to change the timing on them. Hot drat. Believe it or not, I really enjoy those little complaints. It takes 5 minutes to fix, and I get to spend half the day on a field trip out to the site that I'd otherwise have to spend in the office. They say public sector employees are lazy. It's true that some of the stuff we do wouldn't fly in the private sector, but at least we're honest about how we waste our time
|
|
#
?
Jul 30, 2009 22:19
|
|
|
evilweasel posted:Here's a question: Why the gently caress is Massachussets so retarded with their placement of FastLane (our version of EZPass) tollbooths? Every other state does it in a way that doesn't cause problems, while I want to kill someone every time I hit the tollbooths on the Mass Pike. It seems the problem is all the Fast Lane lanes are on the right side, and only the right side, while every other state puts them on both sides. That, unfortunately, is something I can't answer. You're absolutely right, it makes more sense to have them on the left, as that's where faster-moving traffic should be. Could be a legacy issue, or maybe they just want to take advantage of New Englanders' fear of the right lane to extort a bit more cash! Of course, I've seen entire threads in GBS dedicated to Massachusetts' horrible transportation policies, so for all I know it could be an intentional mistake.
|
|
#
?
Jul 30, 2009 22:22
|
|
|
Cichlidae posted:That, unfortunately, is something I can't answer. You're absolutely right, it makes more sense to have them on the left, as that's where faster-moving traffic should be. Could be a legacy issue, or maybe they just want to take advantage of New Englanders' fear of the right lane to extort a bit more cash! Of course, I've seen entire threads in GBS dedicated to Massachusetts' horrible transportation policies, so for all I know it could be an intentional mistake. It's actually not that it's on the right instead of the left, but that it's not on both - in all other states, you can get to an ezpass without weaving, while in Massachussets you have everyone in the left lane weaving to the right to get to the fastlane - it breaks up traffic much more than necessary. But if you don't have an answer, welp, I'd imagine the answer probably is just some idiot stuck them there a decade ago without thinking and they stayed that way.
|
|
#
?
Jul 30, 2009 22:26
|
|
|
evilweasel posted:But if you don't have an answer, welp, I'd imagine the answer probably is just some idiot stuck them there a decade ago without thinking and they stayed that way. You'd be amazed how often that happens. Most of the freeways around here follow roads built over 200 years ago. Some guy builds a road in the middle of nowhere, people build towns around it, then people build more roads to link the towns. Route 32, which goes from eastern Connecticut up into New Hampshire, used to be an old Indian trail, and look at it now!
|
|
#
?
Jul 30, 2009 22:34
|
|
|
A few questions, since you are in my area. What do you think of the Waterbury mixmaster on I-84/CT-8? Think they will ever actually replace it, considering the billion dollar price tag? What was up with that project to make the shoulders on I-84 wider? It seems like they spent millions of dollars and years of time doing something pretty pointless. They even made some of the bridges wider... just to get a full shoulder on the bridge. smackfu fucked around with this message at 23:05 on Jul 30, 2009 |
|
#
?
Jul 30, 2009 22:46
|
|
|
smackfu posted:A few questions, since you are in my area. $3 billion, actually. And yes, the old interchange is rather dangerous, with lots of left exits and substandard curves; we'll find the cash somehow. Here's some funny trivia. Back in the 1950s, when they were preparing to build it, there were dozens of homes that would have to be demolished. The cost was almost prohibitive. However, during the great flood (I think it was in 1955), all of those houses just washed away! Would that all our ROW acquisition could be so easy. quote:What was up with that project to make the shoulders on I-84 wider? It seems like they spent millions of dollars and years of time doing something pretty pointless. They even made some of the bridges wider... just to get a full shoulder on the bridge. Shoulders are a pretty big deal, actually. The presence of a shoulder, whether or not it's being used, lets people drive faster and increases the capacity of the road. It provides a safe place for people to turn off if their car breaks down or they get in an accident (I've seen quite a few cars parked on the shoulder getting rear-ended or sideswiped.) Additionally, during construction, they can be used as extra lanes, and can even be turned into extra lanes in the future if more widening is required. That project was pretty messed-up, though, if you're talking about the one in Southington and Waterbury. The contractor installed the catch basins upside-down, and the inspectors didn't catch it. Needless to say, they got kicked off the project and sued.
|
|
#
?
Jul 30, 2009 23:03
|
|
|
Cichlidae posted:I'd be willing to bet that it's been 5-way for a very long time. All of our policy documents say that intersections are only 4-way and 3-way, and never to intentionally create something more complex. I can't tell what's going on in this picture. I can see that there used to be a connecting road there on the right. Is that all there is to it? Did they just dead-end that road, or what? I can't really tell from the picture.
|
|
#
?
Jul 30, 2009 23:29
|
|
|
Cichlidae posted:That project was pretty messed-up, though, if you're talking about the one in Southington and Waterbury. The contractor installed the catch basins upside-down, and the inspectors didn't catch it. Needless to say, they got kicked off the project and sued.
|
|
#
?
Jul 30, 2009 23:48
|
|
|
Awesome thread! While I don't live in Seattle and have only visited once, I don't understand who thought I-5 was a good idea. There's merging from the left and right, exiting from the left and right, and an extra 4 lanes in in the middle that I can't even understand when I look at it from Google Maps. Is this primarily due to NIMBY and having no space to do anything different?
|
|
#
?
Jul 30, 2009 23:55
|
|
|
A giant metal bird posted:I can't tell what's going on in this picture. I can see that there used to be a connecting road there on the right. Is that all there is to it? Did they just dead-end that road, or what? I can't really tell from the picture. Route 74 used to intersect with 194 and 30 there, making a 5-way intersection. They dead-ended 74 and brought it up to intersect with 194 further north. Wow, guess what? It's time to learn about signal actuation! The material is pretty dry, so let's make it more interesting!  Actuation means that the signal has sensors to tell where the cars are. There are three main kinds of detectors (plus a half dozen more than nobody uses): inductive loops, which are coils of wire that detect metal objects passing above them via an inductive electric field, RADAR detectors in the microwave range, and video detectors which aren't too reliable, but are pretty cheap. Now on a signal without actuation, known as a pre-timed signal, the phases always have the same length, no matter how many cars show up. That's not very efficient, is it? We put in detectors as shown in the lower right of the diagram above. This actuated signal, as the second multi-colored bar shows, has a variable green time. It can be anywhere between the minimum green and the maximum green. Simple, non? We set up a "gap" time, usually 2 or 3 seconds. Once the minimum green has expired, each time a vehicle is detected, the gap timer is reset. If it ever expires (3 seconds go by with no vehicles detected), the green time ends and the signal turns yellow. It is at this point that the controller will look at other approaches to see which one it should serve next. This is called "gapping out." Advanced Actuation Now, if we want to crank up the efficiency, we can add something called "gap reduction." Basically, the farther along in the cycle we are, the shorter the gap is. It might start out at three seconds, but a while later, a car would need to go by every second to keep the green going. Gap reduction brings the next phase in faster under normal conditions, but can leave the green up for a long time if there's a lot of congestion. Another technique is the "added initial." Basically, as cars line up at the stop bar before they get the green, the controller counts them all. The more cars there are, the longer the minimum green is. This is very useful when your detectors are placed far back from the stop bar, and a long queue of cars isn't able to trip the detectors to reset the gap timer. The detectors themselves have two wonderful features: extend and delay. Extend will make the vehicle's detection last longer, which is good for dilemma zone issues (I'll discuss those another time). Delay waits a certain time before placing the call. That's used on right turns where a vehicle will probably turn by itself without needing a dedicated phase. Right turn delays are about 8 seconds, so if you want your light to turn green because you hate right turn on red, just be patient. And finally, each phase has one of three properties: lock, non-lock, or recall. Lock means that if a detector is hit, that phase will be called, even if the car leaves the detection zone. Non-lock, logically, means that the call is dropped if the car leaves the detection zone. Recall is generally used on non-actuated phases, and means that that phase will be called once a cycle, no matter what. It's also the phase that the signal will "rest" on if there are no cars. I hope that wasn't too complicated! Let me know if there's anything you don't understand.
|
|
#
?
Jul 31, 2009 00:02
|
|
|
smackfu posted:And into Southbury. I didn't want to mention the thing with the contractor for fear I'd jinx it since it's finally done. The state just settled with the contractor a couple weeks ago, and recovered the entire cost of the repairs plus a little more for our troubles. I can't really consider it a happy ending, though, since the public was inconvenienced so much. Reasonable Form posted:Awesome thread! Wow, lots of left exits... I'd say that it was just designed back before we realized that left exits and entrances were taboo. Some of them just seem gratuitous, too; That would never fly these days. Seattle has major congestion problems, there's no secret about that. Remember that, even with the best designs, a road that's at three times its design capacity just isn't going to flow smoothly. The extra lanes in the middle seem to be reversible flow lanes, going one way in the morning peak and another in the evening. Can a local goon confirm? As an aside, they have raised lane markings in Seattle, colloquially called turtles. I have no idea how they don't get ripped up by snow plows! We absolutely can't use any here because they'd last exactly one storm before ending up smashed-up on the shoulder. Does it just not snow there, or do they have some special kind of plow? Cichlidae fucked around with this message at 00:13 on Jul 31, 2009 |
|
#
?
Jul 31, 2009 00:11
|
|
|
Cichlidae posted:Wow, lots of left exits... I'd say that it was just designed back before we realized that left exits and entrances were taboo. Some of them just seem gratuitous, too; That would never fly these days. Seattle has major congestion problems, there's no secret about that. Remember that, even with the best designs, a road that's at three times its design capacity just isn't going to flow smoothly. The extra lanes in the middle seem to be reversible flow lanes, going one way in the morning peak and another in the evening. Can a local goon confirm? Yep the lanes in the middle are reversible and they are called "Express Lanes" here. We also don't get much snow so there are pretty much no plows here, thus the turtles are safe.
|
|
#
?
Jul 31, 2009 00:36
|
|
|
Cichlidae posted:As an aside, they have raised lane markings in Seattle, colloquially called turtles. I have no idea how they don't get ripped up by snow plows! We absolutely can't use any here because they'd last exactly one storm before ending up smashed-up on the shoulder. Does it just not snow there, or do they have some special kind of plow? They use rubber edge plows and don't really bother plowing them well. http://seattletimes.nwsource.com/html/localnews/2008551284_snowcleanup23m.html quote:But it turns out "plowed streets" in Seattle actually means "snow-packed," as in there's snow and ice left on major arterials by design.
|
|
#
?
Jul 31, 2009 01:47
|
|
|
weziman posted:They use rubber edge plows and don't really bother plowing them well. The plows here are nasty. They knock over mailboxes so often that people have taken to sticking a 4' by 8' piece of plywood in front of the box in winter to improve visibility. They also peel up pavement markings, carve up islands, and even rip out train tracks. That was a costly mistake. Speaking of costly mistake, remember how I told you that one overhead sign support costs $250,000?  That dump truck driver forgot that his bed was up and took off down the shoulder on I-84. He hit the sign so hard that his truck tipped up in the air, and they needed a cherrypicker to get him down. That's one strong sign, eh? We'd just put it up a few months before. As to the support, well, the foundation is slightly cracked. $250,000 down the drain. Edit: Here's a link to the article for you, including a photo gallery. http://www.wfsb.com/traffic/18851503/detail.html Cichlidae fucked around with this message at 03:07 on Jul 31, 2009 |
|
#
?
Jul 31, 2009 02:29
|
|
|
Ha. That's awesome. I don't suppose you have any higher res of that? As an outsider (  ) looking in, anytime I need to get to Seattle, I think loooong and hard about just how much I want to go, specifically because of I5. gently caress that road.
|
|
#
?
Jul 31, 2009 03:05
|
|
|
This is an awesome thread, thanks for making it! Questions!: My newly adopted home city of Auckland, New Zealand, with its respectable-for-a-relatively-small-city freeway network has taken to installing ramp signals on pretty much every on-ramp. I'm pretty sure they only get used during peak hours, but they're "dumb" i.e they can't tell how congested that section of freeway is before releasing a pair of vehicles to join the freeway, and some of them are on inclining on-ramps that in turn lead to inclining sections of freeway, meaning you have to accelerate harder for longer to reach freeway speeds. Are ramp signals a good idea? Do they help to ease congestion? Also, I was in Bangkok, Thailand a little while ago and came across what I thought to be an awesome feature of their stoplights which I've never seen before or since. Mounted next to the actual stoplight is a countdown timer which will count down the seconds of each phase of the light. I thought this was great as it means you'll always know how long you have to wait at the intersection before the light goes green. How come I've never seen these anywhere else? Is it actually a stupid idea for reasons I'm not aware of? I can't really see any downsides. Am I stupid?
|
|
#
?
Jul 31, 2009 03:29
|
|
|
So about signal actuation: is there any kind of system that lets multiple signals coordinate their greens to keep traffic moving, while allowing cross traffic to still be detected and get their respective greens? I am thinking of one particular stretch of road near my apartment that, no matter what time of day you drive it, you'll hit 80% of the red lights, every time. I'm wondering if it's just because the signals are being tripped by cross traffic and cycling constantly or if there is something more that can be done to improve flow on this main road. So more open-ended: what to engineers do about busy roads with multiple signals to keep traffic moving? Pre-timed all the way or only actuated or some combination of the two?
|
|
#
?
Jul 31, 2009 05:57
|
|
|
Basilson posted:Also, I was in Bangkok, Thailand a little while ago and came across what I thought to be an awesome feature of their stoplights which I've never seen before or since. Mounted next to the actual stoplight is a countdown timer which will count down the seconds of each phase of the light. I thought this was great as it means you'll always know how long you have to wait at the intersection before the light goes green. How come I've never seen these anywhere else? Is it actually a stupid idea for reasons I'm not aware of? I can't really see any downsides. Am I stupid? Even though I'm not a traffic engineer, I think I could see a few problems with this. Say the light is saying there are 3 seconds left on a green or yellow light. This would cause a more reckless driver to gun it and try to make it through the intersection. This could also cause some people to preemptively start moving before a red light fully ended. I'm not sure about NZ, but in the US, the time a light stays green is not always set in stone either.
|
|
#
?
Jul 31, 2009 07:02
|
|
|
Cichlidae posted:Looking at it from this level, that interchange certainly looks up to spec. It doesn't seem to have a technical issue, other than being too close to other interchanges and thus causing weaving, though it could be I'm just not zoomed in enough. Little problems like bad pavement, shoddy guardrails, or low-quality pavement markings and signing can really doom a whole stretch of highway.
|
|
#
?
Jul 31, 2009 07:03
|
|
|
I-5's not bad as long as you avoid peak hours. If you can get a 7-3 or 11-7 kind of shift, you'll miss most of the bad traffic in Seattle.
|
|
#
?
Jul 31, 2009 07:25
|
|
|
Basilson posted:
We have these ramp signals in the UK too and I agree that they are dumb. There is a notorious section of motorway near me that has this and it does the same thing during peak hours where it will let about 4 cars on regardless of how much traffic is on the road. The worse thing is they have appeared along any stretch of Motorway that is near any major town or city, so you can't avoid them either.
|
|
#
?
Jul 31, 2009 08:39
|
|
|
Good stuff. Seems like a really fun job.Cichlidae posted:
Can you tell us a bit more about the Cheshire 84 fiasco and highway construction in general? i.e. Do contractors pretty much have free will as they wish with the project? Im sure they have a contract... but it makes me wonder if they can put in non functional catch basins, what is preventing them from using sub par materials in construction? Also HOV lanes, which mostly get blocked by buses and slow cabbies... why not convert them to 2 additional regular lanes? Why/Who in their right mind thought it was OK to have all major highways in CT run right smack through major cities and intersect in hartford?
|
|
#
?
Jul 31, 2009 09:01
|
|
|
Basilson posted:This is an awesome thread, thanks for making it! This method is known as ramp metering, and, when used properly, it can have a pretty big effect on congestion. It would be best to have sensors in the freeway as well, so it could tell entering vehicles when a gap was present, but even dumb meters work. The reason why is that they decrease the volume entering the freeway, causing motorists to use other, less-used ramps, or even take local roads. Additionally, they break up "platoons" of 3 or 4 cars that would otherwise have trouble merging into heavier traffic. Freeway drivers are more likely to slow down to let in one car than three. Smarter ramp metering systems can respond in real-time to the LOS (level of service) on the freeway, letting more cars onto smooth-flowing areas and cutting down on ramps in heavy areas. Like I said, it has to be implemented properly to work. This means installing the system on every single ramp for a relatively large stretch of highway. Connecticut decided many years ago to install a ramp metering system on a single ramp, and I bet you can imagine what happened. People just started using the upstream and downstream ramps, instead, and the congestion was just as bad as it would have been otherwise. I assume there were lots of complaints, as it's been taken down. quote:Also, I was in Bangkok, Thailand a little while ago and came across what I thought to be an awesome feature of their stoplights which I've never seen before or since. Mounted next to the actual stoplight is a countdown timer which will count down the seconds of each phase of the light. I thought this was great as it means you'll always know how long you have to wait at the intersection before the light goes green. How come I've never seen these anywhere else? Is it actually a stupid idea for reasons I'm not aware of? I can't really see any downsides. Am I stupid? It's been shown that telling a user how long until it's his turn decreases the perceived delay. If you know the next bus is coming in 12 minutes, or that you can cross the road in 22 seconds, it seems significantly shorter than if you weren't informed. Unfortunately, when it comes to traffic signals, it's extremely hard to predict when a phase will begin or end. If you check out the signal actuation lesson from yesterday, you'll see that the length can vary significantly. The only time that a countdown timer would work is on a fixed-time signal. Furthermore, we consider it a threat to safety. If you knew that your light was going to turn green in 3 seconds, wouldn't you start inching into the intersection? Lots of countries, even very safety-minded ones like Germany, offer a "warning" signal before the green begins to let motorists put their cars in gear, but we won't implement the same system here. In fact, we go through the effort of putting visors on the adjacent signal heads so people can't see when the other phases are turning red. Once again, safety trumps capacity.
|
|
#
?
Jul 31, 2009 12:19
|
|
|
Economic Sinkhole posted:So about signal actuation: is there any kind of system that lets multiple signals coordinate their greens to keep traffic moving, while allowing cross traffic to still be detected and get their respective greens? I am thinking of one particular stretch of road near my apartment that, no matter what time of day you drive it, you'll hit 80% of the red lights, every time. I'm wondering if it's just because the signals are being tripped by cross traffic and cycling constantly or if there is something more that can be done to improve flow on this main road. I'll address your first question in Coordination 102, but basically, it's "yes and no." Yes, a coordinated signal can call side phases, but it's very limited in how it can deal with them, because of the time constraints. For your second question, the answer is very simple. If two signals are within 1/4 mile, we'll give serious consideration to coordinating them. Any longer than that, and we usually don't bother, because the platoons of cars break up enough in that distance that coordination wouldn't help. There are some obvious exceptions (signals on two sides of a drawbridge, for example), but it's a good general rule to follow.
|
|
#
?
Jul 31, 2009 12:21
|
|
|
|
| # ? Apr 20, 2024 06:17 |
|
|
ItchyDroopy posted:Also HOV lanes, which mostly get blocked by buses and slow cabbies... why not convert them to 2 additional regular lanes?
|
|
#
?
Jul 31, 2009 13:45
|
|