create new account | forgot password


posted by rick on November 5th, 2008 at 7:45PM

Heh ... I know where he took the video. :)

He talked about two things. I completely agree that letting people in before an offramp would alleviate some traffic jams, but for a different reason. Sometimes the adjacent lane gets clogged up because drivers try to force their way into the offramp, so if you let them it will free up the adjacent lane.

I'm not convinced about the traffic wave theory though. If you're intentionally driving slower than the traffic flow, you're not going any faster than you would in stop-and-go traffic. All you're doing is going slower than you should, then catching up with the traffic afterwards.

I think (in Seattle anyway) the main reason for all the traffic are the way the roads are laid out. One example is where there's an onramp, and the same lane becomes an offramp, and then there will be a congestion of cars getting onto the freeway and trying to exit.

The other example is when carpool lanes end right before a high volume/congestion area. He actually mentioned a specific area in the beginning of the video (SR 520 westbound, towards the bridge). It goes from two lanes plus a carpool lane at 60mph to the bridge, which doesn't have the carpool lane and goes at 50mph.
Link | Parent


 
 

posted by dennisn on November 9th, 2008 at 11:39AM

By the way, how close was that video to where you work/live? :P

posted by dennisn on November 8th, 2008 at 11:51PM

The key was leaving lots of space in front of you, not simply driving slower, and for the same reason as "allowing drivers access to the offramp" (which he also cited, I believe)--allowing drivers free passage anywhere "spreads" the traffic around, and makes a clogged lane less likely. And avoiding/minimizing these clogged lanes, at least intuitively, should maximize net flow.

posted by dsk on November 8th, 2008 at 10:14PM

>I'm not convinced about the traffic wave theory though.

Should be (relatively) simple to simulate.