Tailgating doesn't help

^ Bottom line: play nice.

 

Another article on the same study --

CNN: Cruise control designs could solve traffic jams, study suggests - Dec. 14, 2017

 

I must challenge a number of his claims in this specific Seattle traffic jam, leading up to the Northbound I-5 Express Lanes. I have experienced that many times too, though not as part of daily commuting.

He claims this backup is caused by merge fights. I disagree, asserting instead that the basic cause is the road discontinuity at the entrance (narrow, twisting, less forward sight distance, and drastic lighting transition from sunlight to darkness) causing unfamiliar drivers and 'nervous Nellies' to space out and slow down just as they enter the underground portion, greatly reducing lane throughput capacity. E.g the I-5 main thru lanes (just two at that point, due to long ago design blunders later cemented by the foundations of the overhead Convention Center) have a capacity of about 2000 vehicles per lane per hour when not backed up farther north. While the Express Lanes also have that capacity farther north, and multiple lanes of it too, this specific entrance -- the highest volume entrance to the whole system -- effectively does not. This entrance is a severe choke point, enough so that numerous vehicles that otherwise could use it, instead stick to the main line, or if that is backed up too much, even take the in-between lane to the Seneca St exit (he marked it in the video), drive two blocks on surface streets, and re-enter at the University St ramp.

Mr. Beatty perfectly, if unintentionally, illustrates this himself at 6:25 in the video, where he takes the ramp more than 4 seconds behind the car ahead, a typical spacing during these slowdowns. (His video shows only about half of the choke-point segment.) This following gap is very common in slow traffic, thus limiting this entrance to a measly 900 vehicles per hour. I know quite well that for drivers familiar with it, we can take that ramp at the same speed and car spacing as the main line, i.e. ~2000 vehicles per hour with car spacings a bit under 2 seconds. But the nervous reactions of less familiar drivers at the physical transition, plus the common hesitancy to speed up as traffic clears ahead, cause the reduced throughput and thus a jam. The merge fights he discusses lengthen the jam, but don't normally cause it.

He asserts that his driving pattern causes no more than a 10 second delay, just 1 second per car. But his absurd 1 second per car claim is clearly at odds with his own video. He let in 10 vehicles immediately in front, and they let it at least 2 more visible on tape, and probably plenty more obstructed from view. At the typical 4 seconds per car, that means 48+ seconds of delay added in that single mile of road.

The other wave concepts discussed in this thread concern stop-and-go waves on road sections without any meaningful injection of added traffic or any fixed physical triggers. Killing the standing waves increased both the throughput and the average lane speed.

But this Express Lanes entrance is a very different animal, with very large numbers of cars being injected into that lane. Mr. Beatty is killing the stop-and-go effect, reducing wear on both his clutch and his nerves. But with no video of traffic behind, nor from overhead, he provides no evidence at all of any other benefit whatsoever. Having entered that lane myself numerous times with 2-second spacing, I must claim that his 4-second spacing at the entrance split is part of the cause of these backups.

I believe a much more effective solution would be to straighten and improve the entrance, improve sight distance, brighten the dark concrete and add artificial lighting to smooth the now-abrupt light-to-dark transition. This would reduce the physical triggers that cause nervous and neophyte drivers to slow and space out, which is the primary cause of the standing wave. Then reducing that wave and increasing lane throughput would greatly reduce the volume of 'cheating' that amplifies that standing wave.