Why aren't more traffic light systems synchronized? [Traffic flow physics]

As you can imagine, London has a really big problem with this. An extensive study was done in 2009: https://www.london.gov.uk/sites/default/files/traffic-signals.pdf

One conclusion was:
"6.1.6
������ Traffic signals generally provide significant benefits to road users;
������ It would, however, be beneficial to switch-off traffic signals at some junctions at particular times of day;
������ In particular, there would be a benefit at the junctions studied from switching off during the off-peak, after a full safety assessment.

There was a proposal for a trial but I can’t see if it was done.

Pedestrians were a major consideration as they randomly interrupt traffic flow. That was not taken into account for the study.

I just spotted this in the news:

http://www.msn.com/en-gb/news/uknews/gadget-which-turns-all-traffic-lights-green-trialled-in-uk/ar-AAao1TZ?ocid=UP74DHP

Optimum for heavy traffic.

That brings up a pet peeve of mine.

People complain that there are too many drivers running red lights, but generally they only do that when, like you, they are stuck at a long red light for no reason, since there is no cross traffic. So now they invented red-light cameras to watch the intersection and automatically mail out tickets for going thru the red light*. But that same technology could be used to change the red light to green when there is no cross traffic! So the government could either assist drivers by making the traffic lights demand-driven, or they can treat drivers as suckers to be milked for more fine revenue. Which does your government choose to do?

*Luckily, here in Minnesota a Judicial ruling forced them to turn those off. For now. The company who makes them keeps trying to get our Legislature to change laws to legalize them.

If you hit all greens, they lose out on all the potential revenue from ticketing you. In a big city, those tickets add up. It’s sort of like a road that any experienced driver can safely drive 45mph on being set to a 30mph limit. It pays the city big dividends to keep traffic signals and speed limits less than ideal.

This is the system used to control traffic flow - SCATS.

"SCATS® is a fully adaptive urban traffic control system that optimises traffic flow. Its self-calibrating software minimises manual intervention, which can result in substantial operational cost savings.

SCATS has been continually developed for over 40 years and sold to 27 countries, delivering real and measurable reductions in road travel times and delays."

It’s not the only one. This UC/Caltrans report reviewed several. Los Angeles created its own.

Not only less revenue, also less opportunity to scan for fugitives from the law, uninsured drivers, etc.

I always assumed (but have no cite) that the majority of red light runners are people who just missed the yellow cycle after a green - they went through to prevent having to stop for a full cycle.

Those are certainly the ones responsible for the accidents. The other source of red-light running accidents are the drunks, stoners, or fugitives that just blast through a red without slowing down.

Folks tired of waiting at an empty intersection at 11pm who look both ways, see no cars, and proceed through a red are not the problem red light cameras are aimed (heh) at.

Again, heavy traffic on what road? How far apart are the intersections spaced on this road? How many lanes does it have? How much traffic do the cross streets get? Does the road have a bias towards people traveling in a particular direction at particular times of day? Is the traffic that’s using that road usually going a long distance on that road before turning, or a short distance? When they do turn, are they usually turning on one particular side of the road? Is the road used mostly by small cars, large trucks, or a mixture of both? Are there residences along the road? Businesses? Empty land? Does the road have a median or other divider, or is it flat pavement all the way across? All of these questions are going to be relevant for determining the optimum speed for a road.

I live in a land-locked city. There’s no open land in any direction: population growth is slight, and attributed to apartment houses replacing single dwellings. There are no new streets being built, but traffic used to shift to other through streets when potholes grew to maximum. When I first moved here, traffic was horrible. Flash forward about 5 years; a study was made, and lo and behold! traffic wasn’t snarled again until about 4 or 5 years ago.

Whenever I see a moron hanging up traffic at a turn lane they are:
-Texting. I bet that’s a third of the time.
-Delayed by somebody on the cross street making a left turn after the arrow is red. Another third of the time, or more, depending on the intersection.
-Screwing around with, I assume, the radio.
-Applying makeup/combing hair, but that might be one of thirty instances.

People in our area were not screwing around with smartphones when the traffic studies (and later, adjustments) were made. I firmly believe that it takes a single moron to screw up a city’s traffic.

“Optimum” refers to the maximum rate of flow, or the greatest number of cars per second - like car amps. For minimum resistance you would (presumably, since this is just a thought experiment for me) have a wide, single laned, smooth surfaced, no intersection, no driveways, no visual distractions. To increase throughput, we would look at carrying only cars (lower inertia, smaller breaking distance). I am sure motorbikes would provide an even higher rate, but lets stick to vehicle similar to average traffic.

If you want to increase the number of cars crossing a point (cars/second) then the speed would need to increase, however the faster the cars go the more they spread out (no auto nav cars yet) to adjust for increased breaking distance. Add both these plots on the graph and it must peak around 80km/hr. Maybe there is another non-local maximum closer the speed of light, that we could safely discount.