Comment

Comment on A new study found adaptive traffic signals powered by big data reduced peak-hour travel times by 11% in China’s 100 most congested cities – saving 31.73 million tonnes of CO₂ annually.

<- View Parent

barsoap@lemm.ee ⁨1⁩ ⁨day⁩ ago

this completely ignores larger traffic patterns like arterial roads.

with your idea you are guaranteed to get massive gridlock all along the major roads.

How. Seriously. Show me an adaptive traffic light dumb enough to cause gridlock. Not to mention that gridlock and having arterials, road hierarchies in general, are kinda incompatible with each other and most of the world doesn’t use grids in the first place.

And it’s not like we don’t have central control over here – it’s that all the information necessary to make decisions for a single traffic light is available right there, at the traffic light, because it is impossible to have traffic (or the absence thereof) and that not carrying the necessary information. Roads are wires, so to speak. Central control could make those decisions, but as local information suffices, why would it, regarding traffic lights it’s generally only monitoring. Central control can override things, things like ambulances influence traffic lights in a non-local manner (which is a luxury problem because they are allowed to cross on red anyway), but for basic operation central control could vanish and you wouldn’t see a difference, when a light loses connection but not power it just keeps on operating. Things like information systems telling people where to park need non-local control because they need non-local information.

source

Sort:hotnew top

9bananas@lemmy.world ⁨1⁩ ⁨day⁩ ago
wow, no.

none of what you said is actually true.

“gridlock” happens in non-grid layouts too, the english name is just taken from american road patterns.

“show me…” no. YOU made a claim (that local information suffices, which is a VERY bold claim), so it’s on you to prove that local information suffices.

roads are absolutely NOT “like wires”; they are like pipes. which is why civil engineers commonly use fluid dynamics to simulate traffic.

the rest of what you said is irrelevant to everything else.

seriously, if you make a claim contradicting both the very premise of the post, and common knowledge on the topic, then at least provide a source for that claim, lr explain WHY you think your claim is true.

“all the information is there” is not enough information to verify the claim; it’s a wild guess without evidence to back it up.

if shit where THAT simple, we’d have it figured out 50 years ago… it’s almost like this isn’t the simple problem you desperately want it to be…
source
- barsoap@lemm.ee ⁨1⁩ ⁨day⁩ ago
  
  “gridlock” happens in non-grid layouts too, the english name is just taken from american road patterns.
  
  I said something about road hierarchies, you ignored it.
  
  “show me…” no. YOU made a claim (that local information suffices, which is a VERY bold claim), so it’s on you to prove that local information suffices.
  
  These systems are in operation. You claimed they lead to gridlock. What I get from the Chinese experiment here is that they collected data, threw an optimisation algo on it, and then adjusted local parameters, “err towards giving more green time in this direction” type of deal. They’re still going to use the same type of adaptive, local-control system that’s becoming increasingly common in the last decade.
  
  roads are absolutely NOT “like wires”; they are like pipes. which is why civil engineers commonly use fluid dynamics to simulate traffic.
  
  Vehicles travelling on roads constitute information travelling over roads. Are you trying to deliberately misunderstand what I’m saying. You do not need to look at the app of the parcel carrier to know that your parcel arrived, it’s right there on your doorstep. That’s information. Metaphorically, thus, package delivery trucks are wires.
  
  “all the information is there” is not enough information to verify the claim; it’s a wild guess without evidence to back it up.
  
  if shit where THAT simple, we’d have it figured out 50 years ago… it’s almost like this isn’t the simple problem you desperately want it to be…
  
  50 years ago we neither had the sensors we have now, nor did we have the processing power to use it. Traffic light control was often still done electromechanically. “Adaptive” means a lot more than “pedestrians have a button and there’s an induction coil to detect a car”. Those systems actually solve the local problem optimally which, in the case of traffic management, means that the global problem is solved optimally because the problem has optimal substructure. Don’t ask me for a proof of optimal substructure I just sat on a municipal traffic committee, I don’t actually design those systems. When you observe, them, in low traffic situations, they’re green for everyone because they switch as soon as they see someone arriving and noone else needs to be let through. In higher traffic situations they prioritise throughput, but make sure to not let waiting time for others get exceedingly long, or allow large backups.
  
  source