Pet peeve #2 about roadway intersections

[ewmon]

We’ve all experienced it — stop lights activated for cross traffic (usually one car, maybe two) that, in the meantime, were able to make a right turn on red, resulting in the normal traffic flow needlessly interrupted. I’ve seen this happen a million times (at least ).

 

Modern traffic signal algorithms delay several seconds before initiating the sequence to change the traffic flow. Why then, at the end of the delay and before initiating the change sequence, doesn’t the algorithm simply check the cross road one last time to see if the cross traffic is still there? I’ve programmed computers, and this is an extremely minor re-write of the algorithm.

 

It would save time, improve gas mileage, reduce aggravation, reduce brake wear, etc. Doesn’t this make sense?

 

(This reminds me of Charles Frasier-Smith (book “The man who was Q”) who was working on a spy transmitter for Britain during WW2. He went to a battery manufacturer to ask if they could design a battery with a much greater energy density, and the maker said it was super easy to do. Frasier-Smith basically chewed out the maker for purposely making inefficient batteries at great profit while the rest of Britain suffered.)

[Phi for All]

I've seen a lot of them do just that. You can tell because the crosswalk timer for pedestrians counts down to 0 and then starts back over again but the light stays green.

 

But you're right, some don't. I'll bet it's a setting that has been overridden for some reason, or perhaps something besides a car can trick the sensor into thinking there's a car. How sensitive are the sensors?

[ewmon]

The new sensors now detect bicycles without having to lay them down. I remember having to lay them down on the old sensors. But this one little step in the algorithm could save a lot of resources now going to waste.