The coordination for the major streets is “time based”. What this means is that there is no constant video detection or a person watching traffic in a room controlling traffic. Basically, there are calculations made taking into account distance, speed and direction and then the controllers at each location turn the coordinated direction green in sequence as the platoon is supposed to arrive.
For time based systems there is a background cycle length of usually 120 seconds. During this 2 minute time frame every movement is assigned a specific amount of maximum time and the movements are served in a specific fixed order. So if you picture a clock with two minutes there is only one time in that 2 minute period where the controller will “look” at the side streets and left turns to see if there are any vehicles waiting. If so, then it will give them a green. If not, then it will give that extra green time to the coordinated movements and “look” at the next movement in the sequence. So if you are the first car at a side street or left turn and you happen to get there just after the controller “looks” to see if there was anyone there (and there wasn’t one second before you got there), then it will not look again for about 2 minutes. This can lead to the belief that the signal is not working properly.
Coordination is designed to move large groups of vehicles over long distances with minimal stops. This helps make up for the “extra” time you might have spent waiting on the side street and if you stay on that street for a reasonable amount of time, you will actually be ahead time-wise, gas-wise, frustration-wise and vehicle wear and tear-wise.