Interesting experiment
Ok so I was driving down El Camino to Stanfurd Hospital yesterday and I noticed something interesting. Between my parents' house and Stanfurd Hospital, I spent a grand total of 8 minutes waiting at stop lights. I kid you not (I timed it with the stopwatch on my cell phone).
So I thought of an interesting experiment on traffic flow. Pick a route. It has to be long enough to be able to average out stop signs and lights and such (you'll see what I mean in a bit). Count the number of stop lights/stop signs, and calculate the distance of the route.
Then get a TON of drivers, each of whom has a passenger with two stopwatches and a clipboard. On this clipboard is a piece of paper with a table where the passenger will be listing the amount of time spent stopped at stop signs only (which they can guestimate by looking at their running stopwatch- there shouldn't be so many stop signs on a route that the estimation is that off..). They should also take note of any other out of character delays, like getting pulled over for speeding tickets, or stopping to get something to eat.
Starting at midnight on a Sunday morning of an average week, send one car every thirty seconds to drive that route, and send cars every thirty seconds for a full week. The speed rule is that they can drive as fast as they want so long as they don't exceed 7mph over the speed limit. 10mph if it involves a freeway.
The passenger hits start on their first stopwatch, and doesn't hit stop on that stopwatch until they reach the destination.
The second stopwatch will be used to tally up time spent STOPPED at stop lights. When the driver stops at a stoplight, the passenger pushes start. When the driver begins to accelerate again, the passenger pushes stop. They continue this pattern (without resetting the second stopwatch) every time they stop at a stoplight on this route. By the end of the route, if the passenger is using a conventional stopwatch, it should read the total time spent stopped at traffic lights.
Then, at the end of the route, each tandem records a couple different numbers: the total duration of the trip, minus unexpected delays (like non-routine traffic stops or something, but that wouldn't happen on an average week.. also they should subtract getting pulled over time and other stuff like that). That will give the total time it took to drive from Origin to Destination. Then, the tandem should record the total amount of time spent at stop lights, and then they should record the total amount of time spent at stop signs.
The reason I send a driver every thirty seconds, is because that's a low enough frequency such that it wouldn't effect the average congestion for that time of the week, but a high enough frequency to cancel out outlying data points due to one person's bad luck with traffic lights.
So then you plug all these data points into a point graph. The X-axis would be time, and the Y axis would be a scalar. For each driver, you are going to create different dots (shape or color, whatever) for the following variables:
T = total time passed in transit from Origin to Destination, minus the getting pulled over type time
L = time spent stopped at traffic lights
X = time spent stopped at stop signs
t = T - L - X = time spent actually moving along the route toward the destination.
v = D/t (where D = the distance of the route) = average speed WHILE MOVING over the entire course of the route. The most hilarious results will occur if you pick your route such that it only contains one speed limit. Like driving from Sunnyvale to Palo Alto on El Camino Real.
V = D/T = average speed TOTAL for the whole route.
So for each driver (or each 30 second unit), you plot the corresponding point for each variable.
Over the course of the week, you should be able to draw different trends and averages to see how traffic works at various times of the day and how it might vary from day to day.
For instance, you'll see how much time per mile the average driver can expect to wait at traffic lights at 11am. Or what a driver's average speed will be along that route if they are travelling it at 4pm.
Yesss.. yesss.. If we're going to waste taxpayers' money on stupid studies, we might as well do a stupid study that is at least INFORMATIVE. None of this "Years of research and millions of your dollars show that eating at McDonalds will make you fat" shit.
So I thought of an interesting experiment on traffic flow. Pick a route. It has to be long enough to be able to average out stop signs and lights and such (you'll see what I mean in a bit). Count the number of stop lights/stop signs, and calculate the distance of the route.
Then get a TON of drivers, each of whom has a passenger with two stopwatches and a clipboard. On this clipboard is a piece of paper with a table where the passenger will be listing the amount of time spent stopped at stop signs only (which they can guestimate by looking at their running stopwatch- there shouldn't be so many stop signs on a route that the estimation is that off..). They should also take note of any other out of character delays, like getting pulled over for speeding tickets, or stopping to get something to eat.
Starting at midnight on a Sunday morning of an average week, send one car every thirty seconds to drive that route, and send cars every thirty seconds for a full week. The speed rule is that they can drive as fast as they want so long as they don't exceed 7mph over the speed limit. 10mph if it involves a freeway.
The passenger hits start on their first stopwatch, and doesn't hit stop on that stopwatch until they reach the destination.
The second stopwatch will be used to tally up time spent STOPPED at stop lights. When the driver stops at a stoplight, the passenger pushes start. When the driver begins to accelerate again, the passenger pushes stop. They continue this pattern (without resetting the second stopwatch) every time they stop at a stoplight on this route. By the end of the route, if the passenger is using a conventional stopwatch, it should read the total time spent stopped at traffic lights.
Then, at the end of the route, each tandem records a couple different numbers: the total duration of the trip, minus unexpected delays (like non-routine traffic stops or something, but that wouldn't happen on an average week.. also they should subtract getting pulled over time and other stuff like that). That will give the total time it took to drive from Origin to Destination. Then, the tandem should record the total amount of time spent at stop lights, and then they should record the total amount of time spent at stop signs.
The reason I send a driver every thirty seconds, is because that's a low enough frequency such that it wouldn't effect the average congestion for that time of the week, but a high enough frequency to cancel out outlying data points due to one person's bad luck with traffic lights.
So then you plug all these data points into a point graph. The X-axis would be time, and the Y axis would be a scalar. For each driver, you are going to create different dots (shape or color, whatever) for the following variables:
T = total time passed in transit from Origin to Destination, minus the getting pulled over type time
L = time spent stopped at traffic lights
X = time spent stopped at stop signs
t = T - L - X = time spent actually moving along the route toward the destination.
v = D/t (where D = the distance of the route) = average speed WHILE MOVING over the entire course of the route. The most hilarious results will occur if you pick your route such that it only contains one speed limit. Like driving from Sunnyvale to Palo Alto on El Camino Real.
V = D/T = average speed TOTAL for the whole route.
So for each driver (or each 30 second unit), you plot the corresponding point for each variable.
Over the course of the week, you should be able to draw different trends and averages to see how traffic works at various times of the day and how it might vary from day to day.
For instance, you'll see how much time per mile the average driver can expect to wait at traffic lights at 11am. Or what a driver's average speed will be along that route if they are travelling it at 4pm.
Yesss.. yesss.. If we're going to waste taxpayers' money on stupid studies, we might as well do a stupid study that is at least INFORMATIVE. None of this "Years of research and millions of your dollars show that eating at McDonalds will make you fat" shit.