Sounds good, but I have to laugh at one particular issue...
While this seems to work in some places, for some reason other drivers just don't "get it." See this: http://www.signonsandiego.com/news/j...1m12fixit.htmlAs I noted a few months back, there is considerable evidence that removing all traffic controls – lights, signs, road markings, and even the distinction between streets and sidewalks – can actually make traffic move more smoothly, as well as cut down on the number of accidents and increase the area’s economic vitality.
two months after the slurry seal has been poured, shouldn't the lane markings and turn arrows that were covered by the new asphalt be repainted?
John Hinkle, who lives near Pomerado Road and Higa Place in Rancho Bernardo, waited several weeks before calling the city about the missing road markings on Higa, a short side street.
Hinkle noticed that some drivers were having trouble at the crossing without the extra guidance. Once when he was waiting at the traffic signal to drive straight across Pomerado Road, a motorist in the unmarked left-turn lane drove straight across, too.
One would hope motorists could just figure out such simple things... but apparently the "need to exceed" is just enough to overrule even common sense...
I agree that we -- meaning traffic engineers -- really haven't nailed down such dynamics with a high degree of confidence. Which should be expected given the complications of human behavior. Consequently, we would see a lot of examples where X works and Y works even though on the surface X & Y are incompatible.
Recently, I had an opportunity to discuss a local project with a traffic engineer. One of the interesting tidbits was that accident prediction is pretty rough with the present tools. Consequently, when roads are being designed -- or redesigned in my case -- increases/decreases in the number of accidents and its effect on travel times are omitted from the simulations. I thought that it was an interesting bias to say the least. At least that is what I understood. Note that there seem to be several classes of simulation models -- which I have never really examined. So now that I think about it, I might have interpreted his comments too broadly.
Regarding the simulation software... if traffic flow nor reduced accidents are not metrics for designing roads, what ARE the metrics? What factors does it analyze as being optimum?
In this case, the focus was on travel times on I-66: the freeway that leads out of DC to the west. In its defense, it did take into account a lot of factors such as population growth and contrary effects of the new design to determine a net effect -- one would have "better" merging but drivers would be enticed to use an auxillary lane to jump the queue creating more chaos. I think that several aspects of the model were proprietary so it wasn't clear to me how aspects were modeled. But I think that a limited set of variables are truly dynamic -- i.e., determined within the model -- while a lot of important factors where predetermined outside the model. For instance, I gather that spillover traffic from alternatives and general growth is outside the model. A huge problem, IMO, is that there were no measures of robustness that I could discern. That is, the engineers were unable to give me a confidence interval or any real examples where the model was tested with rigor in similar empirical situations. I thought that it was particularly problematic since the theoretical net effect -- i.e., does it increase or decrease travel times -- was in no way predetermined.
Of course, everything above has the caveat that I actually understood the slides and conversation.