Archive for May, 2011

As discussed in Traffic, sometimes the best way to eliminate tension over ‘late merging’ is to simply eliminate the concept, as the Illinois DOT seems about to do.

“Once people get used to it, it seems to work well,” Wegmeyer said. “As long as traffic alternates at the merger, it should go more smoothly.”

Of course, that “as long as” still presents challenges, as those people in the “open lane” may still feel some sense of priority over those they view as late-arriving interlopers.

(thanks Darrin)

My latest Slate column is up, examining the problems with “Children at Play” traffic signs (the headline, which is never the writer’s, may overstate things a bit…)

If the sign is so disliked by the profession charged with maintaining order and safety on our streets, why do we seem to see so many of them? In a word: Parents. Talk to a town engineer, and you’ll often get the sense it’s easier to put up a sign than to explain to local residents why the sign shouldn’t be put up. (This official notes that “Children at Play” signs are the second-most-common question he’s asked at town meetings.) Residents have also been known to put up their own signs, perhaps using the DIY instructions provided by eHow (which notes, in a baseless assertion typical of the whole discussion, that “Notifying these drivers there are children at play may reduce your child’s risk”). States and municipalities are also free to sanction their own signs (hence the rise of “autistic child” traffic signs).

As someone with an interest in traffic, and a more casual interest in operations research, I was curious about the order history on a recent item from Apple I had purchased. As you will note from the illustration above, the item was shipped from FedEx’s Memphis hub to Newark, where it then made its way to Brooklyn, where I live. But the item wasn’t delivered. Why? The reason given explains everything, yet explains nothing: “Package not due for delivery.” OK. So it wasn’t due for delivery. But couldn’t FedEx have delivered it anyway, given that it was in my home borough?

No. The item then went back to Newark, only to finally be shipped, once again, to Brooklyn, where it finally arrived on my doorstep. Now, I’m no OR genius, and there may be some variant of the Traveling Salesman Problem, or some intricacy of routing and logistics that I’m missing here, but why, for an industry always trying to root out inefficiencies (e.g. UPS’ famous ‘left-turn’ software), would it send my product on an extra round trip, bloating its inventory for a few more days? Perhaps some reader can enlighten me.