Tag Archives: networks

Bad Data: “How Not To Measure Traffic Congestion—Hold the Hyperbole, Please! “

The INRIX report makes several other basic errors. It describes traffic congestion as “gridlock,” a greatly abused term. Gridlock refers to a specific situation in which vehicles in a network are totally stuck due to clogged intersections. It almost never occurs.

Article: How Not To Measure Traffic Congestion—Hold the Hyperbole, Please! | Planetizen: The Urban Planning, Design, and Development Network.

I’ve only just started following Litman.  I should have followed him sooner, I think.  I can’t imagine that gridlock has actually ever happened in Oklahoma.

“Justice” (Season 1, Episode 8): “A System to Cut City Traffic That Just Might Work”

Article: A System to Cut City Traffic That Just Might Work | WIRED.

I like the idea of the token system.  This is not unlike token signaling systems used on British railways on single track branches – if you have the token, you can enter the branch.  The difference here (aside from it being cars, not trains) is that there is more than one token.

At first glance, I like the idea of limiting the number of cars in the congested area, rather than just charging a fee to anyone that comes in.  The challenge, though, is that you’d need to know how many tokens is the “right” number; a non-trivial exercise.  With a congestion charge, there isn’t a hard limit to the number of vehicles because you rely on prices to discourage vehicles from entering the zone.   If you set the prices appropriately, and that has to include a rise in price as more vehicles enter the zone, then it should be OK.  If you set the price too low, however, then more and more vehicles will enter the congestion zone.  You’d have the same congestion and a bunch of revenue.

The thing I don’t like about this method is the idea that you can change the congested area on the fly.  How does this work?  If you’re in a zone that was just designated as “congested” by the controller, do you have the chance to move out of it?  How would you even know?

This is not unlike the Star Trek: The Next Generation episode “Justice” (Season 1, Episode 8), where there are random “punishment zones”.   Anyone caught in them is, of course, put to death (because Star Trek).  Guess where Wesley Crusher ends up?

Though the penalty is less harsh here, I don’t see the “moving zone” working out very well.  Different areas on different days?  Sure.  But no changes during the day.

People are stupid: “Traffic Light Study Reveals Serious Hacking”


Article: Traffic Light Study Reveals Serious Hacking Risk | MIT Technology Review.

“First and foremost, traffic-system administrators should not use default usernames and passwords.”

Unlike traffic signals in my local area, which are networked with fiber (!), these were networked over WiFi.  If details — 5.8 GHz — in a different article are to be  believed, then this is either IEEE 802.11a or 802.11n which are common consumer WiFi specs.

Encrypted or not, leaving credentials at the default is stupid.  In the field of traffic signaling, it probably ought to be a firing offense.  Anyway, change those and risk would be reduced by, oh, let’s say 99%.

Another quote:

“They continue doing the same mistakes that software vendors did 10 years ago.”

No, that’s been going on for a long time. The default credentials for Digital’s VMS OS were SYSMAINT / SERVICE and that’s got to be from 30 years ago.

If you leave the doors unlocked and open, people will walk through them

Yes! “Walker: If a carpenter can’t be a hammer opponent, then I can’t be a streetcar opponent”

Human Transit: If a carpenter can’t be a hammer opponent, then I can’t be a streetcar opponent.

Just going to set this here.  I have been strongly influenced by Mr Walker’s writing and usually agree with him on most issues including this one.  And just because I got him to sign a copy of his book doesn’t mean I’m a fanboy.