Flutterby™! : Transportation costs too much 2012-02-07 22:41:34.144035+00

Transportation costs too much

2012-02-07 22:41:34.144035+00 by Dan Lyke 7 comments

Worth a revisit: David Levinson: Transportation costs too much:

It sure seems like we should be able to build this cheaper. Think about it, $175K for 12 lightbulbs on a timer. What's going wrong?

Some of his examples are a bit underdescribed. For example, the streetlight setup in Chicago has to work 24 hours a day, in weather ranging from 30 below zero to 100 above at different times of the year. This a bit more than 12 lightbulbs and a timer.

I always thought signalizing an intersection cost more like $1M. I guess that's intersections with protected left turns and cross walks.

I ended up sitting next to a transportation engineer on a flight to DC recently. I asked him a question I've been pondering for a while. Why don't signalized intersections have a port that a police cruiser can plug into to power the signals? The problem, if power goes out, the cop has to stand in the middle of traffic and direct traffic. Typically, the cruiser is sitting on the side of the road, engine running and lights flashing.

The engineer said that traffic lights run on something like 1000V and that even the heftiest alternator in a cruiser won't be able to generate the necessary current.

My response was, "Is that kind of current necessary with the LED-based signals normally used today? And is it possible to just power a subset of the LEDs during such a situation?"

I could almost hear the gears in his head start to crank to life. He actually had never thought about that. The current design for traffic signals is still based the legacy of high-power incandescent lamps. For whatever reason, no one is really thinking about how LEDs can disrupt the design of traffic signaling.

Working for the Federal Government, I actually understand EXACLY how this kind of legacy design artifact persists. Look at the production reports for the US Topo program at the USGS. Why the heck does the USGS measure the process in terms of 7.5 minute quadrangles? And that doesn't even start to question why we are making PDFs in the first place.

#Comment Re: made: 2012-02-10 00:33:23.769997+00 by: TheSHAD0W [edit history]

1000V is doable, depending on the amperage; I suspect a voltage converter which would've been unfeasible ten years ago would be practical now. (Look at how 125V converters have scaled up in power lately, and that wall warts now use switching power supplies.)

#Comment Re: made: 2012-02-10 00:42:55.519176+00 by: Dan Lyke [edit history]

1kV is just an off-the-shelf inverter and a 1 to 8 transformer. Even today you could put that together at Radio Shack. If they're still using more current than a police cruiser can generate then I know where we need to start with fuel efficiencies, and I've some choice words for the legislator who's making me put motion and light sensors on my outdoor CFLs...

But, yeah, we've got a whole bunch of "engineers", especially in the transportation space, who are several decades behind modern ~~engineering~~ technology. Part of this is a world in which an automobile platform only ships every 7-10 years ("new model year" my ass), part of this is just laziness.

Cost of Hardware for installation of traffic light: $32,000 Cost of labor to install traffic light: $12,000 Cost of Engineering, traffic studies, administrative costs, grant applications and no telling what other red tape required to get approval through the state DOT and federal approval: $132,000.

Documentation and approval costs rarely ever get factored in by those on the outside looking in. There is a cost to the paper shuffling required and it needs to be accounted for in the project.

I think, though, that the reason the engineering costs so much is that the decisionmakers are still thinking in terms of old technology. And maybe that is really that they're also thinking in terms of old engineering.

Or maybe it's just the cost of the politics involved; when the cost of "failure" (which may actually be a success, but is perceived as bad because it's a change and people don't accept change well) is a thousand outraged voters screaming for blood, it's hard to advance the state of the art.