10 May 2012

Building a "scientific ghost town" to test technology

As reported by Boston.com:
A $1 billion city without residents will be developed in Lea County near Hobbs, officials said Tuesday, to help researchers test everything from intelligent traffic systems and next-generation wireless networks to automated washing machines and self-flushing toilets.
Hobbs Mayor Sam Cobb said the unique research facility that looks like an empty city will be a key for diversifying the economy of the nearby community, which after the oil bust of the 1980s saw bumper stickers asking the last person to leave to turn out the lights...

Bob Brumley, senior managing director of Pegasus Holdings, said the town will be modeled after the real city of Rock Hill, S.C., complete with highways, houses and commercial buildings, old and new. No one will live there, although they could as houses will include all the necessities, like appliances and plumbing.
The point of the town is to enable researchers to test new technologies on existing infrastructure without interfering in everyday life. For instance, while some researchers will be testing smart technologies on old grids, others might be using the streets to test self-driving cars.
A billion dollars.  I don't understand what can be tested in this town that can't be tested with computer modeling.  I thought it was a pork-barrel earmark project, but it seems to be privately funded.  I don't understand.

(and how do you test self-flushing toilets if there are no people?)

12 comments:

  1. I've heard of quite a few of these popping up around the world. I've read articles that say Russia is building one, and I think China too.

    ReplyDelete
  2. "I don't understand what can be tested in this town that can't be tested with computer modeling."

    Answer - Hardware

    Nothing, NOTHING, ever transfers from a computer to actual physical, usable hardware without a great deal of modification.

    -

    an Engineer

    ReplyDelete
    Replies
    1. So then presumably the development companies would rent time/space in the city to test their products?

      Delete
    2. That seems to be the plan, yes.

      After a bit of research I found their website, which contains a nice overview of the facility (including an animation):
      http://www.cite-city.com/Facility/Main/CITE_Design.html

      I personally think it's a brilliant idea! Although I can't help but wonder whether removing the residents will affect the results of some experiments in ways they perhaps didn't predict.

      In any case, as anon said, testing hardware in a realistic environment is absolutely crucial but often impractical (mainly because of stringent safety regulations). So I think this business plan will probably meet some demand.

      Delete
  3. I have to support what Anonymous said. Having worked in supercomputing, you can only model a few variables. While the simulation can prove the concept, it cannot prove out the reality. As Yogi Berra so eloquently put it, "In theory, there is no difference between theory and practice. But, in practice, there is."

    To put it all in perspective, Boeing spent about $32 billion to develop the 787. While they could test many things via computer simulation, they still had to build it and test fly it. Spending $1 billion to build a small city to test out systems that can also be used over and over again, doesn't seem so wasteful.

    Can you imagine being stuck in traffic in one of the big cities in the US, because someone wanted to test an automated traffic control system that had a couple of bugs in it? How about a truck catching fire? I wouldn't want to test that live in a living city. But that fire could have far reaching consequences, such as repair crews messing up when they come in to resurface the road and perhaps damaging or not reconnecting a communications cable. How about dealing with autonomous vehicles?

    One thing for sure, you can simulate many things in a computer, but you cannot simulate how an organic entity (people or animals) might choose to react in a given situation. Consider this: how often do you see someone stuck at a traffic light turn into a corner parking lot in order to circumvent waiting at the light? Most computer simulations don't take behavior like this into account when modeling traffic. They assume everyone is going to stay nicely on the road and play by the rules.

    People complain how the government messes up because it wasn't prepared. This model could be a place where the government can try out new things in order to be properly prepared for whatever.

    ReplyDelete
  4. I pro test city but heres where I think the money is going to waste. They are going to spend a billion dollars to build something that already exists. Abandoned towns and especilly suburban housing projects are all over the place. There no way its cheaper to build one from the ground up than it would be to buy one of those up.

    ReplyDelete
    Replies
    1. It is probably cheaper to build a new, compact city simulator on vacant land especially with the sensor wiring, test equipment, monitoring capabilities, video cameras etc. let along the improved infrastructure than to retrofit an existing town/suburb.
      see:
      http://www.thestate.com/2012/05/08/2268407/hobbs-nm-picked-as-site-of-scientific.html

      Delete
  5. http://www.garbagewarrior.com/
    Renegade architect Michael Reynolds and his green disciples have devoted their time to advancing the art of “Earthship Biotecture” – passive, solar, off-the-grid, sustainable housing. Earthship n. 1. passive solar home made of natural and recycled materials 2. thermal mass construction for temperature stabilization. 3. renewable energy & integrated water systems make the Earthship an off-grid home with little to no utility bills. Biotecture n. 1. the profession of designing buildings and environments with consideration for their sustainability. 2. A combination of biology and architecture.
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    I recommend this movie. It's also posted on youtube.
    Gee, if only he would have had a billion dollars he probably wouldn't have suffered so much for obviously being the wrong kind of visionary.

    ReplyDelete
  6. If google can test their driverless car on public highways, I'm thinking probably not much need for a ghost town. It probably has something to do with a tax loophole for "investment" in R&D.

    ReplyDelete
  7. I'll chime in as another engineer that this sounds like a pretty cool idea. Computer modeling has major limitations. Any good modeler will tell you that it's just another tool in the toolbox - it's never, ever good enough on its own.

    ReplyDelete
  8. As someone who lives in Rock Hill, SC, I have one question: Will I be able to recognize my house in the fake city?

    Probably not, but it's kind of interesting.

    ReplyDelete