West Point Bridge Design Contest 2012

I was wondering why my old post West Point Bridge Designer 2011 was getting so many hits recently, all from searches.  Then it occurred to me that the 2012 contest must have been released at West Point Bridge Design Contest.

I think that kids who have been assigned to do the contest are looking for ways to cheat by copying someone else’s design. Although the contest has changed the design parameters and cost functions somewhat this year, a good solution to last year’s design is also similar to  a pretty good solution for this year’s.  They’d have to change the cost functions a lot to make a different design be much better, but they are trying to stick fairly close to standard engineering practices.  I don’t believe their cost functions, though, as no one around here is able to make a 2-lane bridge capable of carrying trucks for only $166k, which is about what the cheapest design I came up with costs.  I think that they left out the cost of installing the bridge, and may be using prices from the 1960s for materials, as I doubt that a bridge can be built in California for under $2million, and big bridges cost in the billions.  Perhaps they need to scale their costs by 100.  Their site prep costs look like they are just pushing around dirt, not blasting out rocky cliffs and building piers and abutments—and even then they are way low.

In my playing around today I came up with a slightly different design from last year, probably because the width or depth is slightly different in this year’s problem:

A bridge that costs under $166,000 in West Point Bridge Designer 2012.

Although the bridge design program is fun to play with (as addicting as many other computer games), it doesn’t really teach much engineering. I did not learn the pricing model nor much about bridge construction by playing with it—just tried things randomly and scaled up beams that failed and scaled down beams that didn’t. (Well, there was a little more to it than that, but it was more gaming tricks than engineering.) I suspect that a completely stupid algorithm (like simulated annealing or genetic programming) could do a decent job of optimizing a design, particularly if you start with the templates provided.

Because my design is not very different from one of the provided templates, I don’t feel that I’m giving much away by showing this design.  (There are undoubtedly better designs out there, but I’ve spent as much time on this computer game as I can afford to waste.)

Related articles

• West Point Bridge Designer 2011 (gasstationwithoutpumps.wordpress.com)