New Scientist has an interview with Mark Loizeaux, the head of Controlled Demolition Incorporated, one of the biggest demolitions firms in the world. CDI has brought down all sorts of structures—there's a video available of them sending the execrable Seattle Kingdome to the oblivion it deserves—and the interview's worth a read. First off, blowing shit up is cool, but the interview also includes Loizeaux telling us why he thinks computers will never have a major place in the field:

… CAD is used for putting things together where you specify the steel, the concrete, you assume construction methods within parameters of building codes. You assume it was put in using health and safety-approved methods and inspections. It does not allow for weathering, structural fatigue, modification, all the things that don't show up on blueprints. … [With demolition] You move into a different category of structure that is distressed—failed yet standing structures that have failed as functioning structures because they break building codes or have been burnt, struck by lightning or tragically these days bombed or hit by planes. And it frightens me that would-be advancers of the demolition arts think that they can take a program—which is entirely contingent on the data put into it—to analyse what is going to happen in a structural system which is beyond definition. It can be bracketed, it cannot be defined.

Loizeaux has been in the business his whole life—his dad set up the company—and I wouldn't doubt that his innate understanding of the craft of bringing down a building far surpasses anything that could be done with software right now. On the other hand, I think most programmers would agree that this sort of statement—"Here is a rational problem that will not yield to the assistance of thinking machines"—sounds like a challenge. I know almost nothing about CAD or demolitions, but I can still imagine a few fruitful areas of technical development in the field: Do materials scientists need to broaden the ways they define and quantify structural fatigue? Do demolitions experts need better instruments for measuring weathering in the field, and quicker ways to assimilate the resulting data? Does CAD need to be modified and extended to allow this data into its models?

It's easy for me to imagine the role of computers growing in demolitions, but CDI probably won't get on that bandwagon early on. This growth is more likely to be driven by second-tier firms looking to reduce the amount of time required in analyzing a site, not to mention saving costs on payroll by not having to hire as many experts. Such firms probably won't get the high-profile jobs that CDI gets, and as they mess around with new techniques they may suffer the occasional injury or death. (New technology usually has casualties of some sort.) But practice makes perfect, and as time goes on the use of computers will most likely become more efficient, more trustworthy, and more widespread.

As with hundreds of other professions, the growth of technology may make demolitions cheaper and easier, displacing artisans in favor of those trained in the newest software. There will always be the occasional need for the demolitions genius, but the vast majority of the work will go to listless tool users, sipping sodas and pecking at keyboards. If Mark Loizeaux is lucky, it won't happen 'til he's retired.