I’ve been critical of patents that seem to have been issued for ideas that don’t represent much innovation (most recently, this one, and see more discussion here), extending existing “prior art” only a little, or not at all. Many patents, particularly for computer software, describe things that are neither novel nor non-obvious.

Contrast that with this product (pointed out on BoingBoing recently), which really seems to be a neat, original, useful, idea that is a significant step beyond the prior art.

Of course, you know that’s got to be patented. And, indeed, a quick search at the US PTO came up with U.S. patent number 7,107,634, “Method and apparatus for cleaning a conduit” (and there’s an follow-on application pending that adds the “self-cleaning” feature).

Now, it's true that this isn’t computer software, and part of my point has always been that physical inventions are easier to distinguish and to justify than software mechanisms are. When you can actually build it and look at it, you can readily show how it improves on what’s been made before... and it’s also easy to see if it’s substantially the same as something else.

Lots of people in the Boing Boing comments section are criticizing the design of this, and perhaps they’re right — perhaps this is a seemingly good idea that won’t work out in practice. But it sure looks interesting and useful... and it’s certainly an example of the sort of innovation the patent process was meant to foster.