Monday, November 29, 2010

This morning, our cat Carol hopped up onto the tile counter by the sink, purring and looking me in the eyes.

Clearly she formed an intent.
When she was sure that I was watching,
she proceeded to pantomime drinking water from the sink spout.

I turned on a tiny stream of cold water from the tap,
and she twisted her head to align her tongue with the stream and began lapping it up.

No words, yet effective communication.

Then I told her to get down and
wiped down the counter where she sat as she sought a drink of water.

Wednesday, November 17, 2010

The recent reports on the news about the Stuxnet computer virus that has Iran complaining is quite a technical masterpiece, and a far cry from the usual Trojans and botnet attacks that are the usual source of grief for us all. Indeed it seems that someone went to considerable effort to ensure that this virus didn't harm any ordinary industrial systems.

The virus has been discussed in this Ars Technica story: http://arstechnica.com/tech-policy/news/2010/11/clues-suggest-stuxnet-virus-was-built-for-subtle-nuclear-sabotage.ars with new details provided that inform this cometary.

First, the virus checks to see if it is in a very specific context, and targets systems that may be supporting a very particular type of control system that would be regulated in the U.S. as non-exportable as it would be controlling more than 33 variable speed drives capable of high frequency motor control. It checks to see if there are Iranian modules on the system, or one other type of module that could be employed in isotope enrichment plants. If it doesn't find these specific conditions, the virus doesn't affect the system that it infects (though it may try to spread further in an effort to land in a system that meets it's targeting criteria).

Since Uranium enrichment with centrifuges typically uses cascades of thousands of units, each making incremental progress at separating the rare atoms that would be useful in a reactor or weapon from the vast bulk of the material that is essentially inert, messing with the process can contaminate the "pure" product and compromise the effectiveness of the cascade.

The virus is designed to be hard to detect, and it doesn't try to destroy equipment. Indeed for the imputed purpose of this virus, it is more important to delay the production of weapons grade material than to cause a splash in the press.

From the information provided, it appears that the virus effectively delayed to some degree the Iranian military in it's quest to build a weapon. For this I am grateful, and appreciate the amazing operation that managed to get this computer virus into the system. It took knowledge of how the enrichment process works, knowledge of how the Iranian system is built, as well as substantial knowledge and skill in constructing the code that could implement this tactical objective.

Now that the existence of the virus is known, the resultant publicity serves to educate the rest of the system players to the vulnerabilities that we may have in various parts of our infrastructure. I hope that this publicity will spur industrial controls engineers to take a new look at security in our existing systems and to build in more robust security features when new systems are designed. Ideas as to how this may be done are the subject of another article (in press).

Thanks to whoever thought up this important operation, it achieves what a conventional kinetic operation could not and makes a concrete contribution to counter proliferation in a world that is dangerous enough without Iranian nuclear weapons.