IP: more on DO READ -- On centralized, distributed and absolutism -- and the thermodynamics of terrorism

[David Farber <dave () farber net>]

> From: Mary Shaw <mary.shaw () cs cmu edu>
> To: farber () cis upenn edu
>
>
> Dave,
>
> When Jim Morris first described this idea to me a couple of weeks ago, I got
> something very different out of it -- namely a risk analysis heuristic
> something like,
>
> "When analyzing public safety risks, pay particular attention to
> concentrations of releasable energy. Be especially vigilant about
> opportunities to bring multiple concentrations together."
>
> Examples of such concentrations include very tall buildings (potential
> energy), large airplanes full of fuel (chemical energy with some kinetic
> energy thrown in), high dams (potential energy), nuclear power plants.
> snowpack of slopes with avalanche potential, earthquake zones and especially
> high-energy structures in earthquake zones.
>
> All the examples, of course, have other benefits and shortcomings.  The
> energy heuristic doesn't speak to concentrations of humans at risk -- a tall
> building full of people is a different proposition from a tall empty
> building. We have seen a people-centered heuristic operating in the last
> month as large gatherings were cancelled or protected.  This heuristic might
> go,
>
> "When analyzing public safety risks, pay particular attention to
> concentrations of people. Be especially vigilant about situations in which
> the people cannot effectively disperse"
>
> This leads you to ask about occupied buildings that may take considerable
> time to evacuate and to airplanes, where passengers and crew are captive
> while in flight.
>
> It's easy to add heuristics to the list . . .
>
> "mechanisms for rapid propagation of threats" -- The US Postal Service is
> highly distributed, but we are seeing how movement of envelopes can
> distribute spores. We were already paying attention to public utilities. I
> wonder whether the smallpox war game considered shutting down all intercity
> transportation to contain the contagion?  My Dutch friends have told me that
> they handled hoof/foot in mouth disease effectively with "ring vaccination"
>
> "opportunities for destruction of public confidence" -- I agree with much
> recent criticism in IP of the actual effectiveness of many airline security
> measures.  However, to the extent that these measures actually increase
> public confidence, they serve a purpose that we as technologists may not
> fully appreciate.  Also, the recent anthrax attacks are taking toll on
> public confidence that's out of proportion to the actual damage done (at
> least thus far).
>
> . . . and so on.
>
> Avoidance isn't the only possible response, either.  Protection, robustness,
> and mitigation are often good options, too.
>
> I assume that considerations such as these guide the actual practice of
> public safety, but I don't know how explicitly they appear or how
> systematically they're evaluated.  We did miss on flying airplanes into the
> WTC and on anthrax contamination of USPS facilities.
>
> For me, an intrinsic characteristic of engineering is the requirement to
> strike appropriate balances among conflicting goals.  So pursuing one
> heuristic to the exclusion of all others is only very rarely, if ever,
> appropriate.
>
> Engineering also requires finding cost-effective solutions from among many
> and diverse alternatives. Heuristics help to organize the solution search to
> concentrate on good candidate solutions. They remind you to consider
> possibilities that you might otherwise ignore.
>
> Heuristics, like many engineering tools, can help a good (or even adequate)
> engineer find better solutions more effectively. They are not a substitute
> for actually understanding the problem and making sound judgments about
> solutions.
>
> Mary


For archives see:
http://www.interesting-people.org/archives/interesting-people/