Some are more critical than others . . .

Ignoring Application Criticality

by Conrad Weisert, January 8, 2016
©Information Disciplines, Inc.

NOTE: This document may be circulated or quoted from freely, as long as the copyright credit is included.

Levels of reliability and performance
for different kinds of computer application

  1. Application used by internal staff
       Users will be inconvenienced or annoyed by failures.

  2. Application that affects outside people
       Failure will embarass our company.

  3. Application that can cause hardship to outside people
       We may get sued.

  4. Super-critical application
       People may die

That's a presentation slide from a course in systems analysis. The point is that we have to strike a sensible balance between safety and cost.

Most devices fall somewhere between those extremes.

Not surprisingly the gun nuts are deliberately distoring the discussion in their attempts to thwart firearm-enabling control: A garage-door remote control is unreliable or may have weak batteries, they observe, so how could we ever trust an electronic device in a life-or-death confrontation with an armed intruder or a ferocious animal? It's impossible, to build sufficient reliability into such a device, so we should just give up, they advise. Recipients of heart pacemakers are surely grateful that such thinking didn't dominate medical research twenty years ago.

If the opponents are so concerned about the limits of electronic device reliability, we wonder why they aren't voicing equally strenuous opposition to automated vehicles, to aircraft landing controls, to power-grid controls, or to dozens of other critical innovations under development or already on the market?   Could it be because those same experts just like to play with their own lethal toys?


Acknowledgment:   I'm indebted to my colleague and friend, John Collins, for inspiring this article.—CW

Last modified January 8, 2016

Return to IDI home page
Education topics.