Your Toaster Doesnt Know What Year It Is

The original author of this page holds that a massive collapse of society due to Y2K-related failures in embedded computer chips is highly unlikely. In short, MyRefrigeratorKnowsWhatYearItIs.

Your toaster doesn't know what year it is. Your alarm clock doesn't either. Neither does your tv or your car. Your pacemaker doesn't know what year it is. The elevators in your building don't know what year it is either, in most of the cases. The rest of them will think they are 99 years ahead on maintenance, not behind.

Based on my experience in the controls industry, I believe that the majority of control systems don't know or care what year it is. They care about events, not absolute time.

The author's interlocutor, who also has some experience with controls, counters that we can't be so cavalier because we don't know how big the problem may be:

The toaster and PaceMaker? sure seem unlikely to care, but many distributed embedded control systems (traffic lights, power generation & distribution, train routing, telecoms, manufacturing lines, oil & gas pipelines, water & sewage, ...) have date calculations in 'em to help manage load variations, product-data tracking, scheduling, preventive maintenance, instrument calibration, and environmental control. Authoritative estimates of how many control points have such dependencies seem to vary from 2% to 20%. AllenBradley?, one of the largest control manufacturers, gives a figure of 10% for its products.

The issue hinges on the steps each control system takes if it strikes one of these dependencies on 1/1/00. If it gives an alarm and expects operator intervention, that's a problem. If this results in out-of-sync, race conditions, manufacturing-process disruption and ripple effects this is a bigger problem. In general, the more interdependent network elements are with others, the bigger the risk. For example, leap-year miscalculations have shut down large manufacturing/refining operations in recent history: http://www.industryweek.com/cvrstory/html/dco010598.html

What makes this particularly sticky is that these systems are firmware, not software. You can't patch 'em - you have to replace 'em. The US Navy equates the size of this task with replacing the whole US interstate road system.

The issue of the GPS network failing was also raised. (This is scheduled for August 22, 1999, for some outdated receivers.) The concern apparently is that oil and other important shipments will not be able to get to port, resulting in serious economic effects.

Other concerns were raised regarding prisons and bank vaults opening their doors at the wrong time or day, stop lights fouling up traffic because they think it is Tuesday, and similar problems which might arise.

GPS-related failures might also have other economic effects: (quote from http://catless.ncl.ac.uk/Risks/18.96.html#subj9)

''"GPS is now the primary means of distributing time standards throughout the US, and throughout much of the world. (The accuracy of the atomic clocks on board the GPS satellites is second only to those maintained by the primary standards clocks in Washington.) Thousands of large financial computers ultimately take their time calibration from GPS, every day. Interest on overnight multi-billion-dollar short-term electronic-funds transactions is computed at millisecond granularity, derived from the GPS standard."''

The subsequent discussion, um, diverged. Below are some interesting Y2K-related links culled from the discussion.

On the embedded controls issue: http://www.accsyst.com/writers/embedded.htm http://www.bug2000.co.uk/business/emsys.shtml http://www.garynorth.com/y2k/results_.cfm/Noncompliant_Chips

Controls interconnections promoting trouble: http://www.businesstoday.com/techpages/y2kindex03.htm .

Ships at sea: http://www.sjmercury.com/business/tech/docs/058935.htm

RISKS forum, GPS and interest calculations: http://catless.ncl.ac.uk/Risks/18.96.html#subj9 .

On the other hand, see YearTwokBusinessOpportunity and YtwokSecondaryEffects

CategoryYtwok