Embedded systems will not fall prey to Y2K bug ~15 October 1998greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread
I have been asked by some for links to "embeddeds".
15 October 1998
Gartner Symposium: Embedded systems will not fall prey to Y2K bug
by Jonathan Lambeth in Orlando, Florida
The risk of embedded systems crashing because of Y2K is based on ill informed and over hyped information, analyst company Gartner Group warned this week, encouraging users into unnecessary remediation work.
New research by Gartner shows that many embedded systems need not be fixed they will not have a problem. Large scale systems are at serious risk, but your fridge, fax and other everyday devices are safe.
Gartner research director and leading Y2K expert Andy Kyte said companies should focus time on embedded systems that are safety critical.
Embedded systems cannot pose Year 2000 risks unless they have access to a source of persistent date information, he said.
Kyte relates discussions with users who have replaced all their faxes because they were told they were non compliant. The sum total of non compliance Kyte noted dryly was the clock that prints the time of sending.
If any of the audience can find me a fax that wont transmit because of the tick over from December 31st to January 1st 2000, then I will eat it, he promised.
There are an estimated 50 billion embedded chips in the world. The research breaks down embedded systems into three categories.
Microcontrollers are small devices, common in almost all electronics, that have their instructions burnt in at time of manufacture and cannot be reprogrammed like software. They may have some spare addressable memory that can have a date field added but a mere one in 10,000 of these will fail and these will be impossible to identify.
Most of these devices dont even know what planet they are on, let alone the date, said Kyte.
The second category is microprocessors, programmable devices that execute code. These can be at risk if they have a real time clock mounted or communicate with a clock. Less than one quarter of a per cent without clocks are at risk, but around seven per cent with some time dependency will have problems as the clock ticks over into the next century. Only two per cent will continue to have problems once reset.
Finally, large scale embedded systems are most at risk, with more than 35 per cent expected to be non compliant. Typically found in manufacturing, oil and healthcare environments, but can even encompass things like traffic light controllers or aircraft systems. They typically include common PC components although often run proprietary or even site specific applications.
On the embedded systems side, a capital expenditure, Kyte advises users of one get around.
Talk to your chief financial officer. If it costs $1,000 to replace a device, how much should be spent on investigation? Replacement may be more tax efficient, he said.
James Duggan, research director at Gartner said he believed the three areas likely to be worst affected by problems in embedded systems are the oil industry, telecommunications and power grids. He said things such as aircraft were well used to protecting against a single point of failure and were in good shape, though that did not apply equally to the systems that maintained aircraft fleets.
There is one final alternative unless the embedded system does predictive or retrospective calculations, simply leaving a device or system turned off through the tick over into 2000 may well mean it carries on regardless, blissfully unaware of the starting of a new age.
To comment on this story email firstname.lastname@example.org
-- Cherri (email@example.com), September 15, 1999
Of course, this came out before the Van Nuys, California 4 million gallon sewage spill that occured during a Y2K test this year. For that matter, long before the Navy report that Jim Lord exposed, which listed all kinds of potential Y2K failures in such systems.
I think this might be what pollies call "old data", Cherri....
-- King of Spain (firstname.lastname@example.org), September 15, 1999.
This story is over 1 year old Cherri, do you have anything that is recent? A lot has happened since October 1998.
-- current (email@example.com), September 15, 1999.
"your fridge, fax and other everyday devices are safe."
I feel just so much better now. How are the "industrial strength" systems doing? The ones at power plants, refineries, and manufacturers .
"Finally, large scale embedded systems are most at risk, with more than 35 per cent expected to be non compliant."
Yea, we know.
Tick... Tock... <:00=
-- Sysman (firstname.lastname@example.org), September 15, 1999.
Don cha jist haite it wen dose litle freetos break offin an git imbeded in da beahn dip. Dat leafs uh sirprise en dah beahn dip. Buht etleest it aynt sheip dip. Dats hard.
-- no talking please (email@example.com), September 15, 1999.
Whew. No, wait a minute. I wasn't worried about my fridge or my fax. I'm worried about the chemical plant upwind of my family members, about the oil refineries, the safety equipment, medical equipment, etc...
-- Mad Monk (firstname.lastname@example.org), September 15, 1999.
A lot of money can ride on whether or not a fax machine displays the correct date on a fax. Maybe he would like to 'eat' the litigation costs for a company that ends up in court fighting over the date a fax was sent!
-- Dian (email@example.com), September 15, 1999.
KOS, Your using the Navy report, which obtained Y2K readiness information from the utilities themselves (or assumed something if no replys were received), as "evidence" of embedded system problems? So basically, if the utilities provide information to NERC, you dismiss it?. But if they supply it to the Navy and you get it after it's been misrepresented by Jim Lord and it's "bad news", your "evidence" threshold has been met?
I think I understand why you are always looking for someone to mud wrestle, lol...
-- FactFinder (FactFinder@bzn.com), September 15, 1999.
I have noticed that you have been posting interesting stuff lately but this is old and an article.
Here are a couple of great links in regards to Embedded systems.
The Millennium Problem in Embedded Systems - Casebook
5.0 Embedded-System Failure Rates Gartner Group Aug 16 1999
-- Brian (firstname.lastname@example.org), September 15, 1999.
The Navy report exposed by Lord showed many cases of embedded systems failures possible, with obvious serious consequences. Regardless of what assumptions were made, whether NERC was or was not involved, etc., that is a fact. AND THIS NEWS OCCURED LAST MONTH.
The point that I was making, and others who have posted to this thread are making, is that Cherri has posted OLD NEWS (something pollies are always so quick to point out when its a pessimistic article). There is a lot of NEW NEWS, and it hardly is what one would consider happy.
-- King of Spain (email@example.com), September 15, 1999.
Cherri knows it's old, she put the date in the headline.
What's up Cherri, what is your point here? <:)=
-- Sysman (firstname.lastname@example.org), September 15, 1999.
There's one sentence in particular in the article Cherri posted that sums up the whole situation:
Large scale systems are at serious risk, but your fridge, fax and other everyday devices are safe.
-- Linkmeister (email@example.com), September 15, 1999.
A chip's date reliance can be detected in several ways, says Dave Walton, director of AMI Consultancy. "There may be a display of the current date, or the system may generate dated printouts. It may carry out functions on specific times or dates, or it may store information for averaging purposes," he says.
Walton was recently involved with checking a manufacturing plant that had 2,700 devices spread over 30 production lines. "We found 17 non- compliant devices affecting 12 production lines. So only 0.6% of the overall number of devices could have affected 40% of the production lines. We didn't think that they would all fail catastrophically but some could easily shut part of the line down, or even cause damage where, say, a robot arm could lose co-ordination with a hydraulic press."
From a March 11 article in Computerweekly.
The Gartner Group is the most unreliable source of information I've seen. They have put out information that is all over the map...from doom and gloom to no big deal. They seem to be the ones that are "ill informed" in my opinion.
-- Don Wegner (firstname.lastname@example.org), September 16, 1999.
Then in Feb 99, we learned this...
The boy who solved the riddle of Rubik's Cube is now a man and the developer of software that detects where the millennium bug will strike.
At the age of 12, Patrick Bossert shot to fame when he worked out his own solution to the mystifying cube and wrote a book about it that sold 1.5 million copies.
Now, he and a team of software experts at London-based WSP Business Technology have developed Delta-T Probe, a program that electronically scans hardware to identify microchips embedded that process date and time. It can determine which electronic equipment is likely to malfunction when 1999 becomes 2000.
"Only a small percentage of systems will fail to recognize the next millennium, but finding out which ones might go wrong is a huge and costly process," said Bossert, technical director at WSP Business Technology. One in 500 embedded systems will take confuse 1 January 2000 with 1 January 1900 and cause equipment to fail, he said.
For ongoing Year 2000 coverage, visit Wired News' special section.
By Bossert's estimate, there are hundreds of millions of chips buried deep inside equipment that control everything from security systems and fire alarms to production lines, medical equipment, and telecommunications.
Sainsbury's, a British supermarket chain, tested Delta-T at one of its stores in Devon, located in southwestern England.
"We have been working on the millennium bug since 1995, and thanks to Delta-T Probe, we have confirmed that the work we have done on equipment with embedded chips has been spot on," said Trefor Hales, Sainsbury's director of retail systems and innovations.
Seems the estimates of the embedded chips built, sold, and buried, has risen a few billion since this article was published. Now Cheri, do the math.
-- Michael (email@example.com), September 16, 1999.
"Large scale systems are at serious risk, but your fridge, fax and other everyday devices are safe."
I'm just curious, where in the hell is the embedded chip in my refrigerator?
-- Cary Mc from Tx (Caretha@compuserve.com), September 16, 1999.
Have you even read this Cherri?
Did you recycle this because it had a HEADLINE that you agree with and is misleading?
Get a fricking LIFE! No one here gives a damn about whether or not their FAX machine will say 1900 or 2000!
"James Duggan, research director at Gartner said he believed the three areas likely to be worst affected by problems in embedded systems are the oil industry, telecommunications and power grids. He said things such as aircraft were well used to protecting against a single point of failure and were in good shape, though that did not apply equally to the systems that maintained aircraft fleets."
-- nothere nothere (firstname.lastname@example.org), September 16, 1999.
Sysman is right, I am setting up a starting point of articles and information. As I dig them out I will post them and discuss the replies I get to them. It is my hope that those reading them will compare the information in them and the dates they were written to draw their own conclusions.
-- Cherri (email@example.com), September 16, 1999.
Isn't the embedded chip in your fridge the one that even a chisel won't pry loose?
Guess Cherri didn't get it, that the failure of the large scale systems, e.g. power, will cause the failure of your fridge, huh?
no wonder the Polly's don't get it.
-- Beached Whale (firstname.lastname@example.org), September 17, 1999.