Well folks, for those of you who have followed the discussions about embedded systems and programmable logic controller failures, I have just hunted down some pretty good news. As those who have read my former posts know, before I went back to school and got a CS degree, I was in the quality control dept. for the largest coal mining company in the free world (you have to say that, the Soviet mines were all under one roof - so to speak). In that time, I was in and around the operating equipment for many coal mines, handled a lot of various high-tech equipment, and was in close contact (and paid physical visits to) coal fired generating plants. The most common PLC I saw in these plants by far was the Allen Bradley PLC-5 made by Rockwell International, and controlled by their 17xx series controller equipment. Today, I decided to bite the bullet, and hunt down the Y2K info on the Rockwell equipment. Believe me, I was nervous doing this, as if all the AB's failed we would come to a dead stop. Guess what!!! Almost all of them are Y2K compliant to the year 2099. (The ones that aren't are very old part no's, very very unlikely to be in service at any critical point). Most of the 17xx controllers are also compliant, though a few need a bios upgrade to become fully compliant. Moreover the ones that are not do not quit, unless the computer controlling them needs the date as something utterly critical in the controlling software! I wrote some software for such functions, and it would be a rare thing to get the date from the controlling unit, normally you just time out events by telling its clock you want to count seconds. The web site for the compliance issues is:

http://www.ragts.com/webstuff/y2k.nsf/Pages/Brands-Allen-Bradley?OpenDocument

Believe me, this really is good news!

-- Paul Davis (davisp1953@yahoo.com), October 05, 1998

Answers

Glad to hear it. Now if we could only get similar good news on those other 50 billion or so chips that are out there, I wouldn't have to lose any more sleep!!

-- Steve Hartsman (hartsman@ticon.net), October 06, 1998.

Paul:

Sorry you didn't continue to follow the "still antsy about nukes despite NERC" thread in the Utilities category. It could have saved you some anguish. As common as these things are, it's good to know they'll keep working. There is that leap year thing... But as long as they're not powered down, that won't be a problem.

-- Mike (gartner@execpc.com), October 06, 1998.

PAUL: We'll take all the good news we can get. Not that the fact this chip my continue to function 'fixes' Y2K but if the information being provided is accurate, it's one less piece of bad news.

I am just very concerned though about what may be convenient (or blissfully ignorant) claims of compliance from some vendors. Now before all the AB fans/cheerleaders get upset let me say that I am not accusing the good folks at AB of providing inaccurate or misleading information. Indeed, I'm willing, for the the sake of argument,to take them at their word.

But let me give you an example from where I work and I think you'll see the point I'm trying to make. Our IS staff was asked, with no training or direction, to provide a list of the Y2K issues within our individual areas. One staff member was quite enthusiastic. So she got right to work (she's also very diligent) and listed all the computers (i.e. the hardware) in her area. Then she went to the web site of each of the manufacturers and if the web site said "no known problems", she marked the device as "fully compliant". After working all day on this, she announced "Great news! Everything in my area is A-OK." She literally completed awareness, inventory, assessment, remediation (none needed) and testing (none needed I guess) in a single day. What a worker she is.

So her list of A-OK's were incorporated into a larger list of similar A-OK's and forwarded to management who saw that by and large everything was indeed A-OK. (an upgrade here, a new machine there). Management then proceeded to announce to our company (I'm paraphrasing of course) "Our fine IS people have given this issue a THOROUGH evaluation and we now know that our systems are basically in great shape and we anticipate no real problems associated with Y2K." The whole issue has now been pretty much forgotten as we've now moved on to 'more important work'. My protests were brushed aside as being too time consuming and unnecessary - we just don't have time to do the kind of assessment I was pushing for.

So, in order to keep my job and avoid burning bridges (I'm fairly low on the food chain) I did what any good employee and team player would do - I shut up and did as I was told. Besides I'm not part of the 'Official Y2K Task Force' (which by the way, consists entirely of IS folks) and my opinions do not carry all that much weight.

The other day I was talking to this person and mentioned that I was going to need to upgrade an operating system on one of the boxes in my area (similar to hers) in order to bring it into compliance. "You don't need to do that" she said. "I was at their web site and there's no problem with that box. It's fully compliant"

"The hardware is fine," I replied. "It's the operating system that needs updating."

"Oh" she said "I thought they (meaning the vendor) meant the whole system was fine. Are you sure? Hmmm. I wonder if I should have checked the applications on the box as well?"

I was speechless. I realized that the whole department had been thinking in terms of boxes and not in terms of 'systems'. I thought for a moment and said "Well, if you get some free time, it probably wouldn't hurt." She nodded and I left it at that.

I'd say we have just a bit of an awareness problem. Inventory and assessment could probably use a bit of work too, I'm thinking.

Embedded systems? Do we use those?

I have omitted names and details because I have no wish to harm my employer or their public image. I enjoy working there but I do not run the train. Nor would my continued 'nagging' - as it would be percieved, have any positive effect. Though if I persist, it might get me fired. You should have seen the look on their faces when I seriously suggested that a large generator might not be a bad insurance policy.

We employ roughly 1,200 people. What do you think those folks will be doing in 15 months? Shoot, if we can just keep this country limping along with electricity and phones, it will be an extraordinarily good day.

So yes, hopefully the good people at AB have done their homework and provided us with data which we can call 'good news'. We sure do need some. I'll try to keep my personal cynicism in check and if others have similar good news (such as the power generating facities which have announced that they are stockpiling coal) let's definitely hear about it.

Still paddling, was that a foghorn I just heard? Hey! Over here!

Arnie

-- Arnie Rimmer (arnie_rimmer@usa.net), October 06, 1998.

Embedded chips have always been considered "the wild card" of Y2K, and failure estimates have always been real low, as low as 1% (though also as high as 7%). The problem is, all it takes is one to bring down (or cause an undesirable event on) the component that it is used for. Hopefully, good news like this will not cause people to get complacent, because just a few good failures occuring simultaneously could be catastrophic.

-- Joe (shar@pei.com), October 06, 1998.

Arnie:

Your point is well taken. There is no substitute for testing. One difference here - you're talking about a system, not a single component. We trust manufacturers all the time when they make claims about their products. Many A-B PLCs are used in situations where a malfunction (not Y2K related) could cause injuries or death. Someone trusted the manufacturer there. If testing reveals a manufacturer is wrong, it should be made public as soon as possible.

This brings something else to mind - I've been using vendor verification for those items I have no real way to test. Now that I think about it, I have seen a couple of statements saying "We thought this thing did no date processing, but upon further investigation, we find it does." The point is, we should continue to monitor these vendor's web sites to make sure they don't slip in any 'late breaking news.'

I have to agree that awareness is just not high enough everywhere. (Or non-existant in some cases.)

I had a midrange computer sysop tell me that IBM will probably come up with a last minute fix for all their non-compliant stuff, because they just couldn't let all of those installed systems fail. It would probably just cost a bundle. I couldn't believe it!

I also had the following conversation with a Techie at a PC manufacturer.

Me: "Do you have a bios patch for model ####? When I set the date to 2000 and reboot, it comes up 2096."

Techie: "It's not the hardware that causes that, it's the software you're using."

Me: "No, I changed the date right in the BIOS setup program."

Techie: "Oh, then it's probably because you entered a two-digit year code."

Me: "No, I entered 'two, zero, zero, zero.' I rebooted, and the date came up as 'two, zero, nine, six.'

Techie: "Really? Well we don't have a fix for that. I would just go ahead and use the PC. If your software seems to work, I wouldn't worry about it."

Me: "No, I can't do that. I have databases and spreadsheets that need the current date to function properly."

Techie: "Oh."

Me: "Can you help me identify the manufacturer of the motherboard? Maybe they have a patch."

Techie: "Oh, sure."

THIS WAS A VERY SOBERING EXPERIENCE. I'm getting a bit scared right now, wondering how qualified all the 'qualifiers' are.

-- Mike (gartner@execpc.com), October 06, 1998.

I think I would trust Rockwell for two reasons. First is that their equipment, in my experience, has always done what they said it could do. Considering the number of manufacturers about whom I can't make that claim, it is a pretty powerful indication that they take themselves and their job seriously. Second is the fact that in this case they are the manufacturer - not a reseller for something overseas as so many PC makers are - and they will be the folks who land in the hot grease if they claim something is OK and it isn't. I am sure their legal dept. has made them very aware of this.

-- Paul Davis (davisp1953@yahoo.com), October 07, 1998.