Since the GPS timing synchronization signal is used in telecoms, utilities, and banking it seems to me of paramount importance that these industries know if there is a GPS risk so they can plan for it in their contingency planning. (Never mind that the ground systems software upgrade is not scheduled to start testing until September.)

For about a year I have been seeking confirmation that there actually has been an assessment of the embedded systems in the GPS satellites. I have been skeptical of the early claim that "there are no Y2k problems in satellites" because I did not understand how an assessment could have been performed that early.

I believe I have found the source of that information and I believe it indicates that no assessment was done. In a September 1997 online Air Force newspaper is the statement: "The satellites that form the orbiting GPS constellation are compliant to Y2K operational requirements.

Boeing North America, formerly Rockwell International, analyzed Block II and IIA satellites. Lockheed Martin Federal Systems designed the IIR satellites to be Y2K compliant. BNA is designing the Block IIF to be Y2K as well. The EOW rollover does not affect the satellites.

The only Y2K problems in the Space Segment are with a few sections of ground support equipment."

http://www.afca.scott.af.mil/pa/public/97sep/intercom.htm#8

A link is provided at the bottom of the article to the military's 1999 briefing on the Y2k status of GPS satellites, where this statement is quoted almost verbatim. (http://www.laafb.af.mil/SMC/CZ/homepage/y2000 Go to downloads, slide 13 of the PowerPoint presentation) The 1999 slideshow is the only "source" information I have found anywhere that talks about Y2k compliance of GPS satellites themselves. The problem is there is a big difference between "being designed to be Y2k compliant" in 1980-1990 and having made an assessment of the embedded chips in the satellites made in 1997. Many programmers did use four-digit years in their programming in 1980-1990. That has nothing to do with what embedded chips were used in the rest of the vehicle at that time to control the vehicles engine and other operations.

These things were designed before an embedded chip/embedded system problem was recognized. It seems unlikely that the designers designed every chip in the vehicle as well as the vehicle itself (and then didn't use these new "compliant" chips in the rest of their equipment). More likely, they used the same chips they were putting into other military equipment, which equipment it is now acknowledged needs to assessed. I suggest we quit "assuming" that there has been an assessment of embedded chips in the GPS satellites, and start pushing for some evidence. A potential problem with GPS is no military secret. The Government has already admitted the ground system is not compliant and will be late (end of the year), at best, in getting so. The consequences to our country of telecoms, utilities and banking failing to prepare contingency plans for the loss of the GPS synchronization signal would, in all likelihood, be far more complete than any military risk we might face. This is a sin qua non. ("Without it, nothing") If anyone has information demonstrating there has been or has not been an assessment of the embedded systems in the GPS satellites, please post it. If there has not been, then a fundamental element of contingency planning should be the reestablishment of land lines for synchronization of telecommunications, power, and banking.

-- Anonymous, May 23, 1999

Answers

One more piece of information: The Pentagon has stated that "loss of master timing distributed by satellite systems" is a common symptom of Y2k failures. (I beg your pardon? I don't recall anyone else saying this. Aren't they saying "GPS" is not a reliable vendor for utilities and telecoms?)

In doing a search on "Y2K Annex to the Federal Response Plan" on http://usgovsearch.northernlight.com/ I found the Joint Staff Year 2000 Operation Evaluation Guide, set forth in part below. Appendex C contains a Symptoms list of Common Y2k failures, and includes:

"Reduced throughput due to operations at lower rate communication links as a result of loss of master timing distributed by satellite systems."

(See: http://issb- www1.quantico.usmc.mil/year2000/frames/y2kopeva.html)

This raises two questions:

Why would there be a "loss of master timing distributed by satellite systems"?

What impact would this "loss of master timing" have upon telecoms, utilties, and banks, and are they sufficiently advised of it so as to be doing the proper contingency planning? None of the information I have been able to find suggests that they are aware of any possible loss of master timing, nor that they are preparing contingency plans for same. This is an issue that should be addressed, and planned for if necessary. We are getting down to the wire.

JOINT STAFF YEAR 2000 OPERATIONAL EVALUATION GUIDE

Prepared by: Joint Staff, Y2K Task Force, The Pentagon Room 1D825 Washington, DC APPENDIX C

SYMPTOMS LIST OF COMMON Y2K FAILURES

System date upon powering up is 1980 or some year in the distant past Date display (displayed at the "date" command) is different than expected

Unexpected expiration of passwords, licenses, security cards, etc Operating system file operations give unexpected results, e.g., files that are really young are treated as if they are old (the operating system asks if you wish to overwrite a "old file" that you know to be a "new" file)

Software development tools/systems (compilers) that overwrite "old" versions do not perform as expected, e.g., changes made to compiled programs suddenly disappear and problems that have been fixed suddenly reappear after recompilation of modules

Auto-archive or delete functions on files or email are launched immediately after a year 2000 date is encountered

Incorrect calculations involving time spans or timing intervals Applications that depend on file sorting by date for correct operation suddenly fail or present illogical results

Inability to "talk" to applications that are external to the local network

Reduced throughput due to operations at lower rate communication links as a result of loss of master timing distributed by satellite systems

Ability to maintain voice switches numbering plan lost, e.g., cannot add or remove users. This impact will increase with time if not corrected

Loss of ability to pass traffic in record traffic systems due to anti-orbital provisions

Cryptographic operations will be degraded to transmission in the clear; followed by key initialization and distribution problems resulting from expired key

Intel analysis could be impacted if time ordered images are utilized Loss of weapon systems targeting capability due to loss of radar operations

Routers encounter cyclic redundancy check failures

-- Anonymous, May 23, 1999

Snyder,

I think we've beat this particular topic to death in a couple of threads - try using the search function and see what you come up with. The bottom line is that there are different systems that can be used to conduct time based synchronization (WWV comes immediately to mind).

There are other obvious issues that need to be dealt with re: GPS in the electric industry (example: customer locator systems and pole locator systems) that are, on the whole, more convenience items than anything else.

I'm certainly not downplaying GPS issues in general. This is a very important global system on which many companies and industries have become reliant. However, my sense is that for the electric industry, GPS issues are pretty much a non-starter. There is a qualification to that statement: as long as the risk has been recognized and alternative means of land based locator services and timekeeping have been exercised in advance.

-- Anonymous, May 24, 1999

Snyder, I think it might help readers to understand the basis of your concerns if they have some background information, along with some pointers to why you are wondering if the GPS satellites themselves (as apart from the ground control and receiver systems) truly do not have any Year 2000 problems.

I'm going to attempt to do this, as I understand the situation, but please feel free to correct me if I have interpreted your position in any erroneous way.

Readers, there is a progression in the following information which will develop towards the main point, so bear with me!

First, there are levels of GPS satellite development. The first Block I satellites were used to develop the feasibility of the GPS system. There were 11 of these, dated from 1978 to 1985. The Block 2 spacecraft began the operational GPS system and the Block 2A satellites were a slight improvement over the Block 2.

According the the operational status of GPS satellites, there are now 9 Block 2 working satellites, launched from 1989 and 1990. There are 17 Block 2A satellites, launched from 1991 through 1997. And there is 1 Block 2R satellite now in orbit and operational from 1998.

The Block 2R satellites are "operational replenishment" satellites. (they replace, as necessary, the earlier ones) They have some upgraded functions from the earlier Block 2A's. "Additional differences between 2R and 2A are:

additional radiation hardening____ cross link ranging___ reprogrammable micro-processor___ two atomic clocks on at all times (hot backup)

The Block 2 and 2A's were built by Rockwell International. The later Block 2R's were from Lockheed Martin (formerly General Electric). There is even another generation of GPS satellites being developed, but those do not concern us.

The thing to keep in mind is that there are 9, 17, and 1 operational GPS satellites in orbit now and those numbers are in order of the "generations" up to the 2R. Thus you can see that the bulk of the GPS satellites were of the original first two operational series, built by Rockwell.

Now we come to Snyder's big question. The only statement found to date about the Y2K status of the space (satellites) segment of GPS is:

"The satellites that form the orbiting GPS constellation are compliant to Y2K operational requirements." (Sept. 1997 statement) This is apparently based on:

"Boeing North America, formerly Rockwell International, analyzed Block II and IIA satellites. Lockheed Martin Federal Systems designed the IIR satellites to be Y2K compliant."

Now, since there is only one (1) Block 2R satellite operational now, we are mostly concerned about the 26 earlier ones in orbit, launched from 1989 through 1997, and originally built by Rockwell International. The ones which, if we can extrapolate from the upgraded functions of the later version, did not have "reprogrammable micro-processor" capability.

In 1997, we have this statement that the Block 2 and 2A satellites have been analyzed by Boeing. On Rockwell International's web site we can find in the History that in 1996, "Rockwell sells its aerospace and defense businesses to Boeing." This is confirmed on the Boeing History site, "In December 1996, the merger of Boeing and Rockwell International Corporation's aerospace and defense units united the two companies."

Therefore, the timeline for Boeing "analyzing" the Y2K status of those earlier Rockwell satellites is between December, 1996 and September, 1997 (when the announcement about it was published) -- a time window of only around 8 months, and before many businesses had even begun looking at their own systems re Y2K.

If you've followed things so far, I think you can see why Snyder would like to have more definitive information about the status of these satellite systems. To have assessed systems another company designed and built, in the eight months immediately after a major merger acquisition, at a time when Y2K was barely heard of, does place a measure of doubt on the situation. Especially when there has not been any more data on this since 1997.

Is this the gist of your concerns, Snyder? Am I understanding correctly?

-- Anonymous, May 24, 1999

Hi Rick.

In response to, "The bottom line is that there are different systems that can be used to conduct time based synchronization (WWV comes immediately to mind)."

Wouldn't these mostly be telecom based? Did you happen to see my subsequent post about the Pentagon comment that a loss of satellite synchronization is an "expected" outcome of Y2k problems? No one has said that before. They have said there was no problem. Who has studied the consequences of turning off the inputs to thousands of devices expecting input, both telecom and others. Are telecoms essential to utilties?

I am certainly not suggesting there is no solution. I am only suggesting it would be much easier to put one in place (for telecoms and others) beforehand than after.

Best regards,

Snyder

-- Anonymous, May 24, 1999

Hi Bonnie -

Wow! Well, yes. Streamlined, throughly researched, and intelligently stated, that is the reason for my question. (If we weren't both married, I say I was in love! I have been gnawing on this for a year, and you have moved the question forward in a day.) I was unaware of the short timeframe involved. I had only noticed the early date of their "accomplishment". You always do wonderful work Bonnie. That is my question. That, and should telecoms and utilities figure satellite synchronization loss into contingency planning?

We rely on satellites. Anyone else want to help look into this?

Thanks Bonnie!

Snyder Gokey

-- Anonymous, May 24, 1999

Ouch. Excuse me - but - is this entire topic is very annoying?? WHAT is *REAL* about this? Can anyone -please- answer the following questions???

-WHERE ARE the Senate Hearings on these specific issues?? (Beyond the flimsy coverage from last year.)

-And info regarding the 911 dependency on telecommunications & the NRC's' law relevent to same?? (Aren't they supposed to shut down the nukes when local 911 doesn't work?!!? Has that rule/law been covered/changed yet!??!!??????)

-Where are the OFFICIAL DOCUMENTS that guarantee that ALL the GPS receivers that impact on big business survival.... won't malfunction??!!!

-And why has the GPS (military site) only claimed that Military receivers will be ok?? What about everything else??? (hello??)

- What will happen when the rollover in Aug happens? Can anyone describe realistic scenarios??

-WHICH telecommunications are compliant or not?? Via GPS? And WHAT BUSINESSES are dependant upon THOSE that are NOT compliant? - Is there a government &/or business site that thoroughly reviews this (potentially *very* vulnerable) situation???

[[Snyder's questions are ~very~ intelligently relevant, to my mind... And Bonnie and Rick's replies do not soothe my worries at all, indeed this issue has annoyed me for a long time, and I still cannot satisfy myself that the powers-that-be have it ALL under control... no matter how much I research...Can anyone help us??]]

thanks for any thoughtful replies/comments/leads...

God Bless

-- Anonymous, June 04, 1999

I've had so much going on I almost forgot to post this. I received e-mail from someone who identified himself as a Rockwell engineer who retired at the beginning of this decade after five years "as a project engineer on GPS but not directly involved in the clocks or electronics design."

He did not want to post or be identified, even though his information presented an optimistic scenario. He did give me permission to post his views which I summarize here:

[I think]"When Boeing took over Rockwell, the Rockwell people working on GPS just took off their Rockwell badges and put on Boeing's. I feel certain that these Rockwell/Boeing engineers made that compliance assessment and not the Boeing/Boeing engineers. There was not sufficient time nor any technical/financial reason for them to replace the existing Rockwell team."

In a reply I said that such a scenario seemed quite possible, but there were a couple of "ifs" I could think of. One involved the corporate downsizing and streamlining which was still ongoing in the merger timeframe, so that it might be an unknown re Rockwell people taking early retirement packages, etc. The other was that the design and development of the satellites was ongoing for several years before the first was deployed and we couldn't know how many original designers (if any) were still at Rockwell by the time of the merger. And would vendor specifications be difficult to obtain for components manufactured in the early to mid 80's?

This is the reply addressing those questions: "Many of the design engineers of Block I vehicles were in Engineering management in 1990. The merger and the eight-month concept are irrelevant. Vendor specifications are the key documents in any subcontract and are kept in locked cabinets in long term company storage facilities. They could be available for review in a few days."

I can neither verify nor discount this information since I do not know the sender of the e-mail, but in the interests of having as much input as possible on the situation, I believed it should be offered to readers.

-- Anonymous, June 05, 1999

This statement seems possible...

"Vendor specifications are the key documents in any subcontract and are kept in locked cabinets in long term company storage facilities. They could be available for review in a few days."

However, how many entities specified "Year 2000 compliant embedded systems" in 1990? No one knew what they were.

Clearly the specs on the system in general did not specify that be Y2k compliant, insofar as it is admitted the ground software is not, with a billion dollar fix only scheduled to be delivered for testing in September, according to the military. Therefore, I do not see how a review of the specifications could answer the questions that need to be answered.

Thanks for the continuing information and thoughts.

Snyder Gokey

-- Anonymous, June 06, 1999