Question for someone at FAA/Air Traffic Control

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

We have heard much about the FAA over the last 18 months. I believe they have announced "compliance" at least 3 times. At one point there was a quote circulated, purportedly from IBM, asserting that certain key computers in use by the FAA "could not be made y2k compliant". Now we have additional info via Cory's latest report that the FAA is not yet done. I have heard that certain replacement systems/computers were supposedly tested in a test environment but have not yet been installed.

Despite the uncertainty surrounding this issue, its extreme importance, the history of FAA 'mis-statements' and the fact that the claimed IV & V has never been made public, the issue seems to have simply disappeared after the last FAA pronouncement of compliance. This in itself seems quite odd.

Here's the question: Is there anyone here who knows, first hand, the actual status of FAA remediation efforts? Certainly the air traffic controllers working around the country would know if there has been a major system upgrade, if the actual computers have been replaced at their site, etc.

Any information will be greatly appreciated.

I am not particularly interested in third-party speculation (I can do that myself!)

-- Me (me@me.me), October 20, 1999

Answers

What, you don't believe their web site? They're 100% done, 100% installed, 100% tested, and 100% verified! Just ask Hoff, he'll tell ya...

And what does IBM know about Y2K, just because they built the computer?

[Polly mode off]

"The FAA said it was compliant June 30th, and what we sampled was all OK. But, then we found there were 967 systems changes later that were not checked, plus 200 in progress.

...

Another important issue is to do as much end-to-end testing as possible in the time remaining. The FAA, most of its 21 systems are not end-to-end tested yet.

...

Has FAA implemented all their upgrades?

Joel Willemsson: No. We testified last week that STARS would be implemented through December and January. We testified to our concerns on this. Not all new systems are in, by a long shot."

Tick... Tock... <:00=

-- Sysman (y2kboard@yahoo.com), October 20, 1999.


Doncha just know Hoff and Y2K Pro are gritting their teeth...

T-72 and counting.

-- lisa (lisa@work.now), October 20, 1999.


As I crudely understand it, the FAA is bogged down with an ancient air traffic control system--it works okay most of the time, though. That is, IT WILL WORK FINE UNTIL JAN 1 2000!!

The pilots and controllers like the old system. That want it remediated for y2k, not replaced. The FAA is trying to test and install a NEW traffic system called "STARS" by the time of the CDC, and they are literally RUSHING to get it up and going. The problem is that with STARS, aircraft routinely disappear, and the screens freeze up quite often. It is supposed to be better because it is "in color" and more user-friendly, as well as more complex.

STARS is _supposedly_ y2k-compliant, but not terribly reliable as of yet. So far, it can only handle a substantially lower volume of air traffic and still be compatible with public safety. That means that we can expect massive gridlock like never seen before next year.

That is my crude understanding. Anybody want to correct me, or do I have the gist of it?

-- coprolith (coprolith@rocketship.com), October 20, 1999.


I thought that STARS was the long-term solution, and would be implemented over the 1999-2003 period.

Yes, STARS was "re-designed" to be y2k compliant - apprently it wasn't originally designed and tested to be compliant, but 2000 compliancy was added after systems design was started. Raytheon has firmly stated that STARS is compliant, and has been tested to verify 2000 operability/compliance/whatever term they want. So we can assume that the new installations are going to be okay. But only a handful of sites nationally - I think about 15 now, but am not sure - are on-line and operating now with STARS.

The rest have to make do with the older system until all are replaced with STARS. And those older systems are the ones IBM could not maintain nor certify. (The FAA hired some retired IBM tech's; and is these tech's who have said the older systems will work. Not the company itself.)

If anybody has a line-by-line update of this version of this info - please jump in.

-- Robert A. Cook, PE (Marietta, GA) (cook.r@csaatl.com), October 20, 1999.


Link

Wednesday, October 20, 1999

Radical solution to airline delays gets more attention

Growing number of airline executives say privatizing the air-traffic-control system would improve efficiency.

By Stephen Humphries , Special to The Christian Science Monitor

Prompted by consumer complaints and costly delays, a growing number of airline officials are urging a reform that was once unthinkable: Hand over the US air-traffic-control system to a private company.

With a 43 percent rise in passenger numbers expected in the coming decade, there's widespread agreement that the nation's antiquated air-traffic-control (ATC) system must be updated. But the scope of the air-travel problem means that the radical proposal of putting the system into private hands is receiving more high-level attention than ever before.

If such a proposal succeeded, it would be the biggest reform in US commercial aviation since the deregulation of the 1980s - impacting everything from ticket prices to safety.

The notion, though, has long been opposed by major players in the aviation world, namely federal aviation administrators and airline pilots. Their opposition - grounded in the assertion that the current system needs to be updated, not discarded - is unlikely to change anytime soon. Yet the fact that some top airline officials are beginning to lobby for the transfer indicates that the debate over ATC's future will become more animated in coming months.

"This idea is no longer out on the fringe," says Robert Poole, a former aerospace engineer and president of the Reason Foundation, a think tank in Los Angeles. ATC privatization is "the critical next step to preserve the gains from airline deregulation by making the infrastructure flexible and market-driven."

Recently, the push for privatization has been building momentum. The Clinton administration's National Economic Council has been talking one-on-one to members of the House Aviation Committee about privatization models. And during the past few months, top executives at United, American, and Continental have made key speeches calling for the ATC to be turned into a private business.

The reasons to make significant reforms are clear. Current ATC has had an increasing number of runway near-misses, says John Mazor, spokesman of the Airline Pilots Association, and such safety concerns have led to more delays.

The question is: What's the best fix?

At present, US air-traffic control is maintained by the Federal Aviation Administration (FAA) and is funded mainly by passenger ticket-tax revenue.

Privatization proponents want to replace this with direct user-based charges paid by airlines. Delays cased by inefficiency cost the industry billions each year, Mr. Poole says, and any savings brought about by the change would be passed on to passengers.

That's been the experience in countries that have privatized ATC systems, including Canada and New Zealand, Poole adds. In all, 16 countries have gone private.

But critics say America doesn't need to go that far. Privatization is "unlikely, undesirable, and unnecessary," says Mr. Mazor. Like FAA officials, he says the current system can be saved, but only if Congress spends more money on modernization.

The key, he says, is to get Congress to spend some of the Aviation Trust Fund, a trust filled by airline ticket-tax revenue. To this point, Capitol Hill has been unwilling to do so, for fear that such expenditures might push the government even further beyond its tight spending caps.

Critics also say privatization could create a serious safety hazard. They point to Britain, where the Labour government came under fire for its continued support for selling off 51 percent of the nation's ATC, despite an Oct. 5 rail accident involving two privatized trains.

Poole and others respond that the FAA would retain much of its oversight on safety issues. Moreover, a privatized ATC would seek liability insurance from the private insurance market. This, Poole says, would ensure additional scrutiny of its safety by those underwriting the risk.

In addition, a private company could raise money much more quickly to modernize the system - making it safer and reducing delays. Using a satellite global positioning system (GPS) could help air-traffic controllers safely reduce the distances between planes, even under some adverse weather. That would allow more planes to fly in any given corridor.

Poole estimates that GPS would increase capacity at congested airports like Reagan National by 50 percent.

"People pay attention when they see improvements and lower costs coming together," Poole says, "which goes contrary to the idea that the solution is to throw lots more money at the problem."

-- Homer Beanfang (Bats@inbellfry.com), October 20, 1999.



Link

No quick fix for flight delays

October 15, 1999

BY LYNN SWEET SUN-TIMES WASHINGTON BUREAU

WASHINGTON--Frustrated passengers will find no quick fix to air traffic delays plaguing O'Hare and the nation's other major airports, and the airline industry predicts that delays may only get worse, a House panel was told Thursday.

The problem is so severe that air travelers can look forward with some certainty to "death, taxes and delays," said Rep. John Mica (R-Fla.).

Rep. William Lipinski (D-Ill.), the ranking member on the House Transportation and Infrastructure Committee's aviation subcommittee, recounted his horror story of sitting in a plane for seven hours in July on a Washington-Chicago flight that usually takes about an hour and 40 minutes.

At the hearing, the Federal Aviation Administration--which runs the nation's air traffic control system--air traffic controllers and pilots agreed that there is a crisis but disagreed over solutions and whom to blame.

The Air Transport Association issued a dire warning in a report presented to the subcommittee. The organization of air carriers forecast a 43 percent increase in passenger traffic by 2008--and a 250 percent increase in delays "if the air traffic control system is not fixed."

In August, the FAA announced a 21-step program to alleviate delays after a summer when delays dramatically increased. Thirteen of the proposals have been initiated. The FAA blames much of the problem on bad weather, equipment and air traffic control policies.

"We need to say at the outset that delays will never be eliminated," FAA Administrator Jane Garvey told House members. "Our challenge, our job at the FAA is to minimize these delays without compromising safety."

Garvey and others noted that there are many conditions that cause delays: bad weather, limited airport capacity, air traffic equipment problems and airline scheduling.

Randy Schwitz, executive vice president of the National Air Traffic Controllers Association, said controllers are under pressure to cram more planes into congested airspace because airlines want to schedule flights at popular times. For example, a recent Chicago Sun-Times series found that although O'Hare can handle about three takeoffs a minute at most, airlines regularly schedule as many as 20 takeoffs during certain peak times.

The Air Line Pilots Association, asking Congress to guarantee more money to deal with delay-related problems, said the main answer to delays is not the piecemeal FAA approach but a modernization of all air and ground systems.

The latest report from the Transportation Department, issued last week, shows an improvement in on-time arrivals and departures at O'Hare and Midway in August. At O'Hare, 74 percent of arrivals and 77.9 percent of departures were within 15 minutes of schedule that month, which is considered being on time. In July, 66.4 percent of arrivals and 70.1 percent of departures at O'Hare were on time.

At Midway in August, 82.6 percent of planes arrived on time and 78.2 percent left on schedule; in July, 73.8 percent of arrivals were on time, as were 69.2 percent of departures.

-- Homer Beanfang (Bats@inbellfry.com), October 20, 1999.


"A View from the Field" from NATCA is coming soon(will link here); still many concerns... A minor point: Ray Long, no longer is the lead of the FAA Y2K Program Office. He is now in another FAA line of business which is addressing National Airspace System delays. Draw your own conclusions, this disturbing situation is business as usual within the Agency: "tough to hit a moving target." Anyway, updated memo and link should be complete within the week...

-- Dan (Concernedabout@faa.gov), October 20, 1999.

"Gritting my teeth?"

Umm, sure lisa.

Look, I'm tired of posting the link to the FAA site, and the explanation that's been there for the past year. Y'all can look it up yourself.

The GAO is rightfully concerned that the FAA would be implementing new systems around the rollover. See all the info regarding the "freezes" going into effect. Doesn't mean STARS is required for Y2k.

-- Hoffmeister (hoff_meister@my-deja.com), October 20, 1999.


Isn't that what this is about, Hoff, the site says one thing, but the facts say another?

"967 systems changes later that were not checked" doesn't sound like 100% verified to me. "200 in progress" doest't sound like 100% done. "most of its 21 systems are not end-to-end tested yet" doesn't sound like 100% tested.

But the site says 100% everything. We can't argue about 99% of phase x doesn't really mean 99% any longer. They left themselves with 0% wiggle room now.

But the site isn't lying, right Hoff?

Tick... Tock... <:00=

-- Sysman (y2kboard@yahoo.com), October 20, 1999.


BS, Sysman.

You know as well as I do that nobody could fully freeze systems from June 30th on.

That was the whole purpose in the FAA hiring Primeon, to perform IV&V on changes that were required post-remediation.

The GAO is doing their job. You, as a systems person, understand what is required. That changes will be made, for any number of reasons. And that these changes again have to be re-verified for Y2k.

You know this, yet you still spout this stuff.

My question is Why?

-- Hoffmeister (hoff_meister@my-deja.com), October 20, 1999.



Hoffmeister, my question is, why are you such a government shill? Are they paying you? If they aren't, they should be.

-- Ohio Bob (ohiobob@buckeyestate.com), October 20, 1999.

Hoffy,

What does 100% mean then?

I guess the website that you have refered to so many times is meaningless then.

It has taken well over 15 years for the STARS project to be developed. I know because my uncle used to work on the project.

Yes, the problem is trying to maintain the old "look" of the controller screens. No, they are not even CLOSE to being done.

At least over rollover, STARS will work with the reduced number of flights.

-- nothere nothere (notherethere@hotmail.com), October 20, 1999.


As far as Y2k goes, the numbers on the website have nothing to do with STARS.

-- Hoffmeister (hoff_meister@my-deja.com), October 20, 1999.

Why Hoff? Because I don't believe that they are being honest. I think they are feeding John Q. a big BS line, while they're still working (maybe praying is a better word) like mad in the back rooms. I don't see this as normal maint work, and I don't think the GAO sees it that way either. But this is just my opinion Hoff, based on the same GAO reports, and other public comments that we've all been discussing for months. I know, your opinion is different.

You're pretty quiet for a talkative guy here Hoff. I didn't mention STARS. Why no comments on the other points, that do have to do with Y2K? What's the matter, GAO got your tongue? Or maybe you know what the real meaning of 100% is. <:)=

-- Sysman (y2kboard@yahoo.com), October 20, 1999.


See, Sysman, that's where you're wrong.

Like maybe, if you had done some actual research, you'd know all this stuff has already been published by the GAO. See:

http://www.gao.gov/ne w.items/ai99285t.pdf

Course, don't think it ever made it to the forum here, since it basically verifies that yes, the FAA did make their deadlines, and yes, the GAO did do some of its own verification.

And, if you had read the document, you'd see the GAO is specifically talking about modifications to certified-compliant systems, and the effect they have may have. You wouldn't be "guessing" based on a few clipped remarks in the WRP.

Now, the GAO is doing audits, and like any good auditor, they'll find things. Don't think the FAA is any different in this regard. Like I said, the GAO is doing its job. I'm glad they are.

-- Hoffmeister (hoff_meister@my-deja.com), October 20, 1999.



Ya see Hoff, that's where you're wrong. GAO in a bookmark on this 'puter.

"However FAA's work is not yet done... These challenges involve managing modifications to compliant systems, independent verification of systems' compliance, and systems testing"

Once again Hoff, you've only addressed one point, of the three here. Or maybe it depends on how you read that sentence. You read it as verifying and testing only modifications. I don't think that's what it says.

I guess it depends on what the meaning of 100% is. <:)=

-- Sysman (y2kboard@yahoo.com), October 20, 1999.


Folks,

Check out the claimed status of STARS at the following link. A summary status of STARS is as follows:

Renovation complete? N/A, new system

Validation Complete? 8/2/99*, * - Validation and implementation dates below are for EDC only. STARS Y2K progress keyed directly to STARS development progress.

Implementation Date? N/A

ferret

-- ferret (insider@agency.gov), October 20, 1999.


PS Hoff,

And if you wonder why I'm picking on the FAA, aside from their stupid public comments, like "we're 99% done" months before the June "deadline," alot of it has to do with that deadline. They said June 30, and made it 100% right down to the last chip on the last radar tower, on top of the last mountain. Down to the last line of "unfixable" mainframe code. Come hell or high water, they made the deadline, no questions asked, because they said so. 100% 100% 100% 100% on their Y2K status page. What John Q. Public watched on TV, and can look-up on the web. He doesn't pick apart the details like this forum. He reads the headlines. No matter that they were a big concern, of the GAO and others, only a few short months earlier. They did it, by June 30! 100% - see we told you so!

Must be a miracle, that few to date have been able to do.

Or bullshit.

<:)=

-- Sysman (y2kboard@yahoo.com), October 20, 1999.


I see, Sysman.

So you had these startling revelations all along, just sat on them until Cory posts some snippets. Then instead of pointing to the details within the document itself, you make guesses and just more references to the WRP quotes.

Sure, I buy that.

As for the FAA. The things that you quote why you "pick" on them are the very reasons they are so easy to defend. I have no "inside" information on the FAA, and really have no idea how they've handled past projects.

But awhile back, after reading the BS posted constantly, I took a look. And what did I find?

1) An org that defined their project, laid out a plan, and stuck to it. They didn't bow to pressure from the government to push to some arbitrary deadline; they laid it out, determined a timeline, and met it.

2) An org that from the outset wasn't worried just about "mission-critical" systems. Yet again, an org that determined what it would take to remediate all their systems, and again stuck to it.

3) An org that laid out the details of their plan and status, for the public, on the internet. Their project definitions, milestones, and status have always been there.

4) And finally, an org that from the start incorporated IV&V into their project timelines and workplan.

I don't know. Maybe it was because they had so many problems in the past, that they approached Y2k in this way. But from the outside looking in, these are all the signs of a realistic, well-run project.

As for your three "points":

1) ""967 systems changes later that were not checked" doesn't sound like 100% verified to me."

The Y2k effort was verified at the end of June. SAIC performed the verification, and the GAO independantly evaluated it.

The fact that systems were modified later doesn't change the verification. The GAO found insufficient documentation of Change Requests. But systems aren't static, which I would at least guess you understand.

2) ""200 in progress" doest't sound like 100% done"

Again, these are just more system changes. The GAO certainly isn't shy, and if in their review they found these were late Y2k changes, I'm sure it would have been stated.

Again, the very reason they signed Primeon was to provide IV&V of these ongoing modifications. As they had planned to do all along.

3) ""most of its 21 systems are not end-to-end tested yet" doesn't sound like 100% tested. "

The GAO doesn't consider the FAA tests comprehensive enough, and recommends more. Fine. That doesn't mean testing wasn't done, just that the GAO wants more.

Finally, for the other STARS comments. The GAO makes clear their problem is not with STARS needing to be implemented for Y2k, but that implementing it during December and January represents a waiver of the FAA's ARA moratorium on system changes.

-- Hoffmeister (hoff_meister@my-deja.com), October 20, 1999.


Hoff,

I see, spend the first paragraph, plus a few more comments, and attack my honesty. Gee Hoff, I guess the GAO site was never mentioned, or linked on this forum before. I guess I didn't qoute a few snips from an earlier GAO report (when was it, March - April?), with you. That must not be it, with the dozens of other links, under the Y2K folder, in my bookmarks. Do you think that I get all of my Y2K information here? Guess again buddy. How many dozens (hundreds?) of articles have I posted here, with links to other web sites? I get it now Hoff, you don't trust me, but you trust every freaking word on the FAA web site. You just spoke volumes to me, Hoff. If this is the best personal attack that you can come up with, well, it needs some work.

As for your remarks Sir Hoff, yea, this is some great plan. After you're "100% done" throw 967 changes in, retest, reverify, reinstall, spend some more money. But it doesn't have anything to do with Y2K. That's all done. These are all just "normal" fixes, FAA/SP-1 work. Right Hoff? After all, we do know what 100% in June means, don't we?

Oh well, you've got your opinion and I've got mine. When I read the GAO report I see Modifications, AND validation, AND testing. You some how see validation, AND testing OF modifications. Only one little word, that's what we're arguing about I guess.

That, and what the real meaning of 100% is.

Tick... Tock... <:00=

-- Sysman (y2kboard@yahoo.com), October 21, 1999.


Sorry to keep PSing you Hoff, but you said:

"The Y2K effort was verified at the end of June. SAIC performed the verification, and the GAO independantly evaluated it."

But the original quote, from Cory's post is:

"The FAA said it was compliant June 30th, and what we sampled was all OK. But, then we found there were 967 systems changes later that were not checked, plus 200 in progress."

Note the key words here, "then we found". What does this mean to you Hoff, that both the GAO and SAIC did a crap job and missed 967 changes in progress? Does "then we found" mean that these changes have nothing to do with Y2K? Isn't that what they are doing, a Y2K review???

I'll tell you what I think Hoff. I think the FAA tried to hide these 967 problems, and "showed" GAO what they wanted to see back in June, to get the 100% rate that they could brag about. I think GAO did do a good job, and dug a little deeper to find these problems. Did they dig deep enough, or are there another thousand down there somewhere?

Based on nothing more than just a gut feeling, and reading about this stuff for way too long...

I gotta crash Hoff. Happy Thursday.

And happy 100% <:)=

-- Sysman (y2kboard@yahoo.com), October 21, 1999.


Well, I'm going to be in NYC all day. No easy net access. Figured I'ld put this on top, in case Hoff checks in. Be back in about 20 hours or so.

Sorry for rant mode here folks. I'm just trying to figure what 100% means. I really didn't mean some of the things that I said above. I don't believe that FAA tried to hide 967 changes from the GAO. I figure that they tried about 500 or so, and hoped for the best.

Sorry, I've been on both sides of the trench. I know BS when I see it.

Tick... Tock... <:00=

-- Sysman (y2kboard@yahoo.com), October 22, 1999.


Darn, another PS:

Where's my buddy MVI when I need him... <:)=

-- Sysman (y2kboard@yahoo.com), October 22, 1999.


The FAA said it was compliant June 30th, and what we sampled was all OK. ...

Well, if "What we sampled" was okay, I certainly hope (for the sake of those flying) that what we sampled will be what has to operate "will be okay."

Seriously - I know that not all 'decision points" in a program can be checked - there are too many million to get all of them reviewed. (That what "adequate" testing is for, to wring out as many problems as possible before the public is foorced to use the software for its safety.)

Thus - what was sampled? How much? How were these problems missed in a system that people rely on explicitly for their lives? What is the software control proces at the FAA that lets these changes get through un-audited, and apparently un-reviewed?

-- Robert A. Cook, PE (Marietta, GA) (cook.r@csaatl.com), October 22, 1999.


Moderation questions? read the FAQ