Session II Panel Discussion

VICE CHAIRMAN FRANCIS: I think this is a reading test. Let me go through the questions first, and people can answer them, and that will presumably generate a little more interest here as well.

This is for Mike Holtom. How is the data transmitted from the aircraft to the ground station, inverse or constantly in real time?

CAPTAIN HOLTOM: Most of the data is not transmitted. Those five gigabytes a day would cause quite considerable transmission costs.

You can program the more advanced ACMS, aircraft condition monitoring systems, on modern aircraft, like the Boeing 747-400 and the 777, to transmit automatically in the event of certain events happening. Might be an engine failure or an over-speed or an over-heat or even something as small as a go-around, which we do just for those aircraft.

In the event of an engine being close to, shall we say, it's useful life, towards the end of its useful life, if we have any concerns over an engine, we also program that particular engine to send its data automatically after every particular take-off, and we have recalled four aircraft on four occasions back to the take-off point, the take-off place, because we felt that that aircraft with that engine that was on board the aircraft might not make another take-off.

So, rather than have an engine fail down route -- and the one particular that's in my mind is out of Heathrow to Hong Kong. Rather than not be sure that engine would survive another take-off out of Hong Kong, we brought that airplane straight back into London for an engine change, and that saved us probably considerable amount of money.

But most of the data comes off overnight or in transitting our main base at Heathrow or any of the main base stations.

VICE CHAIRMAN FRANCIS: Okay. This is for Max Fuller. Is black box recording technology enough of a "exact science" to justifiably termination a truck driver's employment but not exact enough to be used as evidence in a court of law?

MR. FULLER: Currently, I'm not an expert on black box recorders. I'm probably not the right guy to answer that. Part of our goal is to work with people like the Eaton-Vorad people to have them to continue to develop their product to where we can get it to the point to where hopefully we can profile a driver that does consistently have what we consider maybe bad driving habits and get him out of the cab of the truck.

As far as determining whether the black box recorder is to that level or not, I'm not sure.

VICE CHAIRMAN FRANCIS: Okay. This is for Mr. Westhoff. Are condition alarms conditioned at the RTU/PLC field site or at the control point? Maybe we could explain a little the acronyms here.

MR. WESTHOFF: The conditional alarms that we use are implemented at the central location site. There are some at the field area locations, but the conditions are not generally set at the RTU level. We like to maintain that data in its raw form, and we don't do any filtering from the RTU on in.

VICE CHAIRMAN FRANCIS: This is for Steven. How do you deal with bridge audio where someone comments in negative about ship captain or company when you can replay on ship and captain has immediate access? Do you give anonymity?

MR. SPENDLOVE: A tough one. Well, on a ship, whispers or hearsay are always going to happen, and there's just no way around that. I mean the purpose of the voyage event recorder/data recorder is to record all conversations.

It actually takes quite a professional knowledge to actually go in to the actual recording device and to select the time and the speech that you want to do, and then to actually decode that and listen to it. So, it's more a time-consuming effort, and I don't really think that is really relevant.

VICE CHAIRMAN FRANCIS: So that you're reasonably comfortable with the level of anonymity just in terms of the technical requirements?

MR. SPENDLOVE: Oh, yeah. People say nasty things about everybody, and people that work for me are always calling my name in vain. So, it's just something we have to live with, and I think that a captain of a vessel is professional enough not to let it get to him if he hears, you know, words.

VICE CHAIRMAN FRANCIS: This is for Jeff. Who guarantees anonymity, anonymity from who or whom, FAA, company, lawyers?

MR. BAYLESS: The way the program is set up, the FAA guarantees immunity, and the anonymity is set up in the software, so that even if I wanted to try to identify a crew, I couldn't. The only people who can are the excedance guidance team members, and, so, it's under their purview to make sure that they don't disclose who the crew is.

I don't pick those people, but ALPA picks those people very carefully, so that they don't abuse the huge power that they have.

VICE CHAIRMAN FRANCIS: This is for anonymous, but seeing as Mike and I were talking about this this morning, and BA is getting into helicopter recording and FOQA programs, the question is, compare and contrast FOQA with HUMS or other current flight data recording programs.

CAPTAIN HOLTOM: Thanks, Bob. HUMS, as you probably know, is Health Usage Monitoring Systems, which was originally designed for helicopters to handle vibration and high temperatures in gear boxes and chip detection in the gear boxes of helicopters.

There is a recent development, promoted by the oil companies, and in particular Shell, through the U.K. CAA, and I believe it involves a company called Stuart Hughes, who are experts in HUMS, to expand the use of HUMS to include flight operations monitoring for helicopters because oil companies get a bit fed up with their executives being dumped in the sea.

The sort of money you're saving on helicopters is far different from what you'd be saving in operations on a 747 or a 777. So, we felt we needed to bring this, and this is probably relevant to many of you here, bring down the costs of monitoring data to make it a simple practical option which anybody could apply, and helicopters, as you know, are just a few million dollars compared with large aircraft.

The project has been running now for about nine months, and we're due to deliver all the software and get the aircraft fitted out by June. The U.K. military is very keen on what the civil helicopters are doing because I think they very soon will adopt something similar.

VICE CHAIRMAN FRANCIS: There are two questions here for Gary, which I think perhaps have been partially answered, but I know that given the nature of the questions, he'll be delighted to deal with them again.

What is the status of the FOQA data protection immunity NPRM, and where is the FOQA AC? Isn't it only in draft form? It hasn't been published, has it?

MR. DAVIS: That's correct. The FOQA AC is in draft form. However, we're not going to go into final form until we get ourselves a rule, we don't want to make it a final AC when we don't really have the final rule out there yet. So, that's why it will stay in draft, but we think it is suitable guidance for you to use, and certain numbers of air carriers are using it. So, it is working.

The status of the rule. The best I can tell you is we are in rulemaking. There is some disagreement among the different parties, and it primarily deals with statutory authority as compared to what we would like to see immunized as a part of this program, and when we have a disagreement on what we can immunize and what we can't immunize, then you end up having a problem. So, that's what has held it up.

We do have an administrator's policy statement that she feels is acceptable and is allowing us to start some additional FOQA programs. The FAA is in support of these programs, and we hope that the airlines do voluntarily participate in them.

VICE CHAIRMAN FRANCIS: Gary, while you're on a roll here, there's a question here which says, when will FOQA include regional commuter operations, who likely have lesser experienced pilots? Is there not a greater need there?

MR. DAVIS: Well, FOQA may already include some of the newer commuter operations. I don't have the names of the airlines that have said that they are going to join us, but it's a voluntary program. We're not mandating FOQA for anyone. We want the airlines, we want the pilots, we want everybody to buy into this and want to do it. We feel like it helps the public in general, and if the airlines voluntarily buy into it and accept it, and their pilots buy into it and accept it, the FAA will certainly support it.

So, any commuter airline that wants to start can start, and I think some of them have already asked our FOQA office who is the AQP office out at Dulles for some guidance, and Dr. Tom Longridge is talking to some of those airlines.

VICE CHAIRMAN FRANCIS: Okay. A couple here for Max Fuller that are kind of related.

The first is, what actions do you believe the U.S. DOT should take to advance commercial truck recorder installation and use, and the second is, in your opinion, why is the American trucking industry so reluctant to embrace the on-board computers for hours-of-service?

MR. FULLER: I think there's probably two real issues here. Probably the first one is that this industry really doesn't understand on-board computers, Number 1, and to a certain extent, there has been a move by some of the regulators to almost create a dual standard, where you regulate the guys that have the technology much tighter than you regulate the guys that don't, and because of that, people in this industry have been relatively slow to respond to this type of technology.

At U.S. Xpress, we adopted it. We are having to develop the technology because we think it's the right thing to do. I can tell you I've got some competitors, though, that are very concerned, and I don't think it's that we want to keep the hours-of-service off of data recorders and keep the government from having access to it.

What we do want, though, is a common standard that the whole industry's being regulated by. I think that's more the issue, but truthfully, education of what can be done with these recorders is probably paramount in my industry, and part of the reason is people don't really understand what they're all about.

They think that they are nothing more than a regulatory device, and if they're applied properly, then we can increase our productivity of equipment and probably therefore reduce overall operating costs, and that's truthfully one reason that my company adopted the product.

We think long-term that can give us that competitive edge in being much safer, but also a competitive edge in maybe operating at a much more reduced cost than maybe what some of our competitors are.

Thank you.

VICE CHAIRMAN FRANCIS: Let me make a moderator's editorial comment, maybe if I can, here, and one of the things that in all the areas where we're making a lot of reference to lawyers and litigation and courts, and I think one of the things that it's important to keep in mind, and I think that this is going to come, is in all of these areas, that you're going to have an increased sensitivity on the part of the legal system of the importance and the value of these devices, whether it's aviation or pipeline or marine or wherever it is, and that those who are not joining are going to find themselves potentially at peril in courts for not having taken advantage of what is increasingly perceived as important safety tools in whatever mode you're in.

So, I think that we may not be quite there yet, but I guarantee you there are a few lawyers out there that have had this thought.

I've got another one here for Max. Given the human factors issue has played a central role in many accidents, what additional measures or data-gathering techniques would you support to better capture human factor information? For example, tractor-trailer design, driver training, driver workload, fatigue.

MR. FULLER: Well, I think there's a lot of variables that come into play when you have a truck involved in any type of accident, and the variables are condition of equipment, the technology that the equipment has, the condition of the driver and his skills, and maybe, to a certain extent, just the environment that he's in, and if the driver doesn't have the proper equipment to properly respond to maybe the extreme situation, he can do his best and still fail, and I really think that if you really look at the variables that a driver is confronted with almost on a daily basis, we, my industry, has to give that driver the most comfortable piece of equipment, some of the highest technology they can to be proactive in order to reduce accidents, and also to make sure that he's got the proper training.

One of the elements that we have 80 drivers in our company that have run more than a million miles without an accident, and the three things that they really say are critical in trying to get to that million miles plus without an accident, one is to maintain a distance, keep a distance from any other vehicle out there.

Number 2 is never get in a hurry. No load is that important to where you injure a person, and Number 3 is if you get tired, pull over right then, not an hour or two hours down the road, and I think that if we can continue to instill these thoughts in the drivers as we're training them and give them the proper skills that they really need, plus we give them maybe truck seats, more comfort, as they drive down the highway, I really think that we can improve truck safety tremendously.

If you look at some of the technologies in trucks, they really haven't changed in probably 30+ years. The truck seat is relatively the same as it was probably 20-30 years ago. That technology needs to change.

If you look at truck brakes, there again, most of the laws regarding truck brakes were written before I was born. I'm 46 years old. So, I really think that there's a time that we need to look at what the industry is putting into the marketplace, and I'm not talking about just trucking companies. I really think that you've got to look at the variables that we deal with.

Cars have evolved from drum brakes to disk brakes to ABS brakes. Their stopping distance is probably 50 percent or less of what it was 20 years ago. Trucks haven't changed. We finally got ABS brakes two or three years ago. We need to go on to that next level, adding all-wheel disk brakes and add electronic braking which gives the trailer the same response time that the tractor has, and we can reduce the stopping distance 40 or 50 percent.

Part of it is the rules have changed, but yet trucking hasn't, and I really think that trucking needs to change, and part of that's going to create a need for the federal regulators to change some of their guidelines.

A perfect example is the pneumatic braking law that's in existence. That was written back in the '40s and '50s. That needs to be changed to allow electronic braking.

Thank you.

VICE CHAIRMAN FRANCIS: This is for Mark. Are advances in technology leading gas transmission companies to identify leak and ruptures earlier, and what are some of these advances?

MR. WESTHOFF: Well, probably the most well-known is the implementation of real-time models in terms of increasing the sensitivity that we have in detecting these leaks.

Unfortunately, our experience has been, and several others in the industry have run into the problem, that these will generate as many nuisance alarms as they will actual alarms, and we already have a tremendous problem with alarming in general.

The operators still in my opinion face far too many alarms, have to weed through far too many of them to make the appropriate decisions.

In terms of leak detection of a compressable fluid, it's not a very straightforward problem. I mean just the hydraulics of it are fairly complex, and certainly less amenable to modeling techniques than, say, the liquid pipelines are in that regard.

As far as what the future holds for enhancing our leak detection capabilities, probably better real-time analyses that I talked about earlier today would be one way to do that. Doing tighter system balances and what have you, but leaks on the order of magnitude of a million or two cubic feet a day, which seems like a tremendous amount, and it really is, are very, very difficult to detect.

I've gone through and looked at cases where a significant amount of gas is blown through a remote relief valve, and from the operational data alone, it was virtually impossible to tell that anything was going on there.

So, it's an area that does warrant some improvement, considerable improvement. What those improvements will be, I really can't tell.

VICE CHAIRMAN FRANCIS: Steven, how is the replay of an event on board supervised to enhance the learning process?

MR. SPENDLOVE: When there is an incident on board, the captain would report that incident, of course, in the normal chain of events to head office and himself, and then a discussion takes place whether it would be worth at that moment in time doing that.

The incident would then be sent back to headquarters where you have a professional panel made of many a section of each department on the vessel, and they would then review that information and then decide on a basic training program, whether that be increase in technical training or whatever. But the decision would probably be made from the headquarters.

VICE CHAIRMAN FRANCIS: Jeff, there are a couple of questions here that sort of go together. One is, which aircraft within the United fleet are equipped with QARs and included in the FOQA Program, and the second is, which QAR for FOQA did you select, and why, and is the FOQA report April '96 available to the public?

MR. BAYLESS: Part of being tail-end Charlie was I wasn't supposed to get all these questions. Everybody else was going to get the questions. So, in answer to your question, we have QARs, quick access recorders, on the A-320s, the 777s and on some 737-500s. I think we have 13 on 737-500s, and we're just now equipping them on the 767-300. We have PCMCA cards on the 319s.

In the future, I think we're moving at least in the direction of all PCMCA cards have some type of remote access of the data, whether it's gate link or just physically removing the PCMCA card at the maintenance base and putting it back in the airplane, downloading it, putting it back in the airplane.

The problem is we're already discovering with over 100 airplanes equipped with these recorders, the QARs and the PCMCA cards, it's almost unmanageable to download all the data in our office. Each airplane is equipped. Every two-three days, they remove the recording media off the airplane, send it to Chicago. We download the information and then ship the disks back out to the field, and it's just a tremendous volume.

I think we have something in the order of 7,000 QARs to deal with. So, in the future, as we equip all of our airplanes, we're not going to be capable of handling it just for the volume of data. So, we're going to have to come up with some sort of gate link. That answers the first question, rather long-winded.

The second one was about the report or did I miss one in the middle? Yeah. A Penny and Giles optical unit is what we're using, and as I said, the PCMCA card. In the future, that's still up for contract, I think, as far as what the unit is.

The current technology for the new recording map that is being mandated is we're going to have to put in new DAMS unit. I can't even tell you what the DAM stands for, but it's a black box in the airplane that will have a slot for a PCMCA card, and basically we can slap the card in there and get a free add-on FOQA Program as part of that recording map requirement, the new recording map requirement.

As far as the quarterly report, no, it's not open to the public. I kind of stuck my neck out on a limb just to show you a couple slides from our quarterly report that I showed you today. Actually, we shred that data. It's not available.

We only keep identifiable data on the property for seven days. The rest of the data, the identifiable data gets trashed. The rest of the data is just like stroke counts of we've had so many excedances and so forth, but it's all identified.

VICE CHAIRMAN FRANCIS: Thank you. There are three questions here for Max that sort of have some continuity, and maybe we'll just start with two.

The first is, what is considered the most important post-crash element in the trucking industry?

MR. FULLER: I'm really not sure since I'm not an expert on this issue. Probably determining what caused the accident has got to be a critical requirement, so that we can determine how to prevent those type of accidents from occurring in the future.

To a certain extent, we learn a lot when we have an accident and what went wrong, and the one thing we do in our company is we talk a lot about eliminating the points of failure, whether that's service for a customer or whether it's a safety problem or whether it's an accident. We want to eliminate those points of failure, and we keep striving for perfection, and I think once we can identify a cause of an accident, then we can maybe retrain our drivers, maybe we can buy some new technology to help cover some of the deficiencies that maybe we found once we view what happened in that accident.

VICE CHAIRMAN FRANCIS: And the second is, do you believe recorders should be used for enforcement purposes or should they be protected?

MR. FULLER: That's probably a mixed bag. To a certain extent, I think to really do what the trucking industry needs, they probably need to be used for enforcement at some level, and that's probably a surprise coming from a trucker. But I really think that because this industry, and I'm talking about the trucking industry, the average fleet size, 90 percent of this industry has less than nine trucks, and because of that, you've got such a variety of way people will focus on safety, and I really think that maybe the companies themselves, especially the smaller companies that are less sophisticated, are the ones that the regulators are going to have to spend more time with.

If you don't have a systematic way to identify those companies, you probably can't successfully regulate them. So, I really think long-term, that's the answer. Short-term, I think the technology still has a lot to be developed before we can get to that point.

I really think long-term, that, and maybe coupled with something which is called a pre-pass system, which is an RF transponder, may connect to the black box, and if there has been maybe certain events that has occurred, it will report as this truck goes through maybe one of the inspection stations as it enters a state, and what the state enforcement officers can do is identify this being a problem driver, a problem truck, and hopefully get him off of the highway, and I really think that in order to catch these small fleets, and to a certain extent the large fleets that haven't embraced safety at the level we have, that's probably the only way it's really going to happen.

VICE CHAIRMAN FRANCIS: And, finally, have you measured, underlined, the actual pay-off in accident reduction or operating costs on trucks due to the installation of recorders?

MR. FULLER: Well, actually, at this point, we haven't. Part of the reason is that the recorders are still relatively new in the trucking industry. I know when you get in Europe, they were mandated back in the '70s, but in the U.S. market, they haven't been mandated.

A lot of people put them in trucks maybe in the late '70s or early '80s. Because of the driver turn-over situation, they took them out of trucks because the way the systems are being used was a driver got by with something. When he came in at the end of the trip, then the company called him in, called him on the carpet, and he went to work for someone else, not that the company terminated him, but because he had a lot of options.

So, effectively, the data recording system didn't work. It might have for that one company in that one instance, but for the industry as a whole, it didn't work, and I think that we're still confronted somewhat with those same issues because there is a shortage of quality drivers in the marketplace.

I can have a driver that kills 20 people, and he probably has a job from the time I terminate him to the time he gets home, he probably has two or three job offers, and I really think that that really prohibits long-term data recorders from being a total solution. That's the reason that we look at active controls.

We want to prevent the accident, and to a certain extent, we want to take what we can out of the driver's hands or give him the tools that he needs to make sure that he does prevent those accidents.

VICE CHAIRMAN FRANCIS: Mike, a couple of data questions here. When you mentioned the bad data and data recorders caused by defective sensors, please comment on sensor reliability/levels.

CAPTAIN HOLTOM: It's not very easy for me to do that because I'm not an engineer, and I don't get involved with engineering. But we -- not just bad sensors. We did get a few surprises when we introduced the A-320, Jeff, but the certain systems reset themselves from time to time. So, when you press the button for take-off power, for example, there were several resets amongst systems on that aircraft, and they generated events. So, that's something you have to be aware of, false events.

To give you an example of sensors, on the 777 fuel tanks, we have three fuel tanks, and we have something like 48 sensors in each of those tanks. And I have seen some bad sensors which have given erroneous data, and we were losing fuel apparently at a rate of a ton a minute on a flight out of Bermuda.

Well, rather than divert back to Bermuda, which would have been very attractive, I hasten to add, we did look at the sensors, which we can do on the 777. We could look at practically everything we want to, and sure enough, there was one sensor out of the 48 on one fuel tank which wasn't working properly which gave us indication of a problem.

You've got to be aware of it. I think you're always going to get some sensors failing from time to time, not necessarily at the sensor itself but the connectors to the sensor, the transfer of the data through the aircraft and nearby other systems. In a perfect world, it wouldn't happen. It wouldn't be a problem, but this is the real world we work in. There was one other thing. Lightening, of course, and static electricity does cause problems.

VICE CHAIRMAN FRANCIS: How does your process prevent data overload or overdose in light of increased numbers of parameters in the new recorders?

CAPTAIN HOLTOM: Take the 777 again. Although there are 60,000 parameters possible, we only record 2,000 of those. We do analysis of probably only about 500 of those 2,000 that we record.

For example, the fuel tank sensors, the 48 in each wing tip, we don't need to know what all those sensors are saying. What we want to do is look at what the fuel contents is during the course of a flight. That's one example of reducing overload.

The question that cropped up during discussion at lunch is how do you store this data historically, so that you can go back and look at it over a period of time, and there is quite a lot of hype about data warehousing and data minding these days, but when you're looking at five gigabytes a day, storing half a dozen years' worth is not really a very viable option, and I have to question really the validity of it, and it tends to be more the academics of this world, forgive me if you are one, who want to go back and look at a long period of history.

In the operation environment, if we had a problem yesterday, I want it fixed today, so we don't have a problem tomorrow. We don't want to go and look back too far.

So, when we do store data, which we obviously do, we store the subsets which are going to be useful to us in the future. You're never going to find the time and the resources to go through the large volumes of data that you think might look at one day. It never actually happens, unless you're a university.

VICE CHAIRMAN FRANCIS: This is a question from me. It is apparent that the fatality rate of highway crashes is significantly greater than that of airline crashes. That being the case, what efforts now are being done to promote on-board VDR for vehicular transportation?

And I will, without going into a lot of detail, say that in the last couple of years, we have put out, we, the NTSB, as a result of either an accident or study, have put out some three very strong recommendations to the trucking industry, to NHTSA and to the automobile industry in terms of the importance of moving in this area.

So, I would say that on the one hand. On the other hand, I would say we're sponsoring that which we've been attending for the last couple of days here in hopes that this will be helping as well.

Steve, how do you determine who has access to the tapes? How is the system secured? What is to prevent the 15-minute gap? I hope that means more to you than it does to me.

MR. SPENDLOVE: Yeah. I know where that question came from. Thank you.

VICE CHAIRMAN FRANCIS: Are you thanking me or the question?

MR. SPENDLOVE: No, no, no. I know you're going to cover in the next session, you know, who has access to tapes. It's very difficult. At the moment, it's the ship owner, my company, and we then will share that information naturally after we've had a look at it with the authorities or with legals or anything else.

We have never stood in anybody's way when they have asked us to share information with us. So, until the IMO and the legislation comes out, I'm afraid it will stay with the ship owners. They've purchased the equipment. They're using it for for many things. They're using it for their own training purposes as well. Until legislation comes out otherwise, it will stay with the ship owner.

The question of security. To get into a black box, I do admit it takes 15 minutes, and because it's protected from fire and everything else and reaches the IMO specifications, it's a difficult kit to get into.

So, when I say that usually after an incident, as incidents in the marine industry usually are prolonged over a few hours, we then remove the tape after the incident as it's on a 24-hour play-back anyway. So, when this tape finishes, it reverts to the original 24 hours, all the way back to the beginning.

So, you would never lose that 15 minutes of data, and we surmise that no incident would ever last more than 24 hours without a break in which we could change the tape.

VICE CHAIRMAN FRANCIS: Gary, where can you find information on how to implement a FOQA program, and I guess this could be equally to Mike and to Jeff?

MR. DAVIS: Yeah. The policy that's written is in the advisory circular that our AQP office wrote. The FOQA policy that's draft, that's probably the best guidance that we have out there.

You can call Tom Longridge out at Dulles at his AQP office, and if he has lots of information out there that he can provide for anyone that wants to set up a FOQA program, but the AC is available for that purpose.

VICE CHAIRMAN FRANCIS: Would that be on the FAA Web site?

MR. DAVIS: The draft AC? If it's not, I'll see if I can get it put on there.

VICE CHAIRMAN FRANCIS: Okay. Thanks.

MR. BAYLESS: I just have two cents to add to that, too. The UTRS Web site, I believe, is open. I'm not sure, Doug. Yes, it is open to the public. It's not only the public, but there is information in the UTRS Web site, and I'm sure if you talk to Tom Longridge, he could get you that information.

VICE CHAIRMAN FRANCIS: Thanks. This is to all panel. Has anyone in any transportation mode seen any disadvantage in using and extending the use of recorder systems? No? We're not that close to the break, guys.

CAPTAIN HOLTOM: There is a disadvantage. It just tells everybody how hard my landing was.

VICE CHAIRMAN FRANCIS: That's a personal. This is for Max. With regard to equal protection/equal justice, what do you believe the federal and state government's roles should be in promoting recorder use?

MR. FULLER: I think the issues are somewhat similar to what we heard from the airline industry in trying to make sure that the systems aren't used for improper purposes. They're not used to maybe over-regulate one company versus another. They're not used to discriminate against one driver in my industry or another, and I really think that it's got to start with education, and someone, and it's probably the regulators, are going to be the ones that really have to start the education process, and they're probably going to have to sponsor a program to help people in my industry really understand this technology, and that it's not something where this industry's going to be slammed, and that's kind of the opinion, I think, that most people have.

I think what really needs to happen is with proper education and maybe a development program to help continue developing the system, like we're working with Eaton-Vorad people, to get the technology somewhat more perfected.

I think once you get to that point, then there's going to be more people in the trucking industry probably willing to accept the merits of what data recording can do, and probably take them to that next level to where they can be proactive on safety.

But someone has to lead the pack, and it's probably going to be the regulators that does it. I can't do it with my company. I'm not that big, and I'm not sure that a group of us can really change this industry that significantly. It's probably going to take someone at the federal or state levels to make it happen, and education is probably the first place.

CAPTAIN HOLTOM: Can I comment on that, Bob?

VICE CHAIRMAN FRANCIS: Absolutely.

CAPTAIN HOLTOM: Can I comment on the -- with due deference to my colleague on my left. The regulator leading the industry because I have a personal view on this, and I'll say it's to do with the U.K. regulators. So, I'm quite on safe ground there.

I firmly believe that the regulators do not lead. The industry leads, the regulators follow by looking at better practice and then making it mandatory. So, I don't believe the regulators -- if you wait for the regulator, you'll wait forever. Sorry, Gary.

MR. DAVIS: That's all right.

VICE CHAIRMAN FRANCIS: This is a question with no addressee, and maybe we'll share this. Have you considered utilizing real-time sharing of the FDR data with the ATC data to provide the global safety system? And two, recording the information on the ground instead of aboard the aircraft, and let me just say that there are people certainly that are focused on this issue.

I guess I would harken a little bit and maybe particularly in this country back to Ed Soliday's crawl/walk/run approach to this. There are lots of sensitivities in setting up these programs in aviation and as we've heard in the other modes, and the way to do it, these are programs that are based on trust and confidence and cooperation, and once the egg is broken, the reconstruction of that egg is an enormous problem facing us.

So, I think, sure, this is going to be, and it harkens back a little to what Dan Goldin was saying last night, we are going to be getting to this, but we've got to make sure that when we're moving forward, we're moving forward at a pace that's comfortable for everyone.

In terms of the ATC, I'll just mention there's been reference to pilot self-reporting systems, and American Airlines has got certainly one of the best in this country.

There is also a controller self-reporting program. I think it's in the Dallas-Fort Worth or the Dallas Tower facilities in the Southwest Region of the FAA, which is doing the same kind of thing. So, yes, people are thinking about ATC and operations. Yes, they're thinking about cooperation and more real time, but I think we're going to be moving toward that, but not at the risk of breaking the egg.

MR. BAYLESS: You answered the question better than I could. We have not really looked at that. First of all, it's not feasible.

VICE CHAIRMAN FRANCIS: Don't tell Soliday I quoted him.

MR. BAYLESS: It's not feasible under the technical or the cost of the equipment currently just would not make that feasible. But, secondly, I'd like to address it from a different angle, that we found more value in addressing a system problem rather than an individual problem because, as you know, if you take real-time data and compare it with real-time ATC data, you're going to deal with an individual, and it will be identified and identifiable data.

We've taken a different approach where we take a system -- for instance, Monterey. We've had a lot of ground proximity warning indications in Monterey because of the mountainous terrain. We took the data to the ATC folks, and we said we need to devise a different approach into your approach to avoid this mountainous terrain as we come over the mountains, and sure enough, we're in the process right now, and I'm not really sure where it is in the stages of the FAA, but it's very close to being implemented. We're going to design a new approach strictly based on FOQA data and enhanced ground proximity terrain mapping data to bring the aircraft through the valleys into Monterey and avoid all the mountainous terrain.

We think that's a much more effective approach than dealing on an individual basis.

CAPTAIN HOLTOM: I think it's important to share not the data but what you learn from the data. The data that we're talking about is so complex. I mean, for example, we have operations ranging from a turbo-prop up to a Concorde with sophisticated jet turbines and non-sophisticated in between.

If we gave that data to air traffic control authority or a union of some kind, I don't think the outcome would be very beneficial to either party.

What we learn from the data -- for example, I showed you our approaches into Los Angeles. That's the sort of thing, I think, we should share with the authorities.

There is one slight exception. I think we could perhaps give the air traffic control units more information on TCAS data, but that's part of our general sharing within the industry anyway, and this sort of thing does go on, not only with air traffic control but with manufacturers, and I can assure you if one part of British Airways, such as propulsion or air traffic services people, find there's a problem that we've learned through the data, they'll be banging on the appropriate doors around the world, whether it's with Chinese Air Traffic Control or manufacturers in Spain or wherever it is.

So, each party does it on a one-by-one basis. I'm not sure of the benefits of circulating that widely around the world because that's going to make people very nervous about data and the damage it could do.

VICE CHAIRMAN FRANCIS: This is the important distinction between data and information, I think, that we're always talking about.

Max is, I think, going to win the prize for questions here. Next to human factors, improper brake adjustment is a leading factor in trucking accidents. Will future recording and monitoring systems be able to tell us when truck brakes need adjustment?

MR. FULLER: Actually, I'm not sure if future recording devices will be able to tell that, but there is technologies. Once the electronic braking systems are allowed in the U.S. market, and I'm really talking about down to the point that we bypass pneumatic system, there is sensors along with electronic braking system that can sense brake lining wear, brake lining condition, as well as brake adjustments, and electronically, it can electronically adjust the brakes, and what that will do is basically every time the driver puts on the brake, it will be almost perfectly adjusted. So, when he needs it, it's there.

If you look at the mechanical systems that are in the marketplace today, they have something called an automatic slack adjuster. They're still relatively sloppy compared to what can be done with electronics.

Once you apply electronics, maybe the data recorder can record that information, but until then, it probably won't happen. I think that electronic braking is going to be something similar to what the Internet was for the PC. It gives us a whole new platform of vehicle control that we've never had before.

VICE CHAIRMAN FRANCIS: The last question we have up here is for Mark. Since SCADA systems have been demonstrated to be cost effective in running the business of safety, will industry be more proactive in the use of tools, such as SCADA, remotely-controlled valves, rather than waiting for mandatory regulation?

MR. WESTHOFF: That's a good question. We at our company, I think in advance of the regs, have taken it upon ourselves to instrument a lot of valves in what we call Class 3 locations. Those are locations in which population densities reach a certain level in fairly close proximity to the pipeline right-of-way, and indeed we've gone in and automated many of those valves to be able to be actuated from a remote point, be that in the field or at the central pipeline location.

I think that anything that lessens the impact, mitigates the impact of a rupture or something of that sort is in the company's best interest, and I think it becomes self-evident.

There's a lot of our pipeline, for instance, that goes through the Red Desert of Wyoming, and I don't know if any of you spent a career driving through that, but it's very remote, and it is an area that doesn't see a lot of human intrusion, and I don't know if it will ever be very cost effective to automate those valves in particular, apart from just sending people out there to do this job manually.

But certainly in urban areas or near urban areas, this is probably going to be an area that pipelines will be aggressively pursuing. I know in our case, we are.

VICE CHAIRMAN FRANCIS: This is another NTSB question. The panel said no comment to disadvantages. Would therefore the NTSB issue a call for all vessels entering U.S. waters to have to make a data record while in U.S. waters, even if the U.S. Coast Guard disagrees?

I seem to recall very shortly after I joined the Board four and a half years ago that we made a recommendations to the Coast Guard that they propose, and it was a fairly strong recommendation on the use of recorders, which they, I believe, have taken on to the IMO, and I don't know whether Barry Sweedler's in here, but I believe that that's currently acceptable response, is it not?

MR. SWEEDLER: Yes, sir.

VICE CHAIRMAN FRANCIS: So, that's the answer to that. For Jeff, you mentioned that the engineering community was far behind the pilot community in using FOQA data. What is United doing about this?

MR. BAYLESS: Good question. Our engineering folks have another tool called ACMS or aircraft condition monitoring system, which take real-time snapshots from the aircraft on ACARS, and they can look at the condition of the airplane.

The problem is it takes just snapshots. It takes one or two or three shots on the interior of the air frame components per flight. With the FOQA data, you have it continuously. So, what we've done to get engineering's buy-in is supplied them with data. They've had a problem with an engine, for instance. We had a 777 engine failure, gave them the FOQA disk and let them look for indicators that could have shown that something was happening to that engine prior to the event.

So, with that, they've become very interested, and they've almost swamped us now with requests for data. So, we're moving forward with a new FOQA Program, we're calling it possibly FOQA 2, that would collect all the data and allow engineering to do their own queries on the database, and that's what they really need.

They need about a year's worth of data to go back at a little bit different angle than Mike was saying, but to go back and look at the history of an air frame, a valve component, an engine and so forth, and just to do some data mining to see if there's indicators that they can point to that showed an impending failure, and then look for those indicators in the rest of the database, and perhaps prevent a catastrophic failure before it happens.

VICE CHAIRMAN FRANCIS: Mike, do you want to add anything?

CAPTAIN HOLTOM: No. Yes, I do. Of course, I do.

VICE CHAIRMAN FRANCIS: Are we putting you on an airplane tonight?

CAPTAIN HOLTOM: Yeah. When we first came to an agreement with our union, the pilot unions, they accepted that the engineers would be the ones who were handling the data in the early days, particularly for analysis. Of course, it was engineering analysis which took the higher priority, and that's where, of course, the biggest savings are made within the company, and that pays for the whole program several times over.

VICE CHAIRMAN FRANCIS: We've run out of 3X5 cards, at least up here, and we're 15 minutes before. If there are any questions that have occurred, last-minute questions that have occurred to anyone, we'll try to do it by verbally rather than on the card.

AUDIENCE QUESTION: It's interesting to observe the philosophical differences across the transportation modes regarding the purpose and use of recorder information. I see in the air mode, there's a strong tendency to say we're not using it for any form of punitive action, and in the case of British Airways, they don't even need to know the identity of the crew members.

In the marine, we're finding recorders vindicate the crew members that they didn't do the bad things everybody said, and in the transportation on highways, we're seeing we're getting the bad drivers off the trucks.

I wonder if any of the panelists have changed their views from those that they brought to the meeting having heard all the modes.

CAPTAIN HOLTOM: No, I haven't changed my view. Not because I don't like to suggest I'm intransigent. It's a bit more serious than that in that if you do have a bad truck driver or you do have a bad pilot, and we do from time to time, we have to analyze why we have a bad pilot, and with three and a half thousand pilots, I can't pretend that every one of them is perfect.

We have to analyze our selection procedures, our training procedures, our checking procedures, and the way we handle people from day-to-day as a management team, and when you're handling people, as you know, people have problems. They have stresses. They have divorces. They have children. The dog dies. They don't catch the big fish they were hoping to catch, and they bring their problems to work quite often.

Responsible management has to handle all those differences, and I would like to think that when we set up an operation, and we introduce procedures, and we buy equipment from the manufacturer, we take account of the fact that you could have two people flying together in that piece of equipment at high speed in a very unfriendly environment. Those two people might be having the two worst days of their lives.

I would like to think that everything we set up, the infrastructure we've got around them copes with that. So, we don't actually want to fire the guy who gets it wrong because if he got it wrong today, it could be me that gets it wrong tomorrow because I've made the same mistake because I came through the same system and operate the same equipment that he operated.

MR. DAVIS: Can I add just a few words? From the Federal Aviation Administration standpoint, we don't believe that the purpose of that recorder is to go out and find a pilot doing bad things. We think these pilots are probably -- they have their friends in the right seat, and they have their friends in the left seat. They have a bunch of flight attendants in the back that are watching out what they're doing if they land long or land hard, plus many of them have 250-300 passengers back there in the back that just love to write us letters to tell on them, that they're running a bad show or the company's running a bad show, plus they have air traffic controllers that are talking to them continuously and just love to write letters to say how badly they're doing or how great they're doing. But it's usually how bad they're doing.

We think the process is there to weed out that bad pilot, and this tool is such great benefit to it to help us in the long run, that we don't want to use it for that purpose.

VICE CHAIRMAN FRANCIS: Max, you want to take it?

MR. FULLER: I think that if you look at the trucking industry, it's much more diverse than what you've got in the airline industry. Quite a few more drivers, quite a few major differences in background.

The other thing is if you look in my company, I've got drivers that we give a truck, and they may be gone for three weeks at a time. They may be 2 and 3,000 miles away from our office. If we don't have some systematic way to be able to determine what they're actually doing, then it's hard for us as a company to regulate the quality of the driver that we've got.

If you look at the airline, they're probably going from Point A to Point B and being somewhat tightly controlled. That's not necessarily true in the trucking industry, and I think that you can't compare the airline industry and the trucking and say it's the same because it's not. The industries are substantially different. The environment is substantially different, and I really think that data recorders can give us, the company executives, tools that we desperately need to determine what that driver's doing when he's 2 and 3,000 miles away. Without it, we don't stand a chance.

VICE CHAIRMAN FRANCIS: Anybody else like to comment?

(No response)

VICE CHAIRMAN FRANCIS: One of the interesting issues, as pointed out here, there are enormous differences, but one of the things that I think Max is getting at a little bit is an issue that still is very topical in aviation, and this is the whole issue of record sharing, pilot record sharing, and how do you do this, and what's fair in terms of confidentiality and privacy, and at the same time, where people are moving, and he's talking about two job offers before the fired guy gets home.

Well, there are some similar kinds of things in aviation that the FAA and NTSB and the Congress are trying to deal with in terms of the same kind of issues. Is it exactly the same? No. But certainly there are similarities here.

Well, I thank particularly the panelists. I think it's been an interesting morning and afternoon.