Session I Panel Discussion

MR. SWEEDLER: I'd just like to ask a general question that each of the panelists can give us a short answer on. It has to do with the future. What do we think the future holds for recordings in each of the modes? Maybe that's a big question, but if we can try to answer it in a short way, I'd appreciate it.

Dennis, Mike. What do you see?

MR. GROSSI: Well, want me to go first? We

did cover the future in our presentation this morning, but I would say that the next big development in recorders, once we have the independent power supplies, and that seems to be technically feasible, once we have that in place, will be the implementation of video recordings.

Mike, would you agree?

MR. POOLE: Yeah. Yes, and the only thing I might add is, and it was touched on this morning, I think in the aviation world, the new generation airplanes, because the data is available and used by the airplane to operate it, we're starting to see 11-1,200 and close to 2,000 parameters as routine and normal. So, the future is if the recorders survive, they have a wealth of information on there detailing the operations step-by-step.

MR. GROSSI: I think we might also add the use of flight recorder data for FOQA-type programs and the true accident prevention purposes, and eventually we could even do real-time processing of the flight recorder-type data to improve the fidelity of things like air traffic control communications and handling.

MR. SWEEDLER: Is there any truth about video recorders? What would the cameras be focused on?

MR. GROSSI: Well, there are quite a few concepts being proposed. Some have some problems with the unions are not real excited about it. In fact, the Airline Pilots Association, however, is on board and supporting the use of video recorders, but in a very limited fashion. Their position is to have the cameras focused on the instrument panel and the control panel and the control pedestal, but to keep the image of the flight crew out of the view of the camera.

There's proposals to have cameras externally mounted to show the exterior of the aircraft on a position on the undercarriage. A lot of the external use of cameras would be for ground security purposes, I believe, would be some benefits there which are not necessarily accident prevention areas.

Mike, would you like to add to that?

MR. SWEEDLER: Admiral Lang, any bullets on the future?

ADMIRAL LANG: The future of VDRs for the marine world. We've got a problem. We haven't really sort of started at all yet, and one of the difficulties is the many nations who have the main number of ships, like St. Vincent in the Grenadines, like Belize, like Antigua and Barbuda, some of the names which feature on the Caribbean map, and there are more, they're the people who own these ships, and, frankly, they're not interested.

So, we've got to overcome that hurdle. I have no doubt at all they will come in once people's bottom lines start being affected. If people do ask have you got a VDR on board, if you haven't, they don't use your ship. That's what will make people wake up to them.

But, also, I think that once governments realize the cost benefits from having them, they will support their introduction, but, ultimately, I just want to get the basics, the ability to record what happens on the bridge, the basics just to get the essential features of the ship's movements recorded, and, of course, our difficulty is not just the 30 minutes before the event, sometimes we have to record our information over several days, if not weeks, and I've just finished an investigation where the information had been languishing for 25 years, and it cost HME a lot of money to finally get it back again. Once they realize they can save money by having that instantly available, they will come in. So, we've got to go to the first step before we get on to what I particularly agree a video camera fits on the bridge will make a huge difference to really understanding what they're all about.

MR. SWEEDLER: Captain, I have a tough question for you. Does the Coast Guard equip any of its ships with data recorders?

CAPTAIN ROSS: My understanding is that we have equipped a couple of our ships with some voice recording, and we do use recording instruments in our search and rescue stations and in the vessel traffic services that we operate.

I anticipate that we would expect that we would see more voyage data and navigation-type data recording in the future.

MR. SWEEDLER: Thank you. How about in the rail industry? You talked a little bit about some of the things, but, Myles, what do you think?

MR. SIBLEY: Yeah. I can see the future being influenced from a number of different angles, and I've just jotted down a few since you asked the question in fact.

Train performance from the point of view of the contractual arrangements which are commonly in the U.K. being put in place particularly upon new builds of rolling stock, where the maintenance may be contracted out perhaps to the manufacturer, and very tight and quite heavy penalties could be incurred if they choose a reliability and availability, for example, aren't met and maintained would be one driver.

Rail Track, who charges train operators to use its infrastructure and equally content to penalize them if it incurs delays, etc., is another area where train operators are beginning to use on-board information.

I know this is drifting slightly away from the scenario of accident data recording, but I think that the drivers are there to introduce the technology that can allow it in a number of areas, and, of course, safety has to be my main driver.

Human factors is an issue dear to the hearts of myself and colleagues, and to be fair to the industry, what are root causes of many of the accidents that do take place, issues of fatigue and distraction, indeed the ergonomics of cab design, another driver.

From Europe, we're seeing initiatives that will drive inter-operability between the U.K. and Europe, and some high-level and in some cases quite detailed draft standards are emerging, and again I can see data recorders from an accident prevention and cause being a significant factor there, and, of course, there's this symposium itself.

I should certainly take away, even though it's only just started up and making a number of useful notes of pointers and questions I would like to take back and discuss with colleagues. So, who knows from that point of view.

MR. SWEEDLER: Anything further from Gary or Ed?

MR. WOLFE: As far as the future, I think one thing gratifying is the big railroads add to their motor power fleets the big AC power. We're seeing a cascading of units down to yard service in regional railroads which is giving us broader coverages with recorders that we didn't have in the past.

I think we'll continue to see a trend towards solid state. To me, it makes sense. The maintenance on the eight-track recorders is high at times, and the reliability questions we've raised will continue the trend towards, I think, solid state recorders more and more.

And then, finally, depending on what the colleague on my left here decides with the new FRA rules and what comes out of this process we're going through, will also further dictate movement in the recorder industry, the legislation that comes forward.

So, that's kind of where I see the future.

MR. SWEEDLER: Ed?

MR. ENGLISH: As I mentioned this morning, we're certainly looking at new regulations for a new generation of event recorders. Certainly computer-based, to do away with eight-track recorders, tape recorders, but I think what's even more important in the future is to make sure that we make provisions for new technology, particularly in the rail industry.

We look at positive train control which we know are coming down in the future. We need to be able to make sure that recorders will be able to record the necessary information needed in positive train control.

I'm not sure that we can answer just what data elements need to be recorded in that new signaling system, but that needs to be addressed. Things like electronic pneumatic braking, that certainly needs to be an issue, that is be able to record and the new event recorders.

So, I think that what we're going to look at in the future is new technologies being able to record that data on the recorders.

MR. SWEEDLER: Ed, while you have the stage, what are your views on voice and video recording in the railroad industry?

MR. ENGLISH: A good question.

MR. SWEEDLER: A personal view or FRA's?

MR. ENGLISH: Well, my own personal feeling is that I'm not sure that there is a real need for video recording in the railroad industry. Certainly railroad accident investigation is a little bit different than in the aircraft industry. When there is an accident, the remains are there. The trains don't break up into little tiny pieces like they do in the aircraft industry. So, it's fairly easy to reconstruct the equipment, and with the various data elements that we record, and we have available to us, I think we can reconstruct the accident.

Voice recording, again it's questionable. There are some issues on privacy that people in the rail industry have raised, and I think that Mr. Kauffman tomorrow is going to address that specific issue, and again I think that's still open for debate as far as the rail side is concerned.

MR. SWEEDLER: Okay. Well, there were a number of accident investigations that I've been involved with where I think it would have given my eye teeth to have some kind of video recording of what went on in the cab before leading up to the accident.

But any other panelists want to say something about that particular question? Dr. Lehmann?

DR. LEHMANN: Well, if I may, I would like to give some statements on the highway recording aspect. Let me compare again U.S.A. and the European Union.

In U.S.A., we have round about seven million large trucks and round about 300,000 recording systems installed in them. In Europe, it's just different. Round about five million trucks, and all of them are equipped with recording systems.

So, the question is what is the effect in terms of security, and as I tried to highlight, in Europe, we have decreasing numbers in actual fatalities as well as in fatalities measured with kilometers driven.

In America, it's a little different. Here, we have an increasing number of fatalities with trucks involved. The conclusion out of this could be that U.S.A. maybe should consider a law which is similar to the European one, but I think two effects have to be composed over here.

First of all, such devices have to prove that they are capable of cost saving for fleet management applications as well as they are accident-preventing.

Thank you.

MR. SWEEDLER: Thank you. Captain Ross, you wanted to say something about the future?

CAPTAIN ROSS: I think in considering the maritime sector, it's important to be able to recognize that there's not one maritime industry. There are many. Fishing and liquified natural gas carriers, container ships and tank ships, passenger ships and smaller tug and barge operations. The perceptions of risk, the economics, and the dynamics are different in those various sectors.

In the near term or relatively near term, I believe we're going to see mandatory voyage data recorders on a limited range of vessels. Voluntary use on other classes of vessels and particularly at the high end. Relatively few vessels, existing vessels will be retrofit with voyage data recorders, except possibly passenger vessels and some of the other high-tech vessels, perhaps such as liquified natural gas carriers.

I think VDR will be incorporated into more new builds, particularly as newer vessels are built with digital systems which are perhaps more amenable to recording.

Further into the future, I think mandatory VDR may well be extended to other classes of vessels, and outside of a full VDR, I believe that other recording systems will become more used in recording, particularly in navigation information, for casualty reconstruction and so forth.

MR. SWEEDLER: Good. Dr. Lehmann, you really have hit a cord with the difference between what's happening in Europe and other parts of the world and here in North America.

You might give us some background about the arguments that are being made here in this country about why they're not being used, and if the same arguments were used in Europe and how you overcame those arguments.

DR. LEHMANN: Well, I very much liked the speech of Admiral Lang this morning because he mentioned the same arguments against such devices as we also hearing them over here.

It's the cost. It's the problem that may be shortcomings of the drivers or of the dispatchers will be identified, and having that in mind, the accident preventive character of such devices and the information they deliver to the dispatcher and operator, okay, these arguments are frequently heard.

But in Europe, the situation is rather different, as I tried to explain. Even there, the drivers do not consider this device being a spy in the cab, but they like it because it prevents them from exploitation from their operators, and they want to contribute to road safety as well over there.

MR. SWEEDLER: The particular question I had. When the European Union required these devices, was there general at least acceptance initially on the parts of the owners of the companies or did they fight that in the beginning or what was the climate?

DR. LEHMANN: So, this was a learning process. By the way, tachographs were not developed for accident prevention in those days, but as a fleet management tool, and only after some years of experience, it was identified that they have accident preventive character as well, and then the European Union issued a regulation which in the beginning was to the disliking of the parties concerned, the drivers as well as the operators. But after some years of experience, they changed their mind.

MR. SWEEDLER: So, it moved from business purposes in a voluntary way to how it could improve safety?

DR. LEHMANN: Exactly.

MR. SWEEDLER: Okay. Any other comments on that particular question? The resistance and how it might be overcome? Mike?

MR. POOLE: I'd just like to make one comment, and that is over the last 40 years, the aviation side has evolved, and a lot of the arguments that are going around in the other modes are pretty much the same arguments that were made for the same reasons, and they took a long time to overcome.

People eventually realized the benefits, and there has been, I would say, a significant cultural change within the flight recording industry in the last 10 years, and I sincerely hope that the other modes can shorten that time frame, given the aviation experience.

MR. SWEEDLER: Other thoughts or comments on that?

(No response)

MR. SWEEDLER: With all of the improvements or the rapid improvements that are being made in instant communication, do any of the panelists have any comments particularly in any mode, but I'll start with aviation, on doing real-time processing and recording the information away from the vehicle on ground at a command center or the control center of some kind?

MR. GROSSI: Yes. I think we could improve fidelity of information that's being used in particular by air traffic controllers. If we were to do some real-time processing of the parameters, such as, say, heading for aircraft and brake release. Right now, some airlines are using brake release position information to compute pilot's pay. So, why couldn't we do the same to compute when the airplane is ready to take off?

So, the capabilities are there. It's just a matter of building the infrastructure to use this information. So, I see the potential for it in the possibly not-too-distant future to have real-time processing of the flight recorder-type information.

MR. SWEEDLER: Have there been systems developed that can do that?

MR. GROSSI: Well, ACARS systems right now, the ones that actually are used, like I said, by some airlines actually to compute the pilot's paycheck, when they push back from the gate and so on and so forth.

So, the information's available. It's a lot of information to process real-time, but I think with the right parameters, you could in fact do that.

MR. SWEEDLER: Anyone else want to comment on that?

MR. WOLFE: I think on the railroads, there's already some manufacturers who have produced hardware that will automatically download data as a locomotive passes a wayside point or is in a fueling station, which I think would greatly facilitate a lot of the manual labor intervention to get the data off of these locomotives, and I think we'll probably see much more of this as well.

MR. ENGLISH: And there are several of the large railroads that are doing that now, are downloading data and looking at it in comparatively real-time.

MR. FROSTELL: If I may, on the marine side, I'm a little cautious about real-time analysis. Yes, obviously we will make use of it, but I think the problem with so many maritime accidents, what happens in those few minutes before the event are possibly only quite a small portion of the overall investigation which might have to probe back sometimes several years over the design, the way the company is operated, way people were trained and certificated, etc.

If we start ignoring those for like the attractions of real-time analysis, it might actually get to what underpins the accident. That's why I'm slightly cautious, and it might distract us from that terribly important dimension of accident investigation.

MR. SWEEDLER: Along that same line, has there been any work or any thought to having ejectable recording media that could be more easily retrieved either in aviation, marine or at least in those two modes?

MR. GROSSI: If you read my paper, you'll find out that the original flight recorder concepts had provisions for an ejectable recorder. Nobody picked up on that, except for the military, and there have been some military applications of ejectable recorders, and the problem with ejectable recorders is defining when to eject. You don't want to do it too late. You don't want to do it too soon. So, therein lies the dilemma. So, that's, I presume, why ejectable recorders have not been embraced by the commercial aviation industry.

CAPTAIN ROSS: In developing the technical standards for the voyage data recorder to implement the IMO performance standards, the International Electro-Technical Commission, which is the body that sets those standards, looked at the issue of float-free type of devices for maritime, and the decision, based on the recommendations from some of the accident investigation organizations, such as NTSB and Transport Canada, was that it would be better to have those devices not float free.

There's some significant engineering obstacles and some operational obstacles, and it felt like it would be better to go ahead and recover those from a sunken vessel rather than incur the other detriments associated with float-free.

MR. SWEEDLER: Any other thoughts on that?

MR. GROSSI: The two of us were on that committee, and I think Mike would maybe like to clear the record up there a little bit.

MR. POOLE: Just to clear that up, I believe the IEC's decision was it would be up to market forces to choose which method, and we allowed in the specification for either is acceptable.

Our only concern with the float-free was that the radio beacon which would be used to locate the device if there was a fire on board that exceeded the survivability of that beacon, then the deployment mechanism would be inhibited because we were afraid we wouldn't find it.

MR. SWEEDLER: Any other thoughts on that?

(No response)

MR. SWEEDLER: Here's a question for Dr. Lehmann. You talked about some of the data that you were gathering, had been gathered in Europe showing benefits of the recording devices for the large fleets, for the commercial vehicles.

You did mention the one fleet in Connecticut, but are you aware of data being gathered here in the U.S. that might help move this forward?

DR. LEHMANN: Well, the data material available to us is very poor. There is no market survey, and my colleagues from VDO North America are just commissioning a market survey in order to get more data.

How to put it. We have 300,000 equipments in the market. I think there are seven or eight, maybe 10, suppliers, and, of course, the operators that buy such equipment, they are very reluctant releasing this information because it's of competitive value, and therefore it's difficult to get hold of it.

If I may refer again to Europe, it's very different. In Europe, the operators enforce to keep the data records for one year for every driver in order to make so-called company inspections possible. So, we have two kinds of enforcement. We have road inspections and company inspections.

So, the operator as well as the enforcement agencies, they know everything, if they wish so.

MR. SWEEDLER: Along that same line, I guess it's pretty clear that one way to get acceptance of these recording devices or one of the ways is to show a business or economic benefit. It makes it a lot easier to use these devices for safety if it can be shown that it can have economic benefits.

But we talk a lot about the benefits to large fleets, whether it's a commercial maritime fleet or a commercial motor vehicle fleet, but what about the fishing vessel or the small owner-operator truck, the driver might just have one truck or just a few trucks?

Anyone want to comment on that?

DR. LEHMANN: Well, I don't know the percentage of owner-drivers here in America, but we have some states in the European Union, Italy for example, where 80 percent of the owners have just one vehicle, and, of course, they do not need any fleet management system. They know about their activities.

But on the other hand, nothing prevents them from driving regardless whether they have 20 colleagues or it's just one driver, and therefore also for these people, for these drivers, recording systems make sense.

MR. SWEEDLER: Okay. Dennis, and Mike, here's a question about what is the basis of the 10-minute independent power supply for CDR? What went into that particular figure?

MR. GROSSI: Well, we just looked at the accident history we had, and the most recent accident, the SwissAir accident, actually had the longest time period where we were without information, which was in this case six minutes, and, so, we rounded it up to 10 minutes to cover any event that may even go longer than six minutes. So, that's precisely it.

MR. POOLE: Maybe I'll just add, too, that that 10-minute independent power supply is predicated upon a two-hour CVR. So, we wouldn't like to see 10 minutes on a 30-minute recorder effectively making it a 20-minute recorder, but we felt that compromising, if you will, 10 minutes on a two-hour recorder was an acceptable ratio.

MR. SWEEDLER: Okay. Another question to both of you -- well, at least to Dennis. What is the U.S. requirement to repair a failure in a CVR or FDR? How much time does a carrier have to make a repair?

MR. GROSSI: Well, you can't fly the aircraft with one recorder inoperative, but I believe it is a maximum of three days, and it must be repaired when you reach the next repair facility, but that's not the complete picture. There are other provisions, and I have to admit I don't have that precisely in the top of my mind right at the moment exactly what it is, but it's something along those lines.

MR. SWEEDLER: Okay. Gary, you had raised some questions about limitations on horn sound, and what are the technical limitations for recording horn sounds or other sound data that you referred to in your talk?

MR. WOLFE: Yeah. It's a very controversial point, particularly with the litigation environment surrounding train-car collisions often. There is a requirement, as you know, to sound the horn for railroads at the whistle post in most states. That location is a little different state-by-state, but basically it's about a quarter of a mile.

The problem is when you use pressure transducers sometimes in the horn line to show that air is being delivered to the horn bugles, those pressure transducers become dirty, clogged with dirt, out of calibration perhaps, and they may not actuate precisely when the horn is actually blown. That is one issue which we have tested a number of times and found the horns producing full 96 db of sound which is the federal requirement, yet it's not showing up on the event recorder because the transducer is not quite actuated yet. That's one issue.

And then a broader issue is with magnetic tape technology, where there's anywhere from a six- to 12-second delay between when the horn is sounded and when it actually appears on the tape is crucial when you're moving 90 or a hundred feet a second. Eight seconds -- 800 feet is obviously quite a big deal on the timing of the horn blow.

So, until the technology can precisely identify when the horn is sounded to the second almost, I think it's just going to water up and muddy up a lot of the litigation that derives out of crossing accidents, and that's about all the horn channel is really used for when you come down to it.

The railroads don't really use it that closely for monitoring. They do sometimes, but that is a problematic area of recording the horn, and then the other side is, that I found curious, is in many jurisdictions, when the horn is being sounded on the tape, and when it shows up quite nicely on the tape, and it proves the crew was acting properly, the plaintiffs try to throw it out as evidence because we are not recording precisely the sound, only the electrical contact closure of the switch, and, so, many judges have dismissed the evidence of the horn being sounded because we don't have a vocal recording of the bugle sounding a noise, and, so, it's almost a lose-lose situation, and until the technology improves, I think that is the technical limitations on recordation of the horn sounding for at least the time being.

MR. SWEEDLER: Any other thoughts on that?

MR. ENGLISH: The working group that was established to look at developing the new standards for event recorders has had some discussions with Gary and his group in Atlanta. We have discussed this issue extensively within that working group, and it is an area that a decision has not been reached on how we will resolve that and whether the bell and horn will be recorded or not, but certainly a controversial issue, and one that needs to be addressed further by this working group.

MR. SWEEDLER: Most of the use that we've spoken of today deals with using the data from recorders for accident investigation purposes. There has been some comment about using recorder data to improve safety and accident prevention. This raises a lot of other problems, and I guess we're going to get into that later in the symposium, but do any of the panelists wish to give us some insight into how this data can be used for accident prevention without getting into the invasion of privacy problems that could result?

I know FOQA is one.

MR. GROSSI: Yeah.

MR. SWEEDLER: You might explain what that is.

MR. GROSSI: Right. I realized that after I said that, that not everybody in this audience would know what FOQA mean. It's Flight Operations Quality Assurance. It's actually a pro-active use of recorded data.

I'd just like to mention we have two sessions that will in fact address those specific issues in some detail coming up. So, I don't know that I could really add to anything at this point in the symposium. I would just kind of ask for the questioner's indulgence or patience and wait for those two sessions to come up, and those issues will certainly be answered in great detail in the course of those two sessions.

MR. SWEEDLER: Anyone else want to comment on that? Gary?

MR. WOLFE: Yeah. I'll just say I think the rail industry has been very pro-active in using recorders for other than crash purposes in terms of periodically pulling these tapes out or downloading these recorders to review engineer performance over the road, to look for things like proper brake usage, proper fuel conservation principles, and most of the railroads I've worked with use these things on an almost daily basis instead of just for accident purposes. So, I think the rail industry's making good use of them.

MR. SWEEDLER: But are they being used with individual operators involved, individual locomotive engineers, for just the training purposes, trends, patterns?

MR. WOLFE: Well, yeah, in fact, I've been involved where we pulled the tape, brought an engineer in, sat down and showed him or her the recording of his run or her run and show where the errors were, how he could do better, and had a pretty good feedback session using it in that fashion.

When these things first came out in the '70s, you know, every engineer thought it was going to be Big Brother watching him, and my comment was if you're doing what you're supposed to, this is going to be your best friend you've ever had because it's going to eliminate suspicions and things like this, and I think that's proven to be the case with most people. They realize the event recorder exonerates their behavior in 99 percent of the cases if they're doing things properly, handling the trains correctly, and they have been receptive to feedback when the recorders have been used as a teaching or training tool.

MR. SWEEDLER: Yeah. I think the employee protection aspect has been well proven in Europe, too. I mean the employees seem to appreciate having it.

DR. LEHMANN: So, please let me mention one project we conducted in Europe, by the European Commission. They equipped in the Netherlands and in United Kingdom 340 trucks with recording systems and made a one-year field trial, and the result of this trial was that the vehicles equipped with such devices had a reduction of accidents by 28 percent, and this is not, let's say, a manufacture figure, but a figure which was analyzed by U.K. and Dutch experts, and this could be a magic figure.

My company analyzed several fleets after the installation. I'm not talking about a tachograph now. I'm talking about the accident data recorder. Analyzed several fleets with more or less the same result, and one very important project was the police of Berlin. They equipped all their 417 patrol cars with accident data recorders, and they had tremendous reductions of accidents. They drive like hell and having this device on board, they drive much more careful. I could quote some figures.

The reduction was 20 percent in the average and 36 percent in emergency trips, and the cost reduction, and this is important, was 26 percent. So, these are significant figures. Unfortunately, we can generalize them for the, let's say, normal traffic on roads, and therefore we are just conducting a project which we call "young driver", where under the auspices of the European Union, the vehicles of 800 young drivers were equipped with such devices, and there's another panel, same size and same population of young people, which do not have this device, and at the end of this year, we will know more about the preventive character of such devices for, let's say, the normal transportation on roads.

MR. SWEEDLER: Maybe it's the kind of thing we can have our parents put the device on their young person's car when they first start driving. I recall a personal experience driving in a taxicab in Brussels with the sign in the back, and it being used for marketing purposes only, saying "This vehicle is equipped with an accident crash recorder for your safety".

DR. LEHMANN: Thank you for mentioning that because it's a nice investment from our end.

MR. SWEEDLER: Admiral Lang, since the majority of flags of convenience vessels are regulatory avoidance issues, an editorial comment is being made, but why or how do we get around this issue so we can get approval of the recorders on vessels?

ADMIRAL LANG: The problems arising from the flags of convenience, not everybody may necessarily know what we're talking about. There are various regimes around the world who have less strict requirements on safety and other factors, and therefore ship owners looking for the cheapest option will go to some of these other countries.

I have to say the flag of convenience per se is not necessarily the problem. Some of them are very good, but many are not. How to get around it? There's no doubt about it, the IMO are trying to introduce some form of agreement for wider fitting, and that will take place in due course.

But as I said in the previous answer to a former question, it is the bottom line. If a ship owner whose ships are flagged to one of these states suddenly finds he's not getting the business because he's not meeting the requirements, that will have a greater effect on improving safety than anything else, and the trouble in the maritime world, it is very diverse. It is spread right across the entire globe with different standards applying, and the ship owners in a fiercely competitive world will always go for the cheapest option.

What I would like to see is the people who use ships, either for chartering or take passage in them, actually ask the question, does it come from a flag state which does have a good safety record? And if they would ask that question and listen to the answer, I think that would have a greater effect on some of these nations.

MR. SWEEDLER: Good answer. John, will leak detection systems become a requirement for liquid and natural gas systems? That's probably not a question you can answer, but you might comment on if we're getting to that point or if any companies are using that technology now.

MR. WALLACE: Well, of course, everybody's interested in leak detection. You don't want your product flowing out on the ground for a lot of reasons.

So, all the companies have leak detection programs. Some of them are very sophisticated, and some of them are pretty straightforward. As you might think, the more you spend on leak detection, the finer resolution you would expect to get.

Unfortunately, the more that you spend and the more complex the program becomes, the more sensitive that program becomes to things like an inaccurate transducer or a failure of the transducer to send back good information for whatever reason.

So, sometimes you get to the point of diminishing returns. Some of the most successful companies I've seen have pretty straightforward leak detection programs that are extremely pro-active. They spend their money on testing pipelines and on instrumentation, having accurate instrumentation so they can do line balances which will help them determine if they're losing some kind of product.

Whether or not that's going to be a requirement, though, legally, as you point out, I can't really answer that. I don't know exactly what the penalties can be, but I'm not sure about the regulations right now.

MR. SWEEDLER: Okay. Ed, here's a provocative question for you. Why would the FRA consider an operator's liability concerns when deciding whether or not to record horn operation?

MR. ENGLISH: I don't think that I mentioned operator's liability. I said that it was an issue that the group had discussed on a number of occasions. Certainly it listened to Gary and people from his company made a presentation to the group on what they felt were limitations in recording horns and bells.

We have union members, railroad members in the group as well as members of NTSB. We think it is an issue that is still open, and further discussion needs to be had before a decision can be reached.

We think it's an important issue at FRA, certainly.

MR. SWEEDLER: Here's another one. In FRA's RSAC work, you talked about the crashworthiness standards. Could you just expand a little bit on the impact force and fire standards that are being considered?

MR. ENGLISH: I think if all of you go right next door and look at the NTSB booth, you will see the -

MR. SWEEDLER: Another commercial.

MR. ENGLISH: The draft criteria that the working group has come up with, and as I say, those standards are now in review within the group, and we think we have come up with some pretty good testing criteria.

So, when the session is over, everybody should go next door and look at the NTSB booth, and you will find the criteria listed on the board.

MR. SWEEDLER: Dr. Lehmann, there's a question here about highway recorders. What is considered the most important data elements that are collected on a recording device?

DR. LEHMANN: Well, I have to distinguish between the three different types of equipment I've presented this morning. The tachograph, the most important information there is driving time. You should know that European drivers may not drive longer than nine hours a day, and he has to make a break after 4.5 hours, and the keeping of this rule is recorded on a diagram chart and can be inspected by roadside checks and afterwards by company inspections.

In addition to that, this diagram chart also memorizes the speeds with a very high accuracy of one second. So, this information can be used for speeding enforcement as well as for accident analysis. So, this is what has to be sent to the tachograph.

Coming to the accident data recorder, there the most important information are the longitudinal and transversional acceleration, which is measured by corresponding sensors. Then we memorize the speed which is read out of the tacho-signal, and last but not least, there is a gyro compass in the unit that identifies rotations of the vehicle.

In addition to that, you have 10 inputs for other signals, like ignition on, brake was used, side beam was on, etc. Talking about on-board computers as such, you can identify as many data as you wish. This depends on the individual needs of the driver, respectively, of the operator, whether you use GPS signals, whether you have communication facilities, etc.

But from a mandating point of view, it's speed and time which plays the major role.

MR. SWEEDLER: Good. Thank you. Admiral Lang, in your presentation, you showed a picture of a VDR. The question was saying -- and I noticed that myself, it looked rather large sitting there on the deck, and is that a typical size of a VDR?

ADMIRAL LANG: The question about the size of the VDRs, I think you'll have to wait till Wednesday when Chris Winkley describes the origins of that particular one.

They were in fact fitted in from P&O Ships, and they are rather large, in my opinion, but they are the first generation. Ideally, we do want to have them smaller, but I'm not the technical expert.

But on the other hand, we do have space in a ship. It is not necessarily a criteria. What we need is to get access to it after the event, but perhaps I could suggest you do wait till Wednesday.

MR. SWEEDLER: Okay. Well, I think with that, we've reached the end of our allotted time.