Questions and Answers from Session IV

DR. ELLINGSTAD: We do have a few minutes to take a couple of questions here. And if I can sort them out, we'll keep John talking, if you want to switch your microphone on.

We have a question here. You mentioned that these are GM equipment. Is there an interest in installing these devices from other manufacturers? And is this likely to be available as an after market item?

MR. HINCH: I think that other manufacturers do have interest. It's demonstrated in the fact that they have joined the event data recorder working group. I don't believe -- at least to the best of my knowledge, that other manufacturers are as far along as GM is in building event data recorders and making the data available to the public, but I really believe other OEs are interested and probably will in the near future have event data available.

After market. I think there are a lot of companies that build after market equipment. The problem with after market equipment is it doesn't have the penetration into the vehicle market. There are over 200 million vehicles in America and the chances of us getting very good penetration of those 200 million vehicles from after market is very limited.

DR. ELLINGSTAD: Thank you.

A question for Chris Winkley regarding the proprietary issues with respect to voyage data recorders. And I suppose as the manufacturer of those recorders you may have some perspective on that. But are there issues of standardization and proprietary versus a more open standard for the data that are recorded on voyage data recorders?

MR. WINKLEY: I think if I read the question correctly, the answer I'll give to you is that's the whole purpose, really, for the IEC and it's creation of the technical standard which is currently in its draft form.

Remember, the IMO is a performance standard. It tells you what it's got to do. And the IEC standard tells you how it's going to go and do it.

So I would imagine that in January 2000, assuming that the IEC technical specification is cleared, then there will be a whole myriad of manufacturers who will then be designing to a standard blueprint.

Naturally, anything which the IMO produces is the absolute minimum standard and then it will be up to the impetus of the various manufacturers to make their product better than somebody else's, probably by adding value added engineering or adding different parameters to it.

DR. ELLINGSTAD: Thank you.

I have a question that I'm going to twist around so that Mike can answer it.

Video recording of airline cockpits is not part of current FAA rulemaking and rulemaking should be good for 10 years.

I'm not sure who's conclusion that is.

When it is envisioned that the NTSB will recommend cockpit video recording.

Well, we'll twist that a little bit and ask Mike if that technology is ripe for rulemaking yet and if there are any accident investigation authorities that have indicated a disposition to require or to mandate video recording.

MR. HORNE: Yes. The technology basically is available. The exact use of the technology is under discussion through EUROCAE Workgroup 50. There are several members of that committee here. They are supposed to report through 2000.

There's a recent NTSB recommendation which mentioned video recording and looking towards a requirement for video recording out to 2005. But the Accident Investigation Branch in the UK has been an advocate of video recording for some time. The NTSB, similarly, with various pronouncements from various accident reports.

So, yes, the movement is towards use of video. It is a quietly, quietly approach. Obviously there are big issues with privacy, especially with something where an airline pilot -- it may be the last few seconds of his life that we're recording here and there are big issues to be addressed. I wouldn't like to trivialize that in any way.

DR. ELLINGSTAD: Thank you.

MR. HORNE: Somebody had their hand up over there.

AUDIENCE QUESTION: (Off mike.)

DR. ELLINGSTAD: Okay. I think with respect to that particular recommendation, it's true that we had not yet recommended video recording. I think there was in the background of the recommendations coming out of Swiss Air, some language to the effect that the technology ought to consider requirements that are likely to come down the road in the future.

Turning to Dan and the pipeline industry, this question. It would appear that with advances in technology to acquire data more quickly, that gas transmission companies could use the additional data to improve the line balance calculations to detect leaks sooner.

What is the impediment to this advance occurring?

MR. NAGALA: In the gas pipeline industry, of course, data acquisition rates are important and communication technology certain supports much quicker access of data, assuming that we're not dealing with multi-access radios that have low bandwidth, which is common in a lot of gas pipeline applications.

With regard to detecting leaks in a gas pipeline system via a software analysis method, a real time model as it were, the problems we encounter are really physics, problems of physics.

If you have 100 miles of pipeline and there are 100 miles between pressure measurements and you have a leak at the mid point between those two measurements, it's going to take some time for the effects of the leak to propagate in both directions out to the measurement points.

So if you don't have measurement at every point along the pipeline, your simulation system isn't going to be able to recognize the signature of the leak and be able to deal with it and make the appropriate enunciations in a reasonable period of time.

This has always been the challenge with gas, because gas being a highly compressible product does not propagate those pressure waves very rapidly through the fluid.

In contrast to the liquids industry with that same example, if you've got columnar flow between your pressure measurements and you have a leak at the mid point, the wave front that's created as a result of the leak is going to propagate in both directions at approximately 2500 feet per second. So if it propagates 50 miles, you're talking about 100 seconds before you start to see the effects of the leak and then your leak detection system has to observe those effects for some period of time before it can make a determination. That might lead to maybe two or three minutes depending on the magnitude of the leak.

Of course, if it's a full line break, you're going to see it much quicker than that.

In gas, it's the same way. Typically, the leaks are not full line breaks. Typically, the leaks are third party damage or maybe some kind of corrosion perforation. And the propagation of the pressure indicators that are key components of the leak detection system just take time.

In a gas pipeline system, they could take hours or at least tens of minutes to propagate to the appropriate measurement points.

Now, of course you might wish that wherever you're going to have a leak you had an instrument right there. But unfortunately, that's never the case. So, in gas, it's a very difficult issue. I know a lot of people have been trying to struggle with this but we're still bound by the laws of physics. And until we have more direct observation techniques that are economically feasible to deploy along the pipeline -- and I've heard some talk about fiber mesh. There are some direct burial cable technologies as well. But these are very expensive.

Certainly if you have fiber mesh and you have a third party damage of the pipeline, the fiber mesh is going to give you fairly rapid indicators. But that's not a hydraulic simulation in modeling sort of techniques.

DR. ELLINGSTAD: Thank you, Dan.