area of occupational safety and health effects. At that point, they wanted us to further develop the system; this resulted in a patent application that we have applied for and has been granted. Prior to that, we applied back to NASA for a waiver, which they have granted. In doing so, we have provided NIOSH with a 2d generation EPIC unit. They are now using it with NIOSH and we are now working on the fourth unit.

Our next step from that point, after developing the system and working with Mr. Samos at Langley, was: how can we best get this product on the marketplace. I think the point came up earlier, what processes can be used to speed up the maximum utilization of some of these advances in technology. Frankly speaking, I think I have some suggestions and I will write a book when this is all over. But here is a case where we do have the technology, the system which works great, and it now turns out that part of our problem was and still is, (which we are now working with NASA to resolve) to try and get other government agencies interested in the product. It has been proven to have applications in some of the following areas: NIOSH, which is now conducting a study, as previously mentioned. The Department of Transportation, in Boston, bought a unit with which they are now conducting studies to screen out intoxicated drivers; UMTA, which we are now working with and have just submitted a proposal in which EPIC will be utilized to enhance the area of the screening of bus drivers and motormen, to attempt to better our mass transportation systems in which UMTA has an urgent interest; the JFK Rehabilitation Center in Atlanta, Georgia, also has a unit, and they are now doing some studies in the field of affected individuals due to the fact that their limbs have been damaged by having experienced an accident or deformity by birth.

One other area I would just mention very briefly, is how well the transfer process has worked with JWM on this particular item. By all of the individuals at NASA being so cooperative, I have yet to come to them with a problem involving this TU program to which they have not responded. It has worked well. Again, I would like to publicly congratulate this TU team. Without this TU team, I don't think it would have worked.

Some of the applications of this particular item are found in this brochure that we have attached with the testimony. We have mailed out approximately 1,000 of these brochures. Within the last five months, we have gotten over 3,000 inquiries on the product. The stage we are in right now, with it, is in developing the norms and standards. Once the unit is bought or acquired, they will have the proper values to compare.

Mr. HECHLER. Thank you, Mr. Williams. This is a good illustration of something that sounds pretty far out when you first thing about it. It has practical applications. I don't want to be facetious when I observe that perhaps it would be applied, not only to what you are planning for, but also to get the reactions of bureaucrats and legislators to problems to see how quickly they react.

Do you have any questions, Mr. Wells?

Mr. WELLS. How far away from a commercial marketing program do you think you are, Mr. Williams? You have 3,000 inquiries and it seems implicit in your comments that there are still some problems.

Are these financial or technical or marketing? How would you characterize them?

Mr. WILLIAMS. I think I would be less than fair if I don't say both: financial and commercial. However, the product as a piece of hardware is completed. With the product design as is now shown in that picture and the brochure, it has been proven to meet the requirements of the applications for which it has been designed. In fact, these design features are the result of many, many hours of study, talking to hospital staff and personnel at other institutions. So we think we will have markets for this device. It is a result of having worked with them and coming up with final features such that it can be used effectively.

We thought we were ready to go. Really, we made the brochure and mounted a sales campaign, but we later discovered that we had a device where there is no previous product like it, number one. Secondly, we are dealing in a field where people are not really sure. Everybody would like to know about their dexterity and coordination, specifically in the area of screening applicants for jobs, screening bus drivers, what makes a good motorman, etc. We have intended from a marketing point of view to have worked up a good strategy to go out and sell the device, set up tests around an individual's specific requirement. Unfortunately, in dealing with an area of a new technology and trying to educate the public, at the same time giving them something they can plug in and get a result, we found that is difficult. Our next step we found out is that we have to develop the norms and standards-the software package-and that is an area that we are now dealing with. So we pulled back all previous marketing strategy approaches and said, let's get the norms first. That will be another four months and then we can go back out with the new marketing strategy. Instead of just selling hardware, we will sell with the hardware with a software program as well, which we had not initially intended to do.

Mr. WELLS. Is the Small Business Administration cooperating with your company in terms of financial assistance and cooperation and supplementing the NASA technological assistance?

Mr. WILLIAMS. No, they are not. I think I should explain that. My initial contact with NASA was through SBA. After that point, you are now dealing in an area basically that is the development of a technological design. I brought my problems in this area to NASA and concluded that most of the programs that SBA had dealt with were in the area of loans or in the area called contract assistance. This is a design and development type project. In order to get financing from SBA, it would have to be in the form of a loan. In fact, it is a kind of a project, unfortunately, where it really requires a private company with a big organization with some government group behind it, other than NASA, so that the right publicity can get out. All of the potential customers to whom we spoke before trying to get this out, had the same question: Does X organization endorse it: In the area of medicine, does the AMA or HEW; in other fields, the mass transit industry, we are asked how does UMTA feel about it. They seem to go back to these organizations. For that reason, we then tried to take a marketing approach to deal with government agencies, making proposals for studies or a grant for funding of research, and that is where we are today. Mr. WELLS. Thank you.

Mr. HECHLER. Thank you, Mr. Williams.

Mr. GRAY. One area in which we feel we have made a significant gain is reflected in our joint programs of technology transfer to problems of other Federal Government agencies. I mentioned some of these in my opening remarks and I have Dr. Karl Frank of the Federal Highway Administration to present to you his experience with our joint program to apply the NASA-developed computer analysis system called RANDOMDEC.

Dr. Frank received his undergraduate degree from the University of California at Davis, his Master and Ph.D. degree in Civil Engineering at Lehigh University. Currently Dr. Frank is Structural Research Engineer with the Federal Highway Administration where he is assigned as the contract manager of the study using the RANDOMDEC technique for the detection of structural deteriorations in bridges.

STATEMENT OF DR. KARL H. FRANK, FEDERAL HIGHWAY

ADMINISTRATION

Dr. FRANK. Thank you, Mr. Gray.

I am going to deviate from my written text and use some slides to illustrate the points of this research program. The problem that this research program is addressed to is inspecting bridges for structural deterioration.

Up to 1968, with the passage of the Federal Highway Act and the assignment of the Federal Highway Administration in developing bridge inspection methods and bridge inspection guides for determining the structural integrity of bridges, most bridge inspection was done by visual means. We were looking mainly at how good the paint was. We are looking now, in our research program, at various means of improving structural inspection of bridges.

As an example of what we are looking for, this is a large crack at the end of a plate in a bridge. Through this technology, we hope to develop means for finding cracks much smaller than this so that they can be repaired and the safety of the structure ensured. This is typical of some of the locations in bridges where we have looked for cracking. This is an old bridge. In fact, you can't see the crack, but it is pointed

out.

Our problem in bridges is that they are inaccessible. We have problems with weather and ice and rain. In many cases, the people who are competent to go up and look at the bridges are not competent to inspect them. So we are looking for means of monitoring bridges remotely, looking for signs of structural deterioration. The NASA-developed technique of the analysis of vibrations, called RANDOMDEC, hopefully will provide us with this method. I want to focus on the fact that we are looking at structural cracking caused by fatigue, which is mainly due to truck passages. As you know, there is a trend to increase truck vehicle weight and this problem becomes more acute.

The present means of inspection that we have developed in our research program requires going through the bridge and placing an instrument or a probe on the bridge to detect cracking, as shown here. However, many bridges are inaccessible. If you consider a multi-span bridge with many components, to go through that complete bridge would require a large amount of work. The floor detection problem, as defined by the RANDOMDEC technique, is that we take a random

force input such as the passage of a vehicle over the bridge. We develop, through the bridge structure, a typical response which we monitor through an accelerometer or a strain gauge and amplify it. This noise is filtered and put through a NASA-patented RANDOMDEC computer, with the resulting signature.

In order to evaluate this system for application to bridges, a study was developed in which a NASA contractor, Neilson Engineering Research, in California, did the analysis of data generated here at the Federal Highway Administration Lab in a controlled laboratory experiment. This was to demonstrate the detectability of the system: that is, how small a crack can we find; how far away from the crack you have to be with your transducer. This diagram shows a schematic of the test speed and the random noise generator that we use to generate the right noise in the beam. It was equivalent to a hi-fi speaker; that is, a very low amplitude, with high frequency vibrations. The data was generated here at the laboratory and then sent to the Neilson Company and they used the equipment at NASA Ames to analyze this data. We sent them data without telling them when there was cracking and what size crack. This was a blindfold experiment to make sure that we weren't slanting the results, based on what we observed as far as cracking. We left it up to the contractors to tell us when they thought cracking was occurring. They were able to tell us when a half-in crack was there. We felt it was more than adequate for our needs.

This is a typical cracked beam. It is not very clear in this slide, but we have a crack in the lower part of that beam, underneath the strain gauge. This is a typical RANDOMDEC signature. As you can see, it is a wiggly line and there is a rather complicated procedure by which they develop this. The point of this line is to show that at the top we have the standard, the fingerprint of the beam, and as we completely cycle it, that signature changes and that is what we are looking for: changes in that signature which will signify changes in the structural behavior of the member.

We also found in our laboratory experiments, much to our surprise, that we could be fairly remote from the crack and still detect it. This has great significance, in that we can cover a larger area of the structure with a single transducer.

The second part of the experiment which we performed was to instrument an actual bridge in California. Since we are talking about coordination, particularly the coordination between the different bureaucracies, the involved coordination between NASA, the Federal Highway Administration and the California Department of Transportation, the owner of the bridge. The bridge was instrumented in various locations to determine how these RANDOMDEC signatures varied with different weather conditions and also to determine what types of responses we could expect in a typical highway bridge. This is just an overall shot of the bridge and it is being excited right now by a cement truck which provided adequate field excitation. To give you an idea of what we are talking about, in terms of transducer size, just below that arrow made by the tape is a small accelerometer which was used on the bridge. These were placed on various locations of the bridge, monitored about once a month, and then taken back and analyzed.

This shows the data at various times during the year of a signature at a particular location. The significant point of this slide is that the signature didn't change; we didn't have any structural deterioration and it was stable during these different weather conditions.

In summary, in the results to date, we found that signatures were sensitive to small cracks; that the location of the transducer was not critical; that the ambient conditions provide adequate excitation; and that field measurements were stable. We hope to continue this to provide a useable inspection technique for certain specific structural applications. We will be evaluating two bridges in San Francisco in the next part of the contract. I have been quite happy with the cooperation we have had with NASA in this joint project. In fact, as far as the red tape involved in any bureaucracy, that has been more on the Federal Highway Administration's side than it has been on NASA's side. It really has been a very smooth project.

Thank you.

Mr. HECHLER. What about the state highway departments?

Dr. FRANK. They have been very interested. We have had no trouble at all in getting their help. In fact, just last week, the contractors had discussions again with the State highway department about this bridge and our desire to loosen some bolts on this structure to find out what that does. They have given us the okay to do this, which is something you don't normally get from a highway department. They have been quite helpful, really looking towards some useful results.

Mr. HECHLER. Coming from an area in West Virginia, where the collapse of the Silver Bridge in Point Pleasant, W. Va., caused a heavy loss of life several years ago, I am particularly impressed and interested in RANDOMDEČ and what it can achieve. You mentioned something about the effect of heavier weights of trucks. There has been a fight within the Congress as to whether or not to authorize wider, heavier and longer trucks. In the name of energy, I understand that this may well be slipped through without legislation within the bureaucracy itself. I wonder if you would have an opportunity to utilize your RANDOMDEC to assess the effects of longer, wider and heavier trucks on some of the existing bridges. It would seem to me that your input would be very vaulable to enable us to determine whether or not the existing bridges can sustain that kind of additional weight, length and width increases that are now being proposed.

Dr. FRANK. Very definitely, it has applications in this area. It will show that we can go out on an existing bridge and find out if there is a deterioration over a time, with the increased truck rates. What it will do is to accelerate this. It accentuates the need for this type of thing. Mr. HECHLER. It is later that you think. It is going to happen before you are able to get your RANDOMDEC out.

Dr. FRANK. I realize that and I would rather not comment on it. Mr. HECHLER. This is a good forum to comment.

Dr. FRANK. I think the small loads, as I understand it, that are being discussed would not do injurious damage, but they are much smaller than what was originally proposed, I understand.

Mr. HECHLER. Thank you.

Mr. WELLS. I would like to ask Dr. Frank one question. Dr. Frank, what is the cost of installing a set of transducers on a bridge? Is there some average as to how many transducers per bridge or some other relationship?