Subject: Conversations with Col. Hensley, Dr. John B. Walsh, and Vice Admiral Eli T. Reich at the Pentagon on October 17,
The attached list of background items and questions was delivered a day early to Dr. Walsh. This made it possible for him
to do nearly all of the talking during the 40 minutes of our interview without losing time making complicated explanations.
Dr. Walsh is a Ph.D. electrical engineer, who is in charge of all U.S. military satellites. His message concerning systems
engineering was essentially as follows.
Systems engineering works satisfactorily only if the basic technology is already known. Occasionally in the course of projects
with which he has been concerned, basic investigation has been necessary. It was his reaction that if new knowledge were uncovered
in this fashion, it was ordinarily a case of "lucking out." It was his estimate that when basic investigation is
undertaken in a program which is run on a systems engineering basis, not more than ten or at the very most 20% of the undertakings
He pointed out that the costs of guiding a large project on the basis of systems engineering and the magnitude of the team
which must be assembled to do so are ordinarily much too large for universities to handle. Systems engineering is therefore
to be found almost exclusively in large corporations in industry. Sometimes in the course of work being done in this fashion
by industry it is possible to break out pieces or areas which can be sublet to universities for the purpose of pursuing basic
In the program which is being handled on a systems engineering basis, it is not possible commonly for any one individual to
grasp and to understand all of the details. The top level people, however, must understand all of the pieces in a somewhat
general way well enough to be able to coordinate them.
It is common for the programs to be broken into pieces. Sometimes the manner
in which this is done is rather arbitrary until experience is gained. When projects are broken into pieces for easier grasping
and management, the major problem which arises is the interfaces between pieces. He used as an example one type of rocket
which had an analog computer in it. In the course of development another was made which had a digital computer in it. In order
to make these two devices work together it was necessary to have a DA convertor in one. Finally, both came to have digital
computers and in order to have interchange, both were equipped with DA convertors until the process came along far enough
so that they could have it all digital. This sort of problem becomes especially difficult when the programs get very large.
In regard to the development of the artificial heart, it was Dr. Walsh's opinion that an effort to use a systems engineering
approach could hardly fail to result in disaster simply because so much basic investigation must be done which is not susceptible
to successful attack by systems engineering.
DOD has problems with bureaucracy and personnel ceilings which Dr. Walsh
thought were probably very similar to those at NIH. One mechanism which is employed to get around this difficulty of shortage
of personnel and inadequacy of expertise was to seek a contractor who would use systems engineering and provide technical
direction. Sometimes this technical direction is given directly to the contractors by the systems engineering concern and
sometimes in the interest of maintaining control as well as possible the systems engineering concern gives advice to the military
which then passes it along to consultants. This must be done on a continuing basis. TRW did this in developing the Minuteman
rocket. Aerospace has done the same thing for DOD. The advantages of this arrangement are that these concerns can pay sufficient
salaries to topnotch people who will do it right. He pointed out that the FCRCs (Federal Contract Research Centers) are on
a guaranteed level of effort, and sometimes there is criticism of the arrangement because it looks to some like simply a way
of getting around the Civil Service stipulations. The fear apparently is that top level people who have high salaries also
are likely in government to have a great deal of power, and Dr. Walsh feels that this is a bad combination to give to individuals.
In the case of the FCRCs, however, these people do not have much power inasmuch as the contracts can be terminated at any
time at the pleasure of the government.
In the course of development of the Minuteman, at one time the Air Force turned the entire development over to the TRW. TRW
then hired the necessary contractors and individuals to do the job. The tendency more recently has been to go back to utilization
of TRW more as a continuously monitoring consultant with the actual giving of directions to contractors in the hands of members
of the military.
At the present time the Army is using what are called technical assistance contracts in which experts from the contractor
monitor carefully what is being done by the subcontractors or other contractors and guide the Army in making decisions as
In regard to the NHLI and having committees try to run big programs, Dr. Walsh has found that this is a road to fiasco. One
such committee has been trying to run a large program in DOD. He did not feel that he was free to tell me what it was, but
the running of the program has become just such a fiasco and it is even worse because the committee doesn't seem to have
enough understanding to realize that it is a fiasco. His conclusion is that committees simply are not under any circumstances
capable of running a systems engineering type of developmental work.
In view of the many problems that still exist in the effort to make an artificial heart, Dr. Walsh feels that this is probably
still an area for the lone inventor or at least the independent investigator.
Dr. Walsh suggested repeatedly that it would be very fruitful for NHLI to get a group of top level systems engineers together
to review the patterns of management of the program to develop an artificial heart. He suggested not more than six individuals
and he thought that the appeal of trying to do something that would help mankind would perhaps offset the difficulties created
by the smallness of the per diems offered by NIH.
Dr. Walsh made the following suggestions of individuals who might be appropriate top level systems engineers to have in such
a study group. It will be noted that all of them come from industry. The reason is that no university is able to field a project
running on a systems engineering basis.
Dr. Robert Burnette, TRW
De Lauer, (high officer in TRW, possible president)
Oliver Boileau, Boeing
Howard Burnette, Lockheed
Dr. Seymour Seiberg, Research and Development Associates on the West Coast
Robert Quade, Rand Corporation (There is some question about him since he is not a systems engineer, but rather a systems
analyst. Dr. Walsh thought he would be extremely valuable or of no value at all, but nowhere between for some reason.)
Ivan Getting, President Aerospace Corporation
He thought someone from Bell Laboratories might very well be asked to come. The selection of a person from Bell might be put
up to Brockway MacMillan, the vice president. He incidentally would be very much interested since he has just recovered from
a rather severe myocardial infarction.
Col. Hensley and I stopped in to see Vice Admiral Eli T. Reich. Admiral Reich has had a great deal to do with systems engineering
from the point of view of the military rather than from the point of view of the industrial people such as Dr. Walsh. In general,
Admiral Reich appeared to concur in the messages which I had gotten from Dr. Walsh. Admiral Reich also drove home the point
as Hensely and Walsh had done that systems engineering is not a fruitful pattern of management of basic investigation. (Admiral
Reich was the commander of the submarine which sank the Japanese aircraft carrier Yamamoto during World War II.)
Admiral Reich suggested that it might be very fruitful to talk with Mr. John Riordan, a civilian GS 17 who has worked in the
Pentagon for more than 20 years. Apparently he has a very great knack for the systems engineering approach. He serves now
solely as a consultant for which purpose the DOD provides him with an office in the Pentagon. Col. Hensley will set this up
With regard to the responses to the three questions on the attached sheet which was provided ahead of time to Dt. Walsh, the
answer to question No. 1 as to whether it was feasible to bring truly top quality expertise in depth from outside NHLI into
the day-to-day operation of the program is "yes," but he thought it was irrelevant. With regard to question No. 2,
this appears to have been the subject of all of the discussion with Dr. Walsh. With regard to question No. 3, "Can a prime
contractor take over the whole program making all subcontractors responsible to him" the answer was "no" with
the addition that it would be perfectly all right and easily feasible if all of the technology were already in hand. Since
the technology is not already in hand, this is not a fruitful course to take.
Background and questions to place before John S. Walsh, Ph.D., Deputy Director of Defense Research and Engineering (Strategic
and Space Systems)
1. The program on the artificial heart at NHLI is small by the standards of DOD (about $10 million in contracts).
2. The AH program at NHLI is run primarily by contracts rather than grants, the thesis having been that the work of multiple
contractors can be coordinated and directed from the NHLI.
3. There is insufficient expertise in depth in the staff of NHLI to cover developments properly.
4. Expert task forces have been gathered to review initial proposals and proposals for renewal, but they can meet only periodically
and cannot give continuing guidance. Furthermore, In-house decisions can override the experts.
5. It appears probable that NHLI has failed to develop contracts with the strongest investigators available. This may well
be related to the quality of monitoring which NHLI is able to provide.
6. NHLI cannot acquire appropriate personnel in proper depth or even fully use present personnel because of several factors:
a) Personnel ceilings imposed by the Administration,
b) Proscription of performance of laboratory work on a continuing basis by members of the staff dealing with extramural affairs
of NHLI, either in the laboratories of contractors or in laboratories set up in the neighborhood of NHLI, a fatal obstacle
c) Restrictive ceilings on salaries of high-level personnel,
e) Widely spread geographical distribution of the contractors,
and, finally --
f) Restrictive ceilings on the travel of the staff of NHLI imposed by the Administration, which limits the effectiveness of
even this staff in provision of properly close monitoring of activities in the laboratories of contractors.
7. The above factors have contributed some distressing extravagances and have been costly also in retardation of sound progress.
1. Is it in your view feasible to bring truly top quality expertise in depth from outside NHLI into the day-to-day operation
of the program?
2. If so, in the light of your own experience in DOD, how would you advise that we go about it?
3. Can a prime contractor take over the whole program, making all other subcontractors responsible to him?