responsible for procuring oceanographic instru- 
ments. However, it is imperative that the specifica- 
tions be reasonable. Current military specifica- 
tions, for instance, would often require 
instruments that are overdesigned or overpriced 
for commercial applications. These specifications 
should provide technical guidance and should not 
restrict or freeze a design. Rather, such specifica- 
tions should simplify communications among seg- 
ments of the oceanic community. 
A large number of independent organizations 
gather and contribute oceanographic data to 
NODC. Often substantially different experimental 
results are submitted because their instruments, or 
sensors, while similar, may have been calibrated to 
different standards, operated in a different man- 
ner, or their output data may have been processed 
differently. 
Some specification guidelines already in exist- 
ence can be applied, with or without modification, 
to ocean instruments. However, the establishment 
of standard ocean instrument specifications is a 
major problem that will require considerable effort 
in man-hours and money over a long period until 
agreement is reached on the acceptability of such 
specifications. Voluntary groups have attempted 
to write specification guidelines, but have gen- 
erally failed due to the enormity of the task. 
The panel recommends establishing a perma- 
nently staffed and adequately funded focal point 
in the Government, preferably in a marine- 
oriented agency, to recommend measurement 
standards, prepare standard specifications, and 
perform tests on oceanographic instruments. Some 
efforts of this kind are in progress in the Navy and 
the Bureau of Standards. 
2. Procurement Philosophy 
Measuring characteristics of the marine environ- 
ment absorbs a major part of the National invest- 
ment in oceanography. The present Federal policy 
for developing and acquiring oceanographic instru- 
mentation appears to minimize initial capital cost 
rather than total data cost. Thus, inadequate 
consideration has been given to the total cost of 
obtaining required data, namely the cost to the 
data collector, processor, and user. It has often 
been demonstrated that a more expensive instru- 
ment, by virtue of its versatility and reliability, can 
effect reductions in the ultimate cost cf data. 
V-48 
Procurement policies have frequently over- 
emphasized initial cost because no dependable 
economic and operational performance criteria 
exist to determine the quality and usefulness of a 
given instrument. In addition, procurement specifi- 
cations often fail to take into account the total 
economic implication of each specification. A 
typical example is the requirement for a much 
higher degree of instrument accuracy than is 
required. 
Under present conditions, instrument manufac- 
turers frequently do not have adequate incentive 
to develop equipment or systems that will be more 
cost-effective to the user. This need not be, since 
industry can produce reliable equipment at costs 
commensurate with high quality. Until procure- 
ment policies are changed, many instruments of 
inherent poor quality will continue to be procured 
on a low-bid basis. 
In summary, many instruments perform poorly 
under operational conditions and thus cause need- 
less repairs and delays. This latter expense is 
growing rapidly due to ever increasing vehicle 
operating costs. In addition, these instruments 
often are not designed to minimize the total 
data cost. 
C. Recommendations 
1. Guidelines 
To foster Government procurement of reliable, 
cost-effective ocean instruments, the panel recom- 
mends: 
—Federal procurement policies should emphasize 
lifetime cost, recommend reasonable performance 
standards, and require complete, adequate quality 
assurance programs by instrument producers. 
—General guidelines should be developed for pre- 
paring technical specifications for instrument pro- 
curement, taking into account each category of 
the specifications listed earlier. 
—The Government should recognize the cost impli- 
cations of particular technical specifications for 
data collectors, data processors, and data users. 
—Continuing, effective communications should be 
established among the data users, procurement 
agencies, and instrument manufacturers, encom- 
