ECONOMICS 



The economics of computers used in ocean research is very 

 complex. One has to match the increased cost of technology 

 against the nebulous notion of progress in science. Suppose we 

 break up a scientific experiment into three parts: the planning 

 phase, the experimental phase and the analysis phase. Before 

 the seagoing computer, one planned an experiment, set up a cruise 

 plan, took the data and then returned to the laboratory for the 

 analysis phase. Now it is possible to plan an experiment, plan 

 a cruise and then undergo cycles of experiment and analysis on 

 the same cruise. 



The computer is used in all three phases of the experiment. 

 In the planning phase the computer is used for modelling and 

 prediction and the analysis methods are developed in advance. 

 Experiment modelling can save ship time by eliminating some of 

 the cut and try. The second phase operation requires the 

 computer to be used in on-line or off-line processing or data 

 acquisition. The third phase analysis confirms or modifies the 

 results of the planning phase and in certain experiments one can 

 make changes in the experiment and go through phases two and 

 three again. 



CONCLUSIONS 



From a cost effective basis it appears that a carefully 

 planned medium-sized computer can accomplish most of the tasks 

 undertaken by ocean scientists. However, such a system should be 

 attached to the scientific laboratories and not to the research 

 ships. This last statement is justified by an observation of 

 long standing that bureaucracies carefully engineer creativity 

 out of scientists by the imposition of simplistic rules and 

 arbitrary regulations. 



The philosophy of a shipboard computer must be that of service 

 to the scientist and not control of the scientist,, In this light 

 dedicated computers have the edge since they most directly 

 connect scientist and computer. Medium-sized computers can 

 generate this connection provided they are not assigned too many 

 permanent tasks. The computer center facility is inevitably 

 hopelessly far away from the scientist since the computer entity 

 is larger than most experiments. The fear I have of the computer 

 center and possibly the medium-sized systems is that they will be 

 used primarily for data gathering and graph plotting. The ability 

 to print thousands of pages and plot hundreds of graphs at sea 

 can be a millstone to science if improperly usedo Computer 

 technology changes rapidly and we must be prepared for this 

 change. We must not burden our research ships with any more in- 

 flexible, often inoperable and often out-of-date equipment. If 



40 



