Will computing reduce mountains to molehills?
Scientific Computing World is published from Cambridge, England, an area of the country notorious for the flatness of its landscape, without the merest vestige of a hill. There is a modest irony, therefore, that the common theme of the articles in this issue of the magazine is 'mountains'.
In the quality assurance laboratory, scientists are in danger of being buried in mountains of paper. One of the greatest promises of Laboratory Information Management Systems is that, by putting everything on a LIMS, the paper mountain will be reduced to manageable proportions. The challenges facing the industry are to make such systems secure and user-friendly at the same time.
In the discovery laboratory, researchers have to cope with a data mountain. Partly this is the result of advances in instrumentation: it is now possible to put so many strain gauges on an aircraft wing under test, for example, that the weight of the cabling connecting the sensors itself influences the test rig. Virtual instrumentation - the use of PC-based software interfaced to the sensors - offers far more power than traditional hard-wired instruments. Indeed, the claim made in our article on virtual instrumentation is that the power of the PC-based instrument may bring us back to the good old days when scientific research was carried out by small teams. Traditionally, we think that small teams equate to higher scientific creativity - always a good thing.
In the life sciences laboratory, it is striking how many LIMS players are moving into bioinformatics in the belief that they can contribute to coping with the biological data mountain. LIMS have always played an important role in the pharmaceutical industry, but mainly downstream at later stages in the manufacturing process. Now LIMS are seeking a role in the drug discovery laboratory. It will be very interesting to see the mutual effect: how ingrained laboratory habits change as the documentation process is computerised; and how LIMS evolve to meet the requirements of these researchers.
Data, whether gathered by virtual or hard-wired instruments, in life sciences labs or engineering sheds or out in the field, needs to be analysed as well as managed. Maps are indispensable, but they are not enough. Someone needs to go out and climb the mountain to look at the view from the top.
Appropriately, the data visualisation software review comes from Colorado, in the shadow of the Rocky Mountains. And, as Felix Grant argues in his review, the pre-occupation with the visual appearance of our natural environment evinced in this leader is not misplaced. The way we interpret data is inescapably bound up with our biological and evolutionary history. In words that every manufacturer of scientific software (and not just visualisation products) should learn by rote, he maintains: 'Joy is not to be sneezed at; aesthetic response to visualisations is an important factor in functional effectiveness, in generator as well as perceiver.'
Once all the data has been analysed and the scientific paper is being written, researchers need to cite previous work (having, of course, searched the literature at the outset to ensure that they were not duplicating anyone else's work. Here too there is an information mountain. The techniques to conquer are set out in the autumn issue of Research Information. From the flatness of the Fens, happy mountaineering!
Dr Tom Wilkie