The engineer carrying out a materials survey for a road uses his judgement to locate sampling sites and make observations of the terrain that are relevant to the design of the road. The existence of an engineering data store would enable him to use the experience of his colleagues elsewhere to support his own judgements. Although the actual body of engineering data held in store may be very large, an index designed to locate any item of information need only be relatively simple. Punched cards can be used to provide a simple and efficient index to engineering data, the design of the index depending upon the nature of the material to be stored, and the demands of the user. This report discusses the relative merits of edge-punched cards and optical coincidence ("feature") cards, according to the circumstances of use. If the data are fairly compact, e.g. largely numerical data, then they can be stored on edge-punched cards and rapidly retrieved, although the index is necessarily rather inflexible. If a large volume of miscellaneous, often bulky, data is to be stored, then optical coincidence cards, with their highly versatile means of indexing, are appropriate, although retrieval times may be slower. It is argued that both systems are desirable for the efficient retrieval of engineering data, and a design for an index and storage library is put forward prior to field trials in Ethiopia, where a national data store for engineers is expected to be set up. The structure and operation of the data store are described, particular attention being given to the integration of data acquisition in the field with the processes of indexing and storage in the library, to minimise the work of abstraction and administration. (Author/publisher).
Abstract