A problem solved ...

Press, 25 May 1977, Page 10

A problem solved ...

New Zealand Aluminium Smelters, Ltd, has virtually solved the long standing problem of making production staff at. factory floor level feel at home with data processing.

A feature of the aluminium industry is that the meta! is smelted in a large number of small “pots” — in the case of N.Z.A.S. there are 408. A number of operating variables have to be measured on each pot each day or each shift in order to determine what each pot is doing and what the group of pots is doing. This data was previously processed by hand resulting in a brief daily report and a more detailed weekly report. Often this information arrived too late for adequate corrective action to be taken. All this data is now col-

lected in a form which can be read by an 1.8. M. Optical Mark Reader, a piece of equipment which detects the presence of pencil marks on special stationery.

And the pencilled indications are just “marks” inserted on a “Laundry List” of pre-printed performance characteristics (D.M.R. forms). These D.M.R. forms are filled in on a shift basis and then placed into the Optical Mark Reader which transfers the interpreted pencil marks on to a magnetic tape for processing by the computer.

Following analysis of this data the computer system assists in pin pointing a pending of existing problem in a pot, and the nature of that problem.

As far as John Rogers, Data Processing Superin-

tendent at N.Z.A.S., is concerned, the big breakthrough is in reducing the gap between process computing and batch computing in two ways, viz., by minimising turnaround time between capture and processing, and being able to process data which cannot be measured automatically. Reducing this gap has meant that pot line performance reports are rapidly made available. Thanks to the D.M.R. the turn-around time between data collection and reporting back to the user — which is sometimes measured in days — has been cut to less than one hour. This means that shift operating staff are aware of the current state of performance and repair of each one of the 408 pots. “We have eliminated the transcription stage," notes

Mr Rogers. “We want those reports back in time for every shift change.” It is worthwhile to look at the smelter’s use of data processing in relation to manufacturing. It is, in fact, a well defined policy that investment in data processing systems development is all cost justified and that top priority is given to projects with the highest payback. This explains why relatively little data processing system support has been developed so far for the more traditional roles of D.P. as accounts, budgeting, etc. A comprehensive payroll system is an exception to this, but generally it is noticeable that data processing support is more closely aligned with production departments than service departments.