Non-Productive Time Accounting System
From1986 to 1987, I was the Commanding Officer of 1 Base Workshop Bn. At that time, the workshop consisted of around 300 civilian personnel and 150 military personnel. Having just come from an appointment as the person responsible for the development of EME Computerised Maintenance Systems, I had an intense interest in using small systems to a productive purpose in the workshop. With the help and advice of Lt Stevan Vujovic, I learnt how to program in a language called Turbo Pascal and set about developing the means by which an IBM PC could interface via RS232 with the Machine Assisted Workshop Documentation (MAWD) system, used for workshop production control. MAWD ran on a Perkin Elmer 32/20 Defence Standard Minicomputer. The Defence Computing Services Division had long claimed it was not possible to interface PCs with this mini-computer.
I gained approval to purchase an Amstrad PC512 from Colonel Ford; the Chief of Staff of the 3rd Military District(who promptly retired after giving his approval!). In those days, CSD did everything it could to stop the proliferation of small computer systems as they either saw them as distractions to "real computing" or a threat to their empire. (Not much has changed in that regard!!)
Using a library of routines written by Blaise Computing in the US, provided to me by Stevan, I was able to get the Amstrad to communicate with the MAWD minicomputer; pretending to be a dumb terminal. To achieve this, I had to have a special box built that would allow me to "listen" to the codes going back and forth between a MAWD terminal and the MAWD Minicomputer. I succeeded in getting the PC to fool MAWD into thinking it was simply a dumb terminal. The PC, however, was capable of taking data held in text files, generated by other means, and performing standard transactions in an automated fashion on MAWD.
This system was used to raise dummy jobs for each non-productive time activity on all cost centres to monitor/compute non-productive as well as productive time. At the end of each month the system closed off the jobs and collated the results of time booked to them for each of the nonproductive activities. It was then possible to display trends using graphs produced by a spreadsheet program called SuperCalc. The arrangement performed transactions in 20 minutes that would normally have taken a clerk 10 working days to complete. This was and, at the time of writing, still is the first and only time the Australian Defence Force and the Australian Public Service actually quantified to some exactitude the ratio of non-productive to productive time. The results painted a very positive picture of the productivity of workshop staff at a time when it was conventional wisdom to contract out Defence Force logistic support in the interests of lowering costs.