Perceptions of Timeliness or Just not in time
Production Planning Computerisation
In the mid 1980’s I took over the Production Planning and Control Department of a company called Charcon. Charcon manufactured all types of paving and walling products made from both hydraulically pressed concrete and semi-dry pressed concrete. Later on we moved into the Block Paving market when it became popular with all the supermarkets and DIY sheds and hence the general public. At it's peak as a business and after it had been taken over by ECC, English China Clays it was one of the largest manufacturers of this type. The plant where I worked was over 70 acres in size with 21 presses running double shift all year round and round the UK we had another 10 sites, smaller in size.
This leads me to a slight aside but I need to explain how our marketing/manufacturing/sales process worked. With rare exceptions made no products to order but for stock. Taking a flagstone, you need sand, cement, gravel and water and half a dozen guys working the process. The cement is mixed, poured into the mould and then hydraulically pressed to removes excess water. The product was then palletised and left on the shed floor. for 24 hours to harden then stacked to the yard for 28 days to mature to optimum hardness. Concrete does not set by drying but by a chemical process started when the cement is wetted. We ran processing that was optimised between cost and curing time. We could add more cement and increase the manufacturing cost (cement was on average some 70% of the cost of the product ) but reduce the curing time and we could reduce the cement and increase the curing time but cause a bigger buildup in stock.
Anyway, marketing would forecast the next years sales, manufacturing worth then produce the stock to the yard over the Autumn and the winter, sales would then sell it in the spring and the summer, the high points for using our products and the haulage procurement team would haul concrete.
Given this model there was a strong emphasis in the company coming into the selling season to closely monitor the sales and to trim production to meet shortfalls and not to exacerbate overstocking by swapping and trimming production and my department, Production Planning and Control, had the job of acquiring the production and sales information and then performing this balancing act and as part of this we provided a lot of detailed information to the Production Managers so they could make their decisions.
Our computing systems were for the 80's and the 90's quite sophisticated. The company effectively ran on HP Mini computers with everyone logging in via dumb terminals. All the production data was entered to the system, the system then matured the stock (remember the 28 days on the yard) and then when sales sold the product it was removed from the stock system and invoices were raised etc. We then had separate accounting systems and payroll as well as other applications. All the site in the group were connected to the HP's via permanently connected modems, so the whole companies production was accessible as data.
As good as our system was a critical problem was the data was captured on paper and then was fed into the system by clerks. The data for any one day was usually with one to two days depending on the location and so on the Monday of the following week the data would be crunched so that a picture was taken of ours stock vs. sales position. Because of various factors in the way the system was designed this crunching did not usually finish until the Friday of that week
The Production Team would meet on a Monday to make decisions about how well manufacturing had gone and what steps needed to be made to meet sales targets. The trouble was at this time the data they used was up to two weeks old making their life somewhat difficult as in the selling season demand could change quite quickly, even for something as mundane as grey concrete 5x10 half battered.
At some point we acquired some new report writing applications that allowed us to access the input data and effectively write report just for production that could be run after the close of business (22:00 to allow for shift working) on a Monday and be ready by about 11:00 on the following day. This was obviously a great improvement for the Production team as now their oldest data was only 5 working days and the latest data was only one working day old. After testing the report for a few weeks we (the IT team) persuaded the Production team to move their meeting to a Tuesday later in the morning so we could deliver them the report and they could base their deliberations on it.
This went swimmingly for a while but we were always dependent on the batch processing of the data being input in time and the initial batch processes that yielded our data working without error. Needless to say this did not always happen and so some weeks the data was not ready on the Tuesday am but on the Wednesday am.