In comparison, analysis sometimes feels hard work and slow. Sadly intuition can lead us astray when we encounter a complex problem that looks simple. Here’s an account where simulation analysis uncovered an important blind spot in the ceramics sector.
Managing a ceramic manufacturing process is interesting, because invariably there is a tunnel kiln that operates 24/7, and a make process that only works two shifts per day, Monday to Friday. This sets up a complex storage problem that is irregular and difficult to analyse.
For example, on a Monday the make process needs to produce enough to operate the kiln on the Monday shift, and through the night to Tuesday. In addition, however, it also needs to start creating a buffer of product for firing over the forthcoming weekend. The same is true for Tuesday and so on. On a Friday, the size of the unfired product buffer peaks in preparation to feed the kiln from Friday 6pm through to Monday 6am.
Unfired product is stored on kiln cars. These cars are stored in lanes that feed the kiln. The problem is made more complex because at the start of each day there is even more fired product that has emerged from the kiln overnight. Invariably the manufacturing operation is in a confined space, and so fired and unfired cars have to be stored in the same lanes.
So when the company we were working with regularly ran out of product late on Sunday, intuition said they didn’t have enough kiln cars. And so they purchased new kiln cars. However, strangely this made the situation worse.
The simulation analysis told a different story … it revealed that increasing the number of kiln cars created more congestion. This in turn trapped more unfired kiln cars behind fired kiln cars, and that was the real reason for the lack of supply. And so the real solution lay in removing, not adding, kiln cars – and that was a very brave decision because at the time it was so counter-intuitive. However, soon afterwards I turned up to see unused kiln cars stored in the car park outside!
