Simulation – a new tool to optimise production

Simulation is increasingly being used to speed up and reduce costs in product development. In her Master’s thesis, Afroz Mirzaie Shra has demonstrated the potential for similar benefits in production.

“Using simulation in production is only in its infancy and we still have a long way to go before that alone can form the basis for taking decisions,” says Afroz Mirzaie Shra, a Master’s student at Mälardalens University.

“At present, the data at our disposal is not fully reliable. But similar analyses can indicate where measures might be undertaken to achieve operational improvements and are, as such, already valuable.”

Simulation identifies bottlenecks in production

The focus of Shra’s research was the cylinder block line at the engine machining workshop at Scania in Södertälje. The line comprises 38 production machinery units, of which two are robot cells. The ideal production tempo is six minutes between the 19 operational process steps that are involved in manufacturing the engine block.

The line has lately faced difficulties in upholding its 85-percent operational efficiency target and Shra undertook an extensive database simulation to determine where the bottlenecks occurred and identify those that were easiest to remedy. She found that one particular fine machining unit often went over the allotted six minutes.

Shra introduced a number of parameters to ascertain whether these would affect the rate of output. These included adding an additional operator, reducing tool change time and frequency by ten percent, increasing the buffer stock, solving calibration issues and addressing unclassified disturbances. The data analysis showed that getting to grips with calibration and unclassified disturbances had the greatest impact on operational efficiency.

A useful production analysis

“This analysis has been very useful,” says Anders Westerberg, Cylinder Block Workshop Manager at Scania. “Experience tells us where we may have shortcomings, but this helps confirm guesswork and thereby provides basis for mutual prioritisation within the organisation. We can also test the results of different countermeasures that we could introduce to solve the problems.”

However, Westerberg adds that better input data is needed to achieve greater reliability in simulation analyses. “The manufacturing machine will electronically report a stop but not what causes the stop. That must be manually submitted by the operator. When we see how that information helps us in our daily work, there will be an increased incentive to report stoppages.”