Development of a Pull Production Control Method in the Metallurgical Industry

The comparison was conducted with the following characteristics:

• Demand created within the model based on statistical distributions that are equal in both models.
• The following demand scenarios were performed are shown below in Table 3:
• Results are written in a "KPI (Key Performance Indicator)" Excel file and then compared between the models.

Table 3. Demand scenarios.

No.	Scenario	Description	Name
1	Scenario 1	Demand in an average level around 30 units per week	Base
2	Scenario 2	High demand for all product families compared to base scenario	High
3	Scenario 3	Low demand for all product families compared to base scenario	Low
4	Scenarios 4–10	High demand for one product family and base for the others	High for product family X

Results for all the scenarios were better for the pull-control with higher difference for scenario 2 and less difference for scenario 3. The results in Table 4 are those for the base scenario, scenario 1, and show a clear competitive advantage for the pull control, as it had achieved better on-time delivery with less inventory and idle time at the same throughput. The biggest difference in the results happened when there was a large order intake. In this case, the pull control was able to smooth the demand with the determination of order release and completion dates, as it could act on the bottleneck early on, but the push-control could only react reactively when it was already too late:

Table 4. Simulation results for push and pull-demand production control - base demand scenario.