Comau – availability checking and optimisation
Comau is a supplier of complex automation systems, for example, production lines for car body production. Such a production line consists of a row of workstations, several of which are "manned" by industrial robots. All stations have a fixed work cycle. Information is available on each station's failure frequency (MTBF - Mean Time Before Failure; beta distribution) and the duration of the failures (MTTR - Mean Time To Restore; logarithmic normal distribution). This results in an availability of the whole system or plant to be determined, which can be optimised by suitable distribution of buffer areas in which several bodies can quasi be temporarily stored.
The availability is to be determined for a specific project and optimised by varying the buffers.
Comau provided the necessary data for the system and its stations. Two buffers with 7 and 10 buffer areas respectively were already included. A model was created on the basis of this information and the data was entered. The model consists of a so-called network, which contains all stations and the buffers. To do justice to the statistical inputs, a longer period must be considered for each simulation to obtain stable results values.
Figure 1: Model network
A graphic showing the capacity utilisation of the individual stations is suitable for analysis of the system.
Figure 2: Capacity utilisation diagram
The model was used to assess different buffer scenarios and therefore to optimise the system. The buffer locations and sizes were varied.
A solution with 7% higher availability than the specified planning was only found by varying the buffer locations. However, no additional buffer areas were required, so that this new solution required reorganisation of the system, but did not involve any additional costs for technical materials.
The availability of a system in the planning phase was increased by 7% for the cost of a few days' service.