Principle (Parallel Process Batching): In a batch operation the capacity grows linearly in the batch size, up to the maximum number of parts the operation can process. This implies that the minimum batch size required for stability may be larger than one. However, because the process must collect batches before running them, the WIP and FT at the operation also grows proportionally in the batch size.
Motivation
Some times the nature of the process is such that it naturally requires parallel processing. Heat treatment, burn-in, and plating are examples of processes where multiple parts are processed simultaneously with essentially no increase in the process time. The parallel process batch size is the number of jobs that are processed together. Since the process time is independent of the number of jobs processed at once, the capacity of the operation increases in the batch size. However, since we have to wait for jobs to form a batch, the flow time also increases in the batch size. Since the basic tradeoff is the same (capacity versus wait-to-batch time) as the serial process batch case, the underlying relationship of flow time to batch size is very similar.
Example
|
|
|
|
The above figure illustrates the mechanics of parallel process batching.
Jobs arrive independently to the station, where they are held until they
form a complete batch. Batches queue at the station until it becomes
available, whereupon they are processed as a batch. The following
figure illustrates the behavior of total flow time at the station.
Click here
for the parameters and equations used to obtain the graph. Note that if
the batch size is too small (below five for this example), flow time blows
up because the process has insufficient capacity to keep up with arrivals.
At around eight jobs, flow time is minimized, because the queuing due to
utilization and the wait-to-batch time are balanced. Above the optimum
batch size, flow time increases roughly linearly because wait-to-batch
time increases linearly in the batch size.
|
|
|
|
