Biomedical Engineering Reference
In-Depth Information
Fig. 11 Treemap of batches for four steps (Drug Substance, ChromI, ChromII and ChromIII) of
a process, coloured according to the ''% step recovery'' value obtained in those batches; darker
colour indicates higher recovery
• Which were the batches where the step recovery was highest?
• Which were the batches that had low step recovery?
• Which unit operation had highest step recovery?
On the one hand such quick answers to these questions can provide employees
with faster process knowledge, while at the same time they serve as a powerful
tool to trigger root cause investigations and process improvement initiatives by
representing the process in a concise and easy-to-understand format.
Many types of view can be generated for a process based on a treemap; another
example is presented in Fig. 12 . This view can help a process engineer/scientist to
visually compare all important parameters for a single batch against its respective
specification limits, providing a quick snapshot of all the parameters that are
approaching specification limits. In this example, each step is represented by larger
rectangles, within which the parameters for that step are represented by smaller
rectangles. Since the values of parameters differ in scale, these have been nor-
malized based on the specification limits for the parameter, where 0 % denotes the
lower limit and 100 % represents the upper limit. The colour shade of the smaller
rectangles (parameters) is scaled according to the value obtained after this nor-
malization, where parameters approaching upper limits are coloured red and those
approaching lower limits are coloured green.
2.2.2 Statistics-Based Monitoring
The next level of process monitoring involves various tools and charts that are
based on principles of statistics and need a specialised skill-set and expertise to
understand and infer useful conclusions. Most of these apply statistical process
Search WWH ::




Custom Search