We offer research-backed statistical services to manufacturing and
production facilities that use multistage batch and continuous
production processes.
Our services enable the analysis of data using effective and
economical sampling strategies and state-of-the-art experimental
design techniques.
In a multistage process, the final output is the culmination of a
number of stages. Often, upstream production defects are felt further
downstream in the production process. Production problems - such as
process not performing consistently over time, not achieving desired
specifications within acceptable levels of variation, or defects that
show up in downstream stages - can occur at any stage of the process.
To solve production problems, it is necessary to gather information
on each stage of the process as well as to evaluate output
characteristics.
CSIRO's statisticians can develop sampling strategies to generate
suitable and reliable data at all stages of a multistage process,
allowing problem stages to be identified. CSIRO can then assist
manufacturers to design experiments, either on-line or off-line, to
pinpoint the factors that contribute to the problem.
Through further capabilities in statistical modelling and data
analysis, CSIRO's methodologies also enable evaluation of:
- The variation structure of the data at different stages
- The cross co-variation and correlation structure of the
characteristics of the material at different stages covering input,
throughput and output material or product.
|
