i tend to agree with Michael, but have a few more comments. Applied mineralogy is fundamental to mineral processing (like chemistry is to chemical engineers). Without mineralogy beneficiation processes are based on assays rather than minerals. in simple terms this means copper recovery may be poor because the copper is found in copper oxide rather than copper sulphide (which is floatable).
It is not possible to identify how to optimise a plant's performance without mineralogy. However there are a number of pathways to obtain mineralogy with the two main ones being assay to mineral conversion, and mineralogical systems. In general netir approach is well-utilised (as commented by Michael), mainly beacuse the mathematical methods to fully utilise the data are neither well understood or utilised.
One must be careful about what is meant by data science for which the exact meaning is open to debate, particularly identifying the difference between data science and data analytics. Hence I prefer the question: what problems does the mining industry face that can be solved through data analysis. For which my answer would be: too long to list. However my particular interest is mineral processing optimisation from plant surveys.
To me datascience is not data analysis but is a more holistic view data and its application. I do not think datascience can truly be considered a separate subject area to data analytics; and therefore does not warrant recognition as a well-defined term. However I would tend to regard the problem of "what data should I collect?" as datascience, and "how is the data used?" as data analytics.
I regard things like storing data in Excel spreadsheets instead of relational databases as a perfect example of where datascience is not being applied. A datascientist (and also a data analyst) would argue that the use of unstructured Excel workbooks (rather than a relational database) hinders the utilisation of the data. Excel is here an example. Of course one can think of many examples of where Mining Companies are rendering data near-unusable through poor datascience. Therefore I would argue that wherever there is data,, datascience is required.
i tend to agree with Michael, but have a few more comments. Applied mineralogy is fundamental to mineral processing (like chemistry is to chemical engineers). Without mineralogy beneficiation processes are based on assays rather than minerals. in simple terms this means copper recovery may be poor because the copper is found in copper oxide rather than copper sulphide (which is floatable).
It is not possible to identify how to optimise a plant's performance without mineralogy. However there are a number of pathways to obtain mineralogy with the two main ones being assay to mineral conversion, and mineralogical systems. In general netir approach is well-utilised (as commented by Michael), mainly beacuse the mathematical methods to fully utilise the data are neither well understood or utilised.