Scientific Background
When analyzing certain industries, you can discover that some – originally industry-independent software tools – can become a very popular choice in said industries and become an almost industry-wide solution.
For Spotfire you can encounter this phenomenon in semi-conductor manufacturing and the energy sectors (and this even though Spotfire is at its core an industry independent solution).
There are some concrete reasons for the success in semiconductor manufacturing that we want to look at in this article. The genesis of all these reasons is presumably that Spotfire was invented originally in the scientific community (more precisely: at the University of Maryland) to process and visualize scientific data. In this field the exclusive processing of aggregated data is not enough. It is necessary to make data accessible down to its raw values to come to a reliable understanding.
Processing of Large Data Sets
Spotfire has always been optimized to handle large data sets.
- When loading data you can configure to process data in-memory, in-database or on-demand.
- When creating a dashboard, you can follow the paradigm that you drill down towards the most interesting subsets, reducing the size of the data and enhancing the performance along the way.
- Each plot can be configured to only work an individual subset of the data.
- Data can be combined via custom queries across different data sources/databases and queried more efficiently.
Scientific Visualizations
Spotfire has the capabilities to visualize scientific data properly.
- Scatterplots and map plots can be configured to display „tiled markers“ to act as wafer maps. This enables to display test results (on said wafers) in an intuitive form.
- Map plots allow the inclusion and display of “layers” to overlay various results in one plot.
- Plots can be subdivided into categories on X and Y axis and on separate pages via the trellis functionality.
- Plots can contain calculated curves (“fits”) for example to estimate distributions.
- Naturally Spotfire contains a lot of plot types to display all sorts of data like line charts, heat maps, bar charts and so on.
Spotfire offers pretty, flexible and dynamic visualizations with excellent interactivity to drill, filter and in general explore data easily.
Integration of Advanced Analytics
Spotfire can integrate other analytical platforms.
- With Python and R, you can build and use analytical solutions based on the powerful foundations of the open-source community.
- With the Spotfire® Enterprise Runtime for R (TERR) you can use a professional enterprise-grade R environment for high performance and reliability.
- With Statistica you can integrate a data science tool, that can be used via point-and-click, just like Spotfire itself.
When data visualization and drill-down are not sufficient, these tools are the perfect ways to embed advanced statistics and machine learning and make them available through Spotfire. This way you can build tailored solutions for SPC (statistical process control) and DOE (design of experiments). Additionally, Spotfire supports automation of the UI and data ingestion via IronPython scripting.
Organization-wide Deployment
Spotfire democratizes data and analytics and enables everyone to benefit from insights from data.
- Data Scientists can accomplish their duties in gain insights.
- Data Scientists can design solutions for interested colleagues and make them available to them.
- Solutions built like this can be accessed through the browser. Enabling sharing among any interested party.
Solutions created in Spotfire do not need to stay in expert silos, where they were created. Instead they can be shared and distributed to any user group in any location. Spotfire dashboards can be opened in the desktop application as well as in the web browser via Spotfire Server. This way it is possible to share even with customers and suppliers.
Do you want to learn more about Spotfire in semiconductor manufacturing?
