File:Pgeu2012 rsunny.pdf

In research and development a reliable and stable storage of measured data and metadata is essential.

Up to now these data are mostly stored in simple files. Metadata concerning the specimens are often handled by hand or calculation sheets.

This talk is about three success stories of three different big technical projects. In these projects PostgreSQL is the core software component for storing all the data.

Wind Energy: Testing of large size bearings of wind mills.

   http://www.fag.de/content.fag.de/en/branches/industry/wind_energy/wind_energy_service/largesizebearingteststand/astraios.jsp?

   Recently, Astraios – the most modern, largest, and highest performing large-size bearing test rig in the world - started operation.
   PostgreSQL has to store more then 3 Million measurement points every second continously over a periode of several weeks. In parallel, PostgreSQL has to retrieve the data for evaluation. We explain our technology in effiently handling these hugh amounts measurements.

Automotive Industry: Test center to improve corrosion behaviour http://www.focus.de/fotos/der-inka-test-verlangt-dem-auto-alles-ab-unter-den-extrembedingungen_mid_493179.html

   The AUDI INKA test simulates the whole life of a car within 19 weeks. This is achieved by a well defined sequence of special intensive treatments. The timestamps of all the treatments of the all cars are also stored in the database. Special functionalities of PostgreSQL makes evaluation mutch faster and easier compared to other solutions.

Plant genetics: Automated mass phenotyping by machine vision http://www.gmi.oeaw.ac.at/ http://www.hesotech.de/anwendungsberichte/phaenotypisierung/bilderfassung.html

   At the Gregor-Mendel-Institute in Vienna, the relation between apearance and genetics of plants is investigated. In each experiment, the data of more then thousend individual plants have to be administrated. A robotic system takes pictures of each plant at least twice a day. Beside the climatic data and the results of the online machine vision evaluation, all taken images are stored in PostgreSQL. PostgreSQL makes it possible to setup an efficient statistical evaluation of the experiment, considering all the relations between the aquired data and metadata.