US: Seismic interpretation and reservoir characterisation company Fugro Jason, along with another unnamed major seismic company, have moved to using object orientated databases to manage their seismic data, because they believe it gives them faster results.
The seismic research company has used the database for their data, to run their acquisition and processing modules. The database collects all the sensor data, including GPS, source and receiver positions, seismic response and node position and puts it into the Objectivity database. The project development manager said that the database was “perfectly suited to the demanding data acquisition requirements”.
Fugro Jason uses the database to manage the data from different sources across its geology, geophysics, petrophysics and modelling applications. The company said that analysis which once took days can now be done in minutes.
The theory is that normal databases (commonly known as ‘relational databases’) are not very good at processing data when it means finding complex relationships between the fields.
Relational databases are ideal for tasks which involve putting data into a storage and taking it out later.
But if a task is needed which involves finding complex patterns and relationships between data, then object orientated databases can do the job faster. For example, Amazon manages to trawl its database to notice that several people who bought one product also bought another one (a process which creates 20 to 30 per cent of its revenues).
If companies do not want to switch completely from a relational database to an object database, they can use the existing database to store and catalogue the data, but use Objectivity to analyse the relationships. They can also create external processes run on separate machines.
Source: Finding Petroleum