Hosted by the The Professional Petroleum Data Management (PPDM) Association
Presenters: Neil Constantine & Cyril Gagnaire – “Analytics For Data-Driven Decision Making: A Practical Example”
Abstract: With more than 12,000 petroleum wells in Australia and a rapid release schedule, we’re challenged to meaningfully understand data availability on a project timeframe. Working with a NWS operator using documents loaded to their AWS storage environment, we deployed an analytics approach to assign data and documents by category and predict the effort required to access their content based on format and fidelity. Results were loaded to a BI front end linked to the cloud data store, allowing spatial and filtered selection of preferred documents to feed into a machine-learnt data mining and extraction routine to then load to the interpretation environment. Whilst work in progress, we are targeting a ten-fold reduction in the time to find, QC and load specific data types compared to a legacy manual approach. This ability to shorten load times whilst simultaneously expanding the data footprint fed to interpreters supports true data-driven decision making. One of the key capabilities is to be able to tailor project scope to available time frames and resources by applying processing only to data sources that are fit-for-purpose to support a given decision schedule.