Accelerated Seismic Data Cloud Ingestion through Machine Learning and Automation #seismic #cloud #machinelearning #automation #bigdata #analytics #digitize #metadata #casestudy
22 May 19
10:30 AM
-
11:00 AM
Tracks:
Technical Trends and Innovation
Woodside has focused on moving seismic data to the cloud ahead of other subsurface data types with the objective of moving all seismic field and processed data, supporting files and documentation to enterprise cloud storage from where it will be used for processing, interpretation, specialist services, ingestion to data lakes and analytics. Key drivers behind this cloud investment are liberating data from proprietary application formats, increasing data visibility and availability, quantifying data confidence and reducing the Non Productive Time (NPT) between data reception and delivery of workstation ready seismic to geoscientists. Indexation and gathering of attributes from diverse seismic data sources presented a bottleneck in the seismic cloud ingestion workflow. Structured and unstructured data has traditionally been managed using different applications, workflows and resources. While unstructured data management tools still have value for specific tasks such as managing versioning, technology developments now enable all data to be treated as a heterogeneous attribute source. Through examination of available technologies, Woodside identified technology components solutions for accelerated data ingestion using machine learning and automation to harvest attributes and quantify confidence for individual data objects. The richness of the seismic meta data over time also created opportunities to attempt pattern matching against corporate metrics such as drilling success and application of emerging acquisition, processing and interpretation technologies. In this presentation, Woodside will describe how these technology components have been used in conjunction with sentiment analysis, business rules and semantic mapping applied through machine learning functionality to harvest attributes from seismic data files and documents to accelerate their cloud-based seismic data ingestion process and reduce NPT.