One of the most interesting scientific trends in recent history is the trend towards “open science”. Specifically, the use of open data and open source software greatly increases the accessibility of science. This in turn speeds up R&D and operations. This trend provides huge opportunities for the oil and gas industry.
Academic researchers clearly benefit from open science, but oil and gas companies should not overlook the benefits to their business. Utilizing open science reduces R&D and operational spending. Companies that take advantage of the increased power and lower costs gain a clear competitive edge over their industry peers who cannot or will not participate.
To an outsider it might be difficult to see how oil and gas companies would benefit from open data. Clearly any company would be dis-incentivized to share its R&D or operations data with competitors. However, many companies can and currently do benefit from non-proprietary, open data. Government agencies (like the USGS) aggregate and provide reams of publicly available data relevant to oil and gas companies. Publicly available data sets range from general geoscience surveys to company specific operational data such as well logs, which are actually required to be made available in certain situations.
The bigger opportunity for companies to differentiate themselves within their industry comes from recent advances in open source software coming from the software tech industry. Unfortunately, the software industry often severely lacks domain knowledge of other industries (such as… oil and gas). However, the software industry creates undeniably great technologies for aggregating, storing, exploring, and analyzing data. Oil and gas companies do not need the software tech industry to help them succeed, but these companies certainly can benefit from the technologies that the software industry develops. Disparate technologies such as Spark for general purpose big data and machine learning workloads, TensorFlow for efficiently creating and running complex reservoir simulations on cheap distributed GPU clusters in the cloud, and the SciPy stack (Jupyter, Pandas, NumPy, Matplotlib) for data exploration, data science, and machine learning all have strong and well-established communities, resulting in great documentation, painless learning, and (relatively) easy-to-hire talent. Modern open source software also integrates well with other software; any of the tools mentioned above can be easily integrated with existing and/or critical systems within companies.
To provide a concrete example of how open source technologies can be used to work with open data relevant to the oil and gas industry, I’ve put together a sample Jupyter notebook (see link below) which gets a publicly available mud logging dataset, visualizes the data, and does some lightweight transformation and analysis. While the scope of this example is small, it should demonstrate that compelling visualizations and robust statistical analyses from the SciPy ecosystem can be combined in an easily accessible and shareable format.
Github – Well Log Visualization
Interested in how your company can take advantage of open science? Contact us at info@bpcs.com.
Let’s work together to find your Next Big Thing.
