What sparked off Repsol’s use of IBM’s cognitive technology?
We believe that the technology is applicable in two areas, as an aids to decision making in bidding rounds and in helping with field planning and optimization—where to place injectors and producers. The idea is to make the three fields of complex math/analysis, geoscience and computing accessible and working together. We have already worked internally on technology to optimize field development plans, to increase production, keep costs down and assure safe operations. The cognitive initiative will build on this.
What was this previous project called?
This was our Excalibur project. There came a moment when, in discussion with IBM, we realized there was an opportunity to take this to the next level by adding a cognitive computing component to Excalibur. The idea is to leverage new computing possibilities offering more interaction, machine learning and even reasoning in an advisory context. We thought that this was a great opportunity to do something together.
Where is this today? Are you just starting out or have you already run a pilot?
We are at the early stages of developing the technology. But we started trials as an extension of Excalibur about a year ago. The first results are already there - and we are encouraged by this promising area of innovation.
So what exactly is involved? We have previously reported on IBM Watson’s Jeopardy success. Is this mainly concerned with natural language processing (NLP)?
Sure NLP is a key component. But we are working in other directions too. We are looking at the behavioral and psychological side of decision making and also how we can leverage ‘big data’ access. In fact we want to go beyond data access, to see how these intelligent systems can help us regain control of our huge data sets, realizing their full potential. Making serendipitous links and discoveries across data, text and people.
The IBM release makes a lot of ‘cogs’—is this a marketing term? What exactly are cogs?
Cogs is just a fancy name for the apps and tools we are developing to access and reason with natural language and or big data resources.
In Jeopardy, Watson was fed with large public domain information assets including Shakespeare, the Bible and Wikipedia/dbPedia. What is your ‘feedstock’ for the project?
All public oil and gas related information sources and of course our own databases and information assets will go into the mix. One key idea is to be able to identify relevant analogs for current areas of interest by trawling large subsurface information assets.
Comment—Following our interview, Repsol announced the $8.3 billion acquisition of Canadian Talisman Energy. We don’t know if Watson advised on the transaction.
© Oil IT Journal - all rights reserved.