Hey-ho another year comes to a close and so much that I planned or wanted to do remains undone. OK we did manage to do the survey which gave some pointers as to where we should be taking Oil IT Journal and www.oilit.com next. But as for implementing major changes? Zilch!
The problem is that the time taken to prepare Oil IT Journal has, over the last couple of years expanded considerably. No major changes but a gradual increase in the detail and scope of our coverage which I hope you noticed. I certainly have done as producing the Journal is taking longer and longer. Hold that thought…
~
Just before I left the SEG in Denver I came across the exhibit of the SEG Advanced Modeling Corp. SEAM produces 3D synthetic seismic data sets of various geologies that are used to test imaging algorithms. The EAGE provides similar datasets for its Marmousi models.
SEAM is about to embark on a new ‘Life of Field’ (LoF) project, to build a static earth model, populated with rock physics and reservoir fluid parameters. This will be used to simulate seismics as before, but with repeat models run as oil and gas is extracted from the field, to produce a 4D time lapse seismic dataset. The SEAM prospectus states that ‘Most of the ingredients of such a [data driven production] strategy are already available. What is now needed is a productive way of linking [everything] together in an integrated system.’ SEAM’s starting point will be a ‘highly realistic earth model.’
Now when this was presented to me I had a eureka moment. Having tracked progress of Energistics’ Resqml earth modeling initiative for a while, with notably a review of the book of the project I thought wow! This is tailor-made for Resqml.
The idea of combining SEAM and Resqml, I should say my idea of combining them (false modesty is not my strong suit) has many obvious benefits. One the one hand it is an ‘industrial strength’ test of Resqml’s adaptability to a ‘use case’ not a million miles removed from its original purpose. On the other hand, seismologists, instead of creating and using ‘yet another’ model format will be using an industry standard. This is a no brainer, or so I thought.
Back in the office I fired off a few emails putting my suggestion forward to both the SEAM and Resqml teams. I waited. There was a deafening silence, the like I have not heard (sorry for the strange metaphor) since my March 2010 open letter to the SPE president. Eventually, after a couple more pings and prods, I got a lukewarm response from Resqml along the lines of, ‘it’s not a bad idea, maybe a couple of versions down the road we’ll be able to do something.’
But that of course means missing a great opportunity. I pinged some more and I have to say that I seem to have got some folks thinking about this possibility although I have not heard anything concrete back. I will keep you posted if and when I do.
There are plenty of good reasons why a Seam/Resqml collaboration should go ahead. The spirit of collaboration is in the air—think Standards leadership council for instance.
I can also think of a few bad reasons for not doing this. While people love to talk about collaboration and ‘breaking down the silo walls,’ actually doing something to achieve such is harder and usually out of folks’ comfort zones. It is easier to stay with the tried (and tired) old methodologies of the past rather than adapt to meet someone else’s requirements.
It could be that some facets of Resqml do not lend themselves to the seismic test. It could be that it doesn’t work at all! So there is an element of a challenge in my innocent ‘suggestion’ too. Make it work guys!
During one exchange, my interlocutor expressed ‘intrigue’ at my interest and involvement. This got me thinking. I was intrigued too. Why should I care how the majors and standards bodies run their projects? The simple answer is that this is what editors do, inform and influence.
Actually we have been active in similar fields in the past. Our persistent questioning of the supposed benefits of the semantic web in engineering have contributed to, let’s say, a more moderate presentation of these. We were also, as I learned some time after the fact, influential in mitigating a hook up between the PPDM’s standard data model and Esri’s proprietary technology. Oil IT Journal is an activist as well as an information source.
~
Which brings me back to my workload. I want to keep traveling to and reporting from conferences, to give more attention to the web site and to be active and influential. Regarding the website I have two projects in mind. First I want to bring search in-house and make it smarter—think Lucene maybe. I also want to try to get something along the lines of IBM Watson running our two million plus word information asset—think Uima. I also want to create an ecosystem of websites-of-relevance to what we (and I hope) you are really interested in.
To facilitate all this, we are going to cut back slightly on the number of issues that we publish. Starting in 2015, there will be 10 issues (down from 11) spread out through the year. In keeping with the ‘activist’ spirit, I think that we will revive the French revolutionary calendar with its 10 decimal ‘months.’ Watch out for the inaugural ‘Pluviôse’ edition next year.
Speaking of which, all the best from Teresa and myself for 2015.
@neilmcn© Oil IT Journal - all rights reserved.