Posted 31.08.2018 13:16 by Magnus Tvedt
It started with a discussion on how to solve a drilling problem for a portfolio of development wells, and the experienced drilling engineer asked us if we could do an analytical study based on data for the entire NCS to transfer experiences from all the historic wells.
We got a good description of the problem in the beginning and we started working on it, using our own databases. We were able to reuse our preexisting models which saved us time - and the customer money. We had one update meeting in the middle of the project, where we discussed progress and got some answers which sped up the development.
By applying data science and machine learning within our domain of expertise, with thousands of wells as reference, we extracted experiences on
We are proud to deliver a complex project on time, on cost and on quality. In the discussion which followed, we learned how Aker BP was looking for answers and a fully decisive, data driven model. That will be the goal of our next project.
Also, this being our first data science and machine learning project, we are grateful for the trust and openness we meet in Aker BP, we love being a part of their #dataliberationfront.
Part of the project was to prove integration of Pro Well Analytics to Halliburton's Digital Well Program. This first step shows a valuable add-on to their new suite of well planning products in the data science and machine learning domain.
Clustering of Wells with Machine Learning
In January 2019 we are launching a new era in Well Planning with our machine learning powered well analysis tool. Here is a behind the scenes peek of what you will get when we merge the best of data science with global experience transfer based on real data in well planning.