Norway’s Force industry
body is inviting participation in the 2020 Force machine learning
competition. Interested parties are invited to predicting facies and
stratigraphy from well logs, and map faults on seismic. Force is to
make available some 150 wells, with logs tagged with stratigraphic and
lithological labels. Competitors will use ML to produce pseudo
lithologies for a further 20 more wells that will be held back.
Agile’s Matt Hall is to organize the ‘Kaggle (Google)
style’ competition and will help-out with ML tutorials. More from
Force.
Meanwhile Matt Hall (Agile) invites geos to try their hand at
his own geological coding challenge in a ‘Kata’. Working
from an online dataset of 20,000 lithology codes, coders are invited to
answer questions like ‘what is the total thickness
sandstone?’ More from Agile.
A team of Fugro employees won the geotechnical
machine-learning competition conducted as a prelude to the 2020
International Symposium on Frontiers in Offshore Geotechnics (ISFOG).
Competitors were supplied with a dataset of cone penetration test
results, hammer energy and pile dimensions and had to derive the most
optimum blow count vs. depth for an offshore jacket. A team comprising
data scientists, geotechnical consultants and pile installation
specialists worked on the project for four months! More on the
competition’s Kaggle home page.
AI analytics start-up Contilio won the Industrial Internet Consortium’s ‘Smart Construction Challenge’. The challenge involved construction applications integrating cloud, edge, fog and IoT technologies. Contilio used 3D computer vision and deep learning to provide intelligent insights from 3D data and photos captured at a construction site and processed in the cloud. Contilio won the €25,000 prize and the opportunity to deliver a live proof of concept at the TÜV SÜD International Business Park in Singapore. More from the IIC and the MachNation summary white paper.
© Oil IT Journal - all rights reserved.