AI/machine learning

Fugro Wins Contest With Machine-Learning Model for Pile-Driving Prediction

Using the supplied data set of cone penetration test results, competing teams had to predict the number of hammer blows required to drive the pile a given unit of depth in the North Sea.

Offshore production facility
Fugro

Competing with 60 other teams from industry and academia around the world, a Fugro team came in first in the pile-driving prediction event. The competition ran from April to December 2019, and ended on 1 January 2020.

Using the supplied data set of cone penetration test results, hammer energy, and pile dimensions, competing teams had to predict the most accurate pile installation driving blowcount vs. depth for jacket piles installed in the North Sea—the number of hammer blows required to drive the pile a given unit of depth. The Fugro team combined machine-learning techniques with their geotechnical expertise to develop a stable and reliable pile-driving model.

The competition, organized as part of the International Symposium on Frontiers in Offshore Geotechnics 2020 conference, which will be held in Austin, Texas, in August, was hosted on Kaggle, a subsidiary of Google that is an online community of data scientists and machine-learning practitioners. The Fugro team comprised data scientists, geotechnical consultants, and pile installation specialists from around the world, and worked together for 4 months in the latter part of 2019.