SWIP (Simulation of Wave Propagation) research line participates at the Oil&Gas Conference 2018

• PhD student Mostafa Shahriari gave a poster presentation about the Geosteering Measurements that his research group develops by using Deep Learning algorithms

MATHMODE and the Simulation of Wave Propagation group at BCAM and UPV/EHU have presented their research on Geosteering at the Oil&Gas Conference 2018 this week. The event, which took place at Bilbao Exhibition Center (BEC) on 8-9 October, is a key international conference in its area and gathered major players of the sector, from operators to technologists, consultants, engineers and R&D Managers.

BCAM PhD student Mostafa Shahriari participated at the exhibit area of the conference presenting a poster on his group’s research in this field: interpreting Geosteering measurements using Deep Learning algorithms.

Geosteering is the act of adjusting inclination and azimuth angles of a well trajectory to reach a geological target. Deep Learning algorithms are used to approximate an explicit representation of the inverse function needed to interpret geosteering measurements. Specifically, the group employs convolutional deep-neural networks for the efficient inversion of 1.5D dimensional (1.5D) borehole resistivity measurements and they intend to complement this with transfer learning for the fast inversion of 2.5D and 3D measurements.