公开(公告)号：US20250060498A1

公开(公告)日：2025-02-20

申请号：US18233584

申请日：2023-08-14

Applicant: X Development LLC

Inventor： Alex S. Miller ,  Robert Clapp ,  Aparajit Raghavan ,  Artem Goncharuk ,  Kevin Forsythe Smith

IPC: G01V1/38 ,  G01V1/104

Abstract: Techniques for imaging a subterranean formation include activating an acoustic energy source that is at least partially submerged in a volume of liquid on or under a terranean surface; based on the activating, producing acoustic wave energy that travels through the volume of liquid and to a subterranean zone below the terranean surface; receiving, at one or more acoustic receivers, reflected acoustic wave energy from the subterranean zone; and generating, with a control system, data associated with the subterranean zone based on the reflected acoustic wave energy.

公开(公告)号：US20240369726A1

公开(公告)日：2024-11-07

申请号：US18654765

申请日：2024-05-03

Applicant: X Development LLC

Inventor： Alex S. Miller ,  Robert Clapp ,  Artem Goncharuk ,  Kevin Forsythe Smith ,  Joseph Hollis Sargent ,  Shane Washburn ,  Jonathan Gray Wilfong ,  Allen Richard Zhao ,  Jonathan Blair Ajo-Franklin

IPC: G01V1/22 ,  G01V1/04 ,  G01V1/143 ,  G01V1/34

Abstract: A system includes a mobile vehicle including a geolocator and an active acoustic source configured to generate acoustic wave energy directed toward a fiber optic network that includes one or more fiber optic cables and a distributed acoustic sensing (DAS) interrogator communicably coupled to the one or more fiber optic cables; and a control system. The control system is configured to perform operations including acquiring a signal from the DAS interrogator in response to the acoustic wave energy generated from the active acoustic energy source during movement of the mobile vehicle on or above the terranean surface; determining a geolocation of the mobile vehicle from the geolocator during or subsequent to acquisition of the signal from the DAS interrogator; and determining a location of the at least one fiber optic cable based on the determined geolocation of the mobile vehicle during acquisition of the signal from the DAS interrogator.

公开(公告)号：US20240045089A1

公开(公告)日：2024-02-08

申请号：US18228805

申请日：2023-08-01

Applicant: X Development LLC

Inventor： Artem Goncharuk ,  Nam Phuong Pham ,  Shiang Yong Looi ,  Stuart Farris ,  Robert Clapp

IPC: G01V1/28 ,  G06F30/27

CPC classification number: G01V1/282 ,  G06F30/27

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating realistic synthetic seismic data items. One of the methods includes obtaining a plurality of synthetic seismic data items; obtaining a plurality of real seismic data items; processing each of the plurality of synthetic seismic data items using a machine learning model; processing each of the plurality of real seismic data items using the same machine learning model; determining a range for values for one or more parameters of a synthetic seismic data generator by comparing the synthetic seismic data items and the real seismic data items in an embedding space of the machine learning model; and selecting, as realistic synthetic seismic data items, a plurality of synthetic seismic data items that have been generated with a respective combination of values for the one or more parameters that is within the determined range.

公开(公告)号：US20240070459A1

公开(公告)日：2024-02-29

申请号：US18456792

申请日：2023-08-28

Applicant: X Development LLC

Inventor： Artem Goncharuk ,  Robert Clapp ,  Kevin Forsythe Smith ,  Shiang Yong Looi ,  Ananya Gupta ,  Joses Bolutife Omojola ,  Min Jun Park

IPC: G06N3/08 ,  G06N5/04

CPC classification number: G06N3/08 ,  G06N5/04

Abstract: This disclosure describes a system and method for effectively training a machine learning model to identify features in DAS and/or seismic imaging data with limited or no human labels. This is accomplished using a masked autoencoder (MAE) network that is trained in multiple stages. The first stage is a self-supervised learning (SSL) stage where the model is generically trained to predict data that has been removed (masked) from an original dataset. The second stage involves performing additional predictive training on a second dataset that is specific to a particular geographic region, or specific to a certain set of desired features. The model is fine-tuned using labeled data in order to develop feature extraction capabilities.