Automation of seismic scouting with robotics and machine learning

The accuracy of the survey design for the land seismic data acquisition became crucial for the best data quality and highest productivity with the emerging trend of high-channel high trace density acquisition. Satellite maps remain being a valuable source of information on the ground conditions in the area of interest despite typically lacking the necessary resolution and possibly being out of date for seismic survey planning. We propose using a small, dedicated field crew equipped with UAVs to quickly acquire high-resolution imagery in the most critical areas for the planned survey. Within days, this data can be processed, automatically interpreted using machine learning techniques, and provided to the survey planning specialists.