Method and system for determining a status of a hydrocarbon production site
Methods and systems for determining a status of a hydrocarbon production site are provided. The systems and methods include defining an area about a hydrocarbon well location and obtaining a time series of mobile geolocation data for at least one mobile device located within the area about the hydrocarbon well location. The times series of mobile geolocation data is processed and probabilities that the mobile geolocation data is associated with a drilling activity, a fracking activity, a tie-in activity, and/or an absence of drilling activity, fracking activity, and/or tie-in activity are output. The mobile geolocation data is categorized, based on the probabilities, as being associated with one of drilling activity, fracking activity, or absence of drilling activity and fracking activity.
1. A method for determining a status of a hydrocarbon production site, comprising:
defining an area about a hydrocarbon well location;
obtaining a time series of mobile geolocation data for at least one mobile device located within the area about the hydrocarbon well location;
processing the time series of mobile geolocation data using a recurrent neural network and outputting a probability that the mobile geolocation data is associated with a drilling activity, a probability that the mobile geolocation data is associated with a fracking activity, and a probability that the mobile geolocation data is associated with an absence of drilling activity and fracking activity, wherein processing the time series of mobile geolocation data using the recurrent neural network comprises extracting features representative of the mobile geolocation data in a context layer; and
categorizing, based on the probabilities, the mobile geolocation data as being associated with one of drilling activity, fracking activity, or absence of drilling activity and fracking activity.
2. The method of claim 1 , wherein obtaining a time series of mobile geolocation data of at least one mobile device comprises obtaining a number of mobile devices located within the area about the hydrocarbon well location and an association of the mobile devices with a prior drilling activity, a prior fracking activity, or a prior absence of drilling activity and fracking activity.
3. The method of claim 1 , further comprising obtaining a time series of image data depicting the area about the hydrocarbon well location.
4. The method of claim 3 , wherein obtaining a time series of image data depicting the area about the hydrocarbon well location comprises obtaining a time series of at least one of radar images and optical images depicting the area about the hydrocarbon well location.
5. The method of claim 3 , wherein obtaining a time series of images depicting the area about the hydrocarbon well location comprises obtaining images depicting drilling equipment or depicting fracking equipment.
6. The method of claim 3 , wherein processing the mobile geolocation data and outputting the probability further comprises processing the time series of image data with the mobile geolocation data using a machine learning algorithm and outputting a probability that the mobile geolocation data and the time series of image data is associated with a drilling activity, a probability that the mobile geolocation data is associated with a fracking activity, and a probability that the mobile geolocation data is associated with an absence of drilling activity and fracking activity.
7. The method of claim 6 , wherein the machine learning algorithm comprises a recurrent neural network.
8. The method of claim 1 , further comprising identifying a start date and an end date of at least one of the drilling activity and the fracking activity.
9. The method of claim 1 , wherein selecting an area about the hydrocarbon well location comprises selecting a well location from a database of drilling permits.
10. A system for determining a status of a hydrocarbon production site, comprising:
a memory storing a time series of mobile geolocation data for at least one mobile device located within an area about a hydrocarbon well location; and
a processor communicatively coupled to the memory and configured to:
process the time series of mobile geolocation data obtained from the memory using a recurrent neural network, wherein features representative of the mobile geolocation data in a context layer are extracted;
calculate a probability that the mobile geolocation data is associated with a drilling activity, a probability that the mobile geolocation data is associated with a fracking activity, and a probability that the mobile geolocation data is associated with an absence of drilling activity and fracking activity; and
categorize the mobile geolocation data as being associated with one of drilling activity, fracking activity, or absence of drilling activity and fracking activity.