Salym Petroleum Development
Prediction of failure occurrence
Prediction of failure cause
Health index
Data on 233 failures of four types
69 technological parameters for each well + data on well construction, passport data of ESP, failure and downtime data, etc.
Use of integral features
Hybrid model:
parameters calculated using physical models are the initial features of the machine learning model (engine cooling efficiency, pressure at pump discharge, recovery of TMS data)
A system for detecting and analyzing abnormal conditions based on historical data has been created
A unique interface has been created that allows tracking the condition of complex high-load units in a convenient and accessible format
All enterprise platforms have been integrated into the system
The cost of implementing the software product depends on the number of mechanized extraction methods and the volume of functionality implemented in the 'CycleOp' software product.
To calculate the cost, please leave a request and we will get in touch.
