A model to predict failures in turbomachinery used in the oil & gas industry
Client X operates within the oil and gas industry. They have recently been experiencing failures from a crucial piece of turbomachinery. There are 32 sensors that record measurements (flow relations etc) relating to the machine. Each failure event results in an unplanned shutdown and has large cost implications, which is why the client is keen to address this problem.
Client X has provided some historical data and would like a predictive solution in order to take a proactive approach for any future failure events.
Data provided: • Historical time series data relating to the 32 sensors for an approximate 2-year period (February 2015 to February 2017). • A list of when failures have previously occurred.