A leading coal miner, using tens of trucks to operate mines around the clock, was experiencing huge production losses from unanticipated in-field failures. Even though the customer had thorough asset utilization and work management data in which root causes of the failures were embedded, they lacked the tools to decipher causes from the raw data.
GE Digital’s Data Science team was brought on to mine the customer’s data for effective insights. In doing so, the Data Science team leveraged text mining techniques to identify top 10 failure modes, generated component-level survival distributions and aggregated estimated mean time between failures (MTBF) across all components to predict remaining useful life (RUL).
Incisively analyzing data that had been otherwise confined to siloes within the customer’s organization, the Data Science team recommended that the customer implement inter-inspection time changes and best practices for fleets across mines. As a result, the customer implemented these practical actions to meaningfully minimize unanticipated failures and optimize maintenance operation.
The GE Digital Data Science team leveraged text mining techniques to identify top 10 failure modes.