Data-driven insights to improve heavy equipment maintenance and reduce downtime.
Challenge
A heavy equipment dealership wanted to enhance their maintenance recommendations and predict engine failures more effectively. They sought to improve their understanding of customer equipment health in the field beyond what their Condition Monitoring Agents (CMA) could provide. Existing practices struggled to detect anomalous time-series patterns in engine sensors and anticipate engine faults, leading to reactive maintenance and potential equipment downtime.
Solution
Arundo developed a solution centered on predictive analytics. A survival analysis-based, regularized regression model was created using engine sensor data to predict engine faults. This model provided a probability distribution of expected engine faults over time. Additionally, models were developed to detect anomalous patterns in engine sensor data. These models were designed to integrate into existing CMA workflows, aiding in daily triage and activity prioritization based on the latest engine telemetry data.
Impact
The implementation resulted in enhanced insights into the health of heavy equipment assets in the field. The dealership gained predictive capabilities that improved their ability to anticipate engine failures and identify anomalous patterns. This allowed for more proactive maintenance, potentially reducing downtime and improving overall equipment reliability.
