Maximizing vehicle uptime and reducing maintenance costs are critical objectives in modern automotive systems, making efficient resource utilization a top priority. One of the key factors is engine oil life or degradation, which directly affects the engine performance, longevity, and overall vehicle efficiency/fuel economy. Most vehicles tracks engine oil life solely on a fixed mileage interval while few uses dedicated sensor, which is costly and requires service and maintenance.
As the engine oil degrades, it reduces Oil Total Acid Number (TAN) increases while Oil Total Base Number (TBN) decreases. It is recommended that maximum usable life of the engine oil is up to the crossover point between oil TAN and TBN (as the engine oil degrades). Vehicle driving pattern governs the occurrence of crossover points with respect to vehicle mileage. Based on this fundamental concept, an XG-Boost machine-learning algorithm is trained using vehicle Controller Area Network (CAN) channels and varying oil TAN and TBN parameters, derived from the vehicle-level measurement data available for the entire life cycle of engine oil in operating condition. The developed model based on CAN channels like engine rpm, engine torque, gear position, engine power, coolant temperature and odometer readings accurately predicts engine oil TAN and TBN parameter. The cross over point of TAN & TBN is accurately forecasted as seen in correlation results. An interactive user interface is designed and developed to display the deterioration in terms of remaining useful life of the oil to customers in vehicle driving condition.