Application of multiple information sources to prediction of engine time on-wing

Abstract

The maintenance and operation of commercial turbofan engines relies upon an understanding of the factors which contribute to engine degradation from the operational mission, environment and maintenance procedures. A multiple information source system is developed using the Pratt & Whitney engine to combine predictive engineering simulations with socio-technical effects and environmental factors for an improved predictive system for engine time on-wing. The system establishes an airport severity factor for all operating airports based upon mission parameters and environmental parameters. The final system involves three hierarchical layers: a 1-D engineering simulation; a parametric survival study; and a logistic regression study. Each of these layers is combined so that the output of the prior becomes the input of the next model. The combined system demonstrates an improvement in current practices at a fleet level from an R2 of 0.526 to 0.7966 and provides an indication of the relationship suspended particulate matter and engine degradation. The potential effects on the airline industry from city based severity in maintenance contracts are explored. Application of multiple information sources requires both knowledge of the system, and access to the data. The organizational structure of a data analytics organization is described; an architecture for integration of this team within an existing corporate environment is proposed.