Sensor fusion for vehicle health monitoring and degradation detection

Health monitoring and degradation detection systems help to address problems at an earlier stage and reduce maintenance costs. A vehicle problem is characterized by high bandwidth vibration signals and oil data that allow one to diagnose a wide variety of failures from relatively few measurements. Reliable health monitoring and degradation detection require the integration of research in the diverse fields of instrumentation, data acquisition/storage, signal processing, approximate reasoning, feature extraction techniques and multi-sensor data fusion. Data fusion is the synergistic use of information from multiple resources in order to assist in the overall understanding of the condition of a system. It offers a more complete view of the whole situation and a more accurate evaluation of the condition based on information from individual sensors. Vibration analysis and oil debris analysis can be sensitive to failures of different type or degradation stages. So the combination of these two types of information should give more reliable and accurate detection of vehicle degradation and failure.