Data Fusion and Digital Modeling

The concept of data fusion originated in the 1970s and was first applied in the military. With the growing complexity of bridge structures, single-source data could not satisfy the demand of bridge health monitoring. Hence, it was essential to synthesize data from different sensors in order to evaluate the bridge condition more precisely. Therefore, data fusion plays a vital role in making decisions for bridge maintenance. The current bridge health monitoring data centers have collected numerous types of data, hence it is imperative to perform data fusion. This chapter will present a systematic overview of the data fusion algorithms, data fusion-based condition assessment methods, and data fusion platforms within the intelligent bridge operation and maintenance framework. It should be noted that, in a strict sense, data implies unprocessed unstructured numerical sets; while information implies processed data with significance. The general notion of data covers raw data (sensor data), processed data and validated data (that is, information). In this chapter, our data refers to the general notion of data unless otherwise stated.