Performance of modified wood in service - multi-sensor data fusion and its multi-way analysis

Recent developments in the field of electronic sensors and analytics provide new opportunity for accurate characterization of materials often based on portable and non-destructive methods. By using several complementary techniques, the material description is precise and complete. The data provided by multiple equipment, however, are often not directly comparable due to different resolution, sensitivity and/or data format. The complexity related to the data fusion step and its further interpretation often leads to not complete exploitation of the available data. This paper presents a multi-block approach used for merging experimental data collected by measurement of modified wood in service. Characterization of samples appearance (colour and gloss) is merged with spectral data that decodes information regarding chemical composition. Alternative approaches for data fusion on the low-, mid- and high-levels are introduced, discussed and confronted with the standard approach (single sensor data interpretation). Finally, the trial to analyse the data with multi-way method is presented and interpreted.