Fig 1.
Promethee-based decision-making improvement using the following methods: Method A) P-PFA—Principal factoring with rotation and communality analysis; Method B) P-sPCA—Integration of sparse PCA into the Promethee methodology; Method C) P-SRD—Sum of Ranking Differences to qualify the consistency of the criteria.
Fig 2.
Methodological steps of the Promethee method.
Fig 3.
Types of generalised criteria (P(d)—Preference function).
Fig 4.
The Promethee-GAIA plane of the I4.0+ criteria—Red dots represents the alternatives and blue vectors denote the loading vectors of the criteria.
Fig 5.
Methodological steps of the P-SRD method.
Fig 6.
Criteria selected by the P-PFA method—The loading values are colour coded, the greener, the higher the value.
Fig 7.
Interpretation of criteria selected by the PFA method indicating two major criterion groupings: 1) employment rates of young people measured according to the years since their highest level of education was completed and 2) human resources in the fields of science and technology coupled with the number of research institutions in and news appearings concerning I4.0-related fields.
Fig 8.
Criteria of PCs selected by the P-sPCA method and their corresponding loadings.
Fig 9.
The rank of criteria according to the P-SRD method.
The x axis shows the SRD values. Criteria belonging to the same criterion group are located on the same level on the y axis.
Table 1.
Promethee-ranked performance of the top-ten NUTS 2 regions with regard to the most determinative criteria—’Human Resources in Science and Technology (Percentage of total population)’ concerning people employed in science and technology (HS5), Scientists and engineers (HS9) and people studying in tertiary education and/or employed in science and technology (HS7).
Fig 10.
Map visualization of the Promethee-based I4.0 readiness ranking of European NUTS 2 regions based on the two loadings of P-PFA.
It provides a visual representation of how regions perform/rank considering the two major criterion groups. Sub-figure A) represents regional rank based on the first loading, which includes criteria of employment rates and job opportunities. Sub-figure B) represents regional rank based on the first loading, which includes criteria of academia sector and its employees in R&D fields and I4.0-related news.
Table 2.
Summary of the advantages and utilization potential of the proposed methods.