Empirical study of manufacturing enterprise collaboration network: Formation and characteristics

• A partner selection method considering the uncertainties is proposed.
• An empirical study of a manufacturing network is carried out.
• Statistical features of the network are analyzed.

With the rapid development of society and economy, market competition is getting increasingly fierce. It is almost impossible for a single enterprise to respond promptly to the changing market. As a result, more and more enterprises tend to improve their competitiveness by means of collaboration. Specialization and cooperation arise between enterprises, forming manufacturing enterprise collaboration network (MECN). MECN is a complex network which comes into being by means of numerous enterprises selecting partners. Therefore, partner selection is a key issue in MECN. Considering that existing partner selection methods are subjective and theoretical to some extent, an objective and practical method called gray comprehensive evaluation is adopted in this paper. Besides, probability is introduced to simulate the preference and uncertainties. Correlations between each index and the optimal index are calculated according to which sort the possible partner enterprises, and the enterprises nearest to the optimal indexes are more likely to be selected. An MECN driven by an airplane as the final product is then studied as a case, and the statistical characteristics of the network are analyzed. Results show that probability-based gray comprehensive evaluation is an effective way to model MECN, and the network is of self-organization, complexity and localization.