Based on the bilateral trade data from 2002 to 2019, this paper investigates the characteristics and evolution mechanism of the industrial robots trade network of the Belt and Road. The study finds that there exists regional heterogeneity in the trade network, showing a typical core-periphery structure. The node status of the network is evolving steadily while changing. Asia has developed into the “core” of the industrial robots trade network, and forms a multi-polarization pattern with China as the hub. The endogenous mechanism test shows that the reciprocity structure effect drives node pairs to actively engage in reciprocal trade; the negative expansion structure effect drives outdegree star nodes to inhibit the increase of dependent nodes; the transitive closure structure effect drives the network to produce transitive closure triangle, promoting the formation of trade clusters; the time dependence effect not only maintains the relative stability of trade relations, but also promotes the formation of new relations. The economy volume, industrial employment proportion, manufacturing technology level and human capital level have the sender (receiver) effect, Geographical, linguistic and trade organization proximity promote trade cooperation. Since the implementation of the Belt and Road Initiative, the endogenous structure effect and convergence effect of nodes have changed significantly.