The emerging graph neural networks for intelligent fault diagnostics and prognostics: A guideline and a benchmark study

•A practical guideline on leveraging graph neural networks (GNNs) for realizing intelligent fault diagnostics and prognostics.•A novel intelligent fault diagnostics and prognostics framework based on GNNs is established to illustrate how the proposed guideline works.•In this framework, three types graph construction methods are provided, and seven kinds of graph convolutional networks with four graph pooling methods are investigated.•A benchmark study of these models is carried out on eight datasets, including six fault diagnosis datasets and two prognosis datasets.