Reliability assessment and maintenance scheduling of flowmeters are of great importance for process industry companies because of the need to ensure the production quality and reduce operational risks and costs. In practice, the failure process of flowmeters is complicated due to multiple failure modes arising from different electro-mechanical parts. Besides, there generally exists heterogeneity among the flowmeters that is caused by several factors (i.e., covariates) such as operational conditions. In this paper, we use the nonhomogeneous Poisson process (NHPP) model with covariates to assess the reliability of flowmeters and then derive the optimal age-based preventive maintenance (PM) policy based on the estimated reliability model. The effects of covariates are incorporated in the reliability model by means of a proportional intensity function. We apply the maximum likelihood method to estimate the model parameters and adopt the random weighted likelihood bootstrap procedure to address the statistical uncertainty. Real-world flowmeters failure data from a process industry company are used in the case study. The estimated intensity functions are shown to reasonably fit the failure data and the obtained optimal PM policies provide cost-efficient and covariate-specific maintenance strategy for the company.