Machinery OEMs often have to choose between different bearing suppliers when designing a new device. Although the same bearing model from different suppliers may have the same external form factor and load rating, there are often differences in internal geometry, material quality, and surface treatment from one supplier to another that will cause the performance and expected life to vary in the field. These slides will outline Sentient Science’s approach towards measuring and quantifying these hidden bearing parameters. We show how we combine material, geometry, and surface treatment metrics into high-fidelity predictive models which we use to help OEMs decide which bearing supplier to choose for their application.