Embedded smart bearing is a kind of structural unit with the characteristics of self sensing, self diagnosis, self-adaptive and self-regulation, which is one of the development directions of high-end bearing at home and abroad. Based on Ioannides-Harris theory, the load distribution of rolling bearing is analyzed. Combined with the finite element simulation model, the influence of outer ring grooving on bearing reliability is analyzed, and the radial grooving structure of outer ring of rolling bearing is determined. According to the theoretical analysis and numerical simulation, the embedded bearing vibration monitoring system is designed by combining the vibration sensor with the bearing body. By using the bearing life testing machine, the bearing life test is carried out and the vibration signals of smart bearings with different structures are collected. The experimental results show that the amplitude of vibration signal of smart bearing is higher than that of traditional bearing pedestal monitoring method, and it can effectively monitor the running state of the bearing; compared with the structure of axial grooving, the service life of the proposed smart bearing is similar to that of the ordinary bearing, which provides a new method for the condition monitoring and fault diagnosis of the rolling bearing.