Effective monitoring and diagnosis of manufacturing processes are of critical importance for manufacturing processes. If critical conditions are continuously monitored and controlled, problems can be detected and solved during the processing cycle, resulting in less damage to tools, less downtime for machine repair and maintenance, higher productivity, and less energy consumption. Driven by a widespread need for better process monitoring and diagnosis techniques, signicant efforts have been taken to advance sensors and data analysis technology for manufacturing processes. However, current technology still evidently lags behind practical needs. The conventional sensors used are normally large in size and are either attached to the surface that might be far away from critical locations to avoid interference with the operation of the machine or destructively inserted into critical locations through appropriate channels in the components. As a result, it is difcult to provide measurement with a high-spatial and temporal resolution at distributed critical locations.