Vehicle Defect Detection Using Sensor Inference

The patent uses sensor systems on externally visible components to create information about the state of internal components without directly sensing them. The sensor data from accessible locations is used to infer the condition of hidden components, effectively copying the diagnostic information from where sensors can be placed to where defects need to be detected.

The patent introduces an intermediary diagnostic approach where sensor systems on externally accessible components serve as mediators to detect defects in internal components. Instead of placing sensors directly on all components, the system uses intermediate measurements from accessible locations to infer the state of inaccessible components through signal processing and evaluation.

The patent implements preliminary defect detection by continuously monitoring sensor systems during normal operation and immediately after accidents. The evaluation unit processes sensor data in real-time to identify defects before repair is needed, and can automatically trigger emergency calls. This preliminary detection eliminates the need for time-consuming expert appraisals later, as defects are identified and reported automatically.

The diagnostic system performs self-diagnosis through automated sensor evaluation and defect identification. The evaluation unit automatically processes sensor signals, determines whether critical situations have occurred, and can independently trigger emergency notifications without requiring external expert intervention for the detection phase.