Research Data Gathering Using Ancillary Codes and ML Prediction

The patent creates a virtual copy of the monitoring device's data collection capability by training a machine learning model to replicate the device's reading patterns. When the physical device fails to read ancillary codes completely, the virtual copy generates predicted readings to supplement the incomplete data, thereby maintaining data completeness without requiring additional physical monitoring devices.

The system performs preliminary training of the machine learning model using historical reading data from monitoring devices before actual data collection begins. This preliminary action enables the model to learn the characteristics of ancillary codes and reading patterns in advance, so that when equipment failures or environmental interference occur during actual operation, the pre-trained model can immediately generate accurate predictions to complete the data sets.

The patent replaces the mechanical approach of deploying multiple physical monitoring devices with an intelligent software-based solution. Instead of using multiple hardware devices to read ancillary codes and hope for complete data collection, the system uses a single monitoring device augmented by a machine learning model that predicts missing data, thereby reducing hardware complexity while maintaining or improving data accuracy.

The machine learning model acts as an intermediary between the physical monitoring device and the final data set. Rather than directly relying on multiple physical devices to capture all data, the intermediary model processes the partial readings from a single device and generates the complete data set by predicting missing information, thereby simplifying the overall system architecture.