DTC Rulebook Generation for Predictive Vehicle Maintenance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Vehicle Health Management (VHM) solutions are primarily based on preventive and reactive maintenance, providing information only after a malfunction has occurred, necessitating a predictive solution that can discover Diagnostic Trouble Code (DTC) patterns within telemetric data for early detection of malfunctions.
Innovation Solution
A Diagnostic Trouble Code (DTC) rulebook generation system that processes telematics trace data and malfunction occurrence data to extract ride records, label them as healthy or faulty, train machine learning models, and generate DTC rules, which are then compiled into rulebooks for predictive maintenance, using models like logistic regression, decision trees, and sequence models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional preventive and reactive maintenance solutions are used, then the system provides information after malfunction occurs, but it cannot detect malfunctions early
Solution Approach 1:
The system performs preliminary actions by continuously analyzing telemetric data and DTC patterns before malfunctions occur. Machine learning models are trained on historical data to predict potential issues, enabling the system to alert operators about impending failures before they actually happen, thus transitioning from reactive to predictive maintenance
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring vehicle telemetric data and comparing it against learned patterns from training data. The machine learning models provide feedback loops where prediction results are validated against actual malfunction outcomes, continuously improving detection accuracy and enabling early warning signals before failures occur
2Measurement precision
If machine learning models are trained on all telematics data, then detection accuracy improves, but processing time and computational resources increase
Solution Approach 1:
The training process is segmented into multiple stages: initial model training on a subset of representative telemetric data, followed by iterative refinement using only new anomaly patterns discovered after each deployment cycle. This segmentation allows the system to achieve high detection accuracy without requiring continuous processing of entire historical datasets
Solution Approach 2:
The system applies partial action by training models on carefully selected subsets of telemetric data that are most representative of actual malfunction patterns. Rather than processing all available data continuously, the system processes only relevant new patterns discovered in each monitoring cycle, reducing computational overhead while maintaining detection accuracy
3Measurement precision
If DTC rules are generated with high precision threshold, then false positives decrease, but the number of detected malfunctions decreases
Solution Approach 1:
The system implements dynamic precision thresholds that adapt based on contextual factors such as vehicle type, operating conditions, and historical malfunction patterns. The precision requirement is made flexible rather than fixed, allowing the system to adjust detection sensitivity dynamically to balance false positives against detection coverage across different operational scenarios
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A Diagnostic Trouble Code (DTC) rulebook generation system comprising a processing circuitry configured to: obtain: telematics trace data records and malfunction occurrence data records; and extract one or more ride records from the obtained telematics trace data records and the obtained malfunction occurrence data records, wherein at least one ride record of the ride records is for a given vehicle; label at least one ride record of the ride records as a healthy ride or a faulty ride, wherein a ride record associated with a given vehicle where no malfunction occurrence data records are associated with the given vehicle is labeled as a healthy ride, otherwise the ride record is labeled as a faulty ride; train one or more machine learning models on one or more subsets of the labeled ride records; determine one or more DTC rules utilizing at least one of the trained machine learning models.