Vehicle Sensor Type Identification via Data Pattern Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In automotive vehicles, unidentified sensors connected to a controller area network (CAN) lack clear association with their respective types, making it difficult for drivers or technicians to determine the specific type of sensor without manual identification.
Innovation Solution
An external device that obtains input data streams from unidentified sensors, detects changes in vehicle parameters, and identifies the sensor type by associating changes in data patterns, using reinforcement learning or convolutional neural networks to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If sensors are connected to the CAN network with unique identifiers, then sensor data can be collected and transmitted, but the sensor types cannot be automatically identified without manual intervention
Solution Approach 1:
The system performs preliminary actions by collecting training data during normal vehicle operation, storing data patterns associated with specific sensor types before actual identification is needed. This pre-collection of data enables automatic identification without manual intervention during diagnostic operations.
Solution Approach 2:
The system uses feedback mechanisms where sensor data patterns are continuously monitored and compared against stored training data. The identification process provides feedback loops that refine the matching between observed data patterns and known sensor types, enabling accurate automatic identification.
2Measurement precision
If manual identification of sensor types is performed, then accurate sensor type determination is achieved, but time and labor resources are consumed
Solution Approach 1:
The system implements self-service by automatically identifying sensor types without requiring technician intervention. The identification process uses stored training data and observed data patterns to autonomously determine sensor types, eliminating manual identification steps and associated time losses.
Solution Approach 2:
The system creates copies of sensor data patterns from training sessions and stores them for comparison during identification. By working with copied data patterns rather than requiring direct manual analysis of live sensor data, the system achieves rapid automatic identification while maintaining accuracy.
3Adaptability or versatility
If training data is collected during vehicle operation, then sensor type identification can be automated, but additional data collection and processing requirements are created
Solution Approach 1:
The system achieves multi-functionality by using the same data collection infrastructure for both normal vehicle operation monitoring and sensor identification training. The existing sensor network and data processing pathways serve dual purposes, reducing the need for separate dedicated training data collection systems.
Solution Approach 2:
The system merges the training data collection process with regular vehicle operation data collection. By combining these functions into a unified data gathering and processing framework, the system reduces overall complexity while maintaining the ability to collect comprehensive training data for sensor identification.
Data Source
AI summary
A method for identifying a type of a sensor includes obtaining a first input data stream from the unidentified sensor in a vehicle, wherein the input data stream includes an identifier representing the unidentified sensor and a first data pattern associated with the identifier, detecting a change in a vehicle parameter of the vehicle, obtaining a second input data stream from the unidentified sensor in the vehicle, wherein the second input data stream includes the identifier and a second data pattern associated with the identifier, the second data pattern being different from the first data pattern, associating the change in the vehicle parameter of the vehicle with a change from the first data pattern to the second data pattern, and identifying the type associated with the unidentified sensor based on detecting the change in the vehicle parameter.