Sensor Data Fusion Using Curation, Linking, and Conditional Entropy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensor data fusion systems face challenges in accurately fusing heterogeneous, partially heterogeneous, or homogeneous data sources, leading to inefficiencies in computational processing, storage demands, and the inability to create actionable data.
Innovation Solution
A system and method for sensor data fusion that utilizes a computer processor with multiple engines (curation, link, fusion, inference, and validation) to curate, link, and fuse sensor data from various sources, creating a unique dataset with new data points and validating their accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensor data from multiple sources is fused using traditional methods, then data integration is achieved, but computational processing efficiency deteriorates and storage demands increase
Solution Approach 1:
The patent segments the sensor data fusion process into distinct functional modules: a curation engine that pre-processes and validates individual sensor data streams, a link engine that establishes relationships between curated data, and a fusion engine that combines linked data. This segmentation allows each module to operate independently and efficiently, reducing overall computational complexity while maintaining fusion accuracy.
Solution Approach 2:
The curation engine performs preliminary actions by pre-processing, validating, and curating sensor data before it reaches the fusion engine. This preliminary curation reduces the computational burden on subsequent fusion operations by ensuring data quality and relevance upfront, thereby improving processing efficiency without compromising fusion accuracy.
2Reliability
If traditional sensor data fusion methods are used, then data integration is achieved, but storage requirements increase
Solution Approach 1:
The link engine extracts only the relevant relationships and features from curated sensor data that are necessary for fusion, rather than storing and processing all raw data. This extraction of essential information reduces storage requirements while maintaining the accuracy needed for effective data fusion.
Solution Approach 2:
By performing curation and linking operations before fusion, the system prepares data in a condensed, relationship-rich format that requires less storage space. The preliminary organization of data into curated and linked structures eliminates redundant information storage while preserving fusion accuracy.
3Productivity
If heterogeneous sensor data is fused without validation, then processing speed is maintained, but data accuracy deteriorates
Solution Approach 1:
The curation engine performs validation and quality checks on sensor data before it enters the fusion process. This preliminary validation ensures data accuracy is established upfront, allowing the fusion engine to operate at full speed without needing to perform additional validation checks during fusion, thus maintaining both speed and accuracy.
4Loss of information
If comprehensive sensor data fusion is performed, then actionable data is generated, but system complexity increases
Solution Approach 1:
The fusion system is segmented into specialized engines (curation, link, fusion) that each handle specific aspects of data processing. This segmentation reduces system complexity by assigning clear responsibilities to each module, making the overall comprehensive fusion process more manageable and maintainable while still generating actionable data.
Solution Approach 2:
Each engine in the segmented architecture serves multiple functions within its domain: the curation engine validates, cleans, and organizes data; the link engine establishes relationships and contexts; the fusion engine combines data and generates actionable insights. This multi-functionality within segmented modules reduces overall system complexity compared to a monolithic fusion system.
Data Source
AI summary
Systems and methods of sensor data fusion including sensor data capture, curation, linking, fusion, inference, and validation. The systems and methods described herein reduce computational demand and processing time by curating data and calculating conditional entropy. The system is operable to fuse data from a plurality of sensor types. A computer processor optionally stores fused sensor data that the system validates above a mathematical threshold.