Marine Data Prioritization for Machine Learning Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in efficiently collecting and transmitting relevant data from marine vessels for training machine learning algorithms, particularly in the marine environment where remote communication is difficult due to coverage, quality, latency, connectivity, and cost issues.
Innovation Solution
A method and system that utilize sensors on marine vessels to collect data representative of encountered situations, prioritize this data based on relevance criteria, and transmit it via remote data communication links when available, ensuring that high-priority data is transmitted efficiently for training machine learning algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If all sensor data are transmitted regardless of relevance, then data completeness is improved, but transmission cost and bandwidth consumption increase
Solution Approach 1:
The patent extracts only the most relevant data from the complete sensor dataset based on predefined relevance criteria (such as proximity to hazards, weather conditions, or navigation critical parameters). This extraction process filters out redundant or less important data, enabling selective transmission that reduces bandwidth consumption while maintaining essential information for machine learning training.
Solution Approach 2:
The patent applies different transmission strategies to different data based on their local relevance characteristics. High-relevance data (e.g., data from sensors detecting potential hazards or critical navigation parameters) are prioritized for transmission, while low-relevance data are filtered out. This local quality differentiation optimizes the transmission volume by focusing resources on the most important data segments.
2Productivity
If data are prioritized and filtered before transmission, then transmission efficiency is improved, but data representativeness may be compromised
Solution Approach 1:
The patent implements a feedback mechanism where the relevance of transmitted data is continuously evaluated against the original dataset characteristics. This feedback loop ensures that the filtering and prioritization processes maintain adequate data representativeness by adjusting selection criteria based on observed patterns in the data, preventing over-filtering that would compromise the training quality of machine learning algorithms.
Solution Approach 2:
The patent performs preliminary actions by pre-defining relevance criteria and establishing data prioritization rules before the transmission process begins. These pre-established guidelines ensure that data selection is systematic and unbiased, maintaining representativeness while improving transmission efficiency. The preliminary setup includes defining which sensor types, time periods, or parameter ranges should be prioritized based on operational requirements.
3Loss of information
If remote data communication is used for marine vessels, then data accessibility is improved, but communication reliability deteriorates due to coverage and connectivity issues
Solution Approach 1:
The patent performs preliminary data processing and prioritization actions on the marine vessel before transmission to shore. By pre-filtering and ranking data based on relevance criteria, the system ensures that when communication becomes available, the most critical data are ready for immediate transmission. This preliminary action reduces the impact of communication delays or interruptions by minimizing the total data volume that needs to be transmitted.
Solution Approach 2:
The patent segments the data transmission process into priority-based batches rather than attempting to transmit all data simultaneously. High-priority data are transmitted first when communication becomes available, followed by lower-priority data. This segmentation strategy improves communication reliability by ensuring critical information is delivered even if overall transmission is interrupted or delayed, while maintaining data accessibility through the prioritized transfer mechanism.
Data Source
AI summary
A method comprising, by at least one processing unit, obtaining data collected by one or more sensors of at least one marine vessel, said data being representative of one or more situations encountered by the marine vessel during its voyage, prioritizing data according to at least one relevance criterion, wherein when the marine vessel is located in a zone in which at least one remote data communication link meets a criterion, transmitting at least some of the data using the at least one remote data communication link, wherein data are transmitted according to priority determined for the data, thereby facilitating transmission of relevant data for the purpose of training one or more machine learning algorithms (e.g. deep learning algorithms) providing output based on these data.