Endpoint Sensor Data Prediction for Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated sensor data collection systems face challenges in managing increased communication bandwidth and data storage requirements due to high-frequency data collection from a large number of sensors, leading to inefficient resource utilization and potential data collisions.
Innovation Solution
Endpoint devices collect data at fine granularity but report at coarse intervals, allowing the central data collection point to predict values and send exception reports only when predictions exceed a tolerance, reducing unnecessary data transmission and storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If high-frequency data collection is implemented from a large number of sensors, then real-time or near real-time data availability is improved, but communication bandwidth requirements and data storage burdens increase significantly
Solution Approach 1:
The patent extracts only the essential information by having endpoint devices predict sensor values and only transmit data when predictions exceed a tolerance threshold. This removes unnecessary data transmissions while maintaining data availability, directly resolving the contradiction between fast data availability and bandwidth consumption
Solution Approach 2:
The system performs preliminary prediction of sensor values at endpoint devices before actual data collection. By predicting what the sensor values will be and comparing with actual readings, the system prepares in advance to avoid unnecessary transmissions, thus improving data availability speed without proportionally increasing bandwidth usage
2Measurement precision
If data collection frequency is increased to support demand billing and real-time monitoring, then measurement precision and system responsiveness are improved, but communication overhead and data collision risks increase
Solution Approach 1:
The patent implements periodic data transmission based on exception conditions rather than continuous high-frequency collection. Data is transmitted periodically when prediction errors exceed thresholds, maintaining measurement precision for billing while reducing communication overhead and data collision risks associated with constant high-frequency transmissions
3Reliability
If frequent sensor readings are collected and stored centrally, then data accuracy for operational decisions is improved, but data storage requirements and processing loads increase
Solution Approach 1:
The patent applies local quality by performing prediction and exception detection at the distributed endpoint devices rather than centrally. Each endpoint device independently determines what data needs transmission based on local prediction accuracy, reducing the volume of data that must be stored and processed centrally while maintaining data accuracy for operational decisions
Data Source
AI summary
Endpoint device, central data collection point, and associated methods for collecting data over a communication network between endpoints and the central collection point. Actual measurements from a sensor are obtained by the endpoint device at a relatively fine time granularity. The endpoint device generates reports for receipt by a central data collection point. The reports include regular reports containing a portion of the actual measurements representing sensor measurements at a relatively coarse time granularity, and exception reports, containing information representing one or more of the actual measurements that differ in frequency or granularity of regular report measurements. Each of the exception reports is generated in response to a determination that at least one of the actual measurements differs from a predicted value for that at least one of the one or more actual measurements by an amount that exceeds a pre-established limit.