Sensor Reading Verification via Associated Sensor Querying
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In modern systems with multiple sensors, such as server farms and IoT systems, sensor malfunctions can lead to inaccurate readings, causing costly responses to false alarms and delaying detection of true problems due to the time it takes to identify and discard erroneous data, especially when dependence relationships between sensors are not considered.
Innovation Solution
The system leverages dependence relationships between sensor readings to rapidly detect and flag inaccurate readings by querying additional sensors, adjusting query rates based on trustworthiness values determined through predefined conditions and machine-learning models, and triggering appropriate remedial actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If sensor query rate is increased to detect malfunctions faster, then detection speed improves, but system resource consumption and processing load increase
Solution Approach 1:
The system pre-identifies associated sensors and their dependence relationships before malfunctions occur. When a target sensor reading is received, the system already has a ready list of associated sensors to query, eliminating the need for real-time analysis and enabling rapid verification without ad-hoc computational overhead.
Solution Approach 2:
The system introduces an intermediary verification layer that uses associated sensors as mediators to validate target sensor readings. Instead of directly analyzing the target sensor data for anomalies, the system queries associated sensors whose readings can indirectly verify the target reading's accuracy, reducing the need for frequent direct querying of the target sensor itself.
2Measurement precision
If multiple sensors are queried to verify readings, then measurement accuracy improves, but device complexity and processing time increase
Solution Approach 1:
The verification process is segmented into distinct phases: (1) receiving target sensor reading, (2) identifying associated sensors based on pre-established relationships, (3) querying only those specific associated sensors, and (4) comparing readings. This segmentation allows the system to verify readings systematically without overwhelming complexity, as each phase handles a specific subset of operations.
Solution Approach 2:
The system applies different verification strategies to different sensors based on their specific dependence relationships. Each target sensor has its own customized set of associated sensors identified through machine learning models, allowing the verification process to be tailored to local characteristics of each sensor group rather than applying a uniform complex verification protocol to all sensors.
3Reliability
If sensor readings are verified using associated sensors, then reliability of detection improves, but loss of time due to additional queries increases
Solution Approach 1:
The machine learning models that identify associated sensors and their dependence relationships are trained and configured in advance, before actual sensor verification is needed. This preliminary action creates a ready-reference framework that enables rapid verification during operation, as the system doesn't need to perform complex analysis in real-time but simply queries pre-identified associated sensors.
4Productivity
If query rate is dynamically adjusted based on trustworthiness, then productivity improves, but device complexity increases due to dynamic adjustment mechanisms
Solution Approach 1:
The system dynamically adjusts the query rate for sensors based on their calculated trustworthiness values. Sensors with higher trustworthiness are queried less frequently, while those with lower trustworthiness are queried more often. This dynamic adjustment is driven by automatic trustworthiness calculations rather than manual configuration, allowing the system to adapt to changing sensor performance over time without increasing operational complexity.
Data Source
AI summary
Systems and methods described herein leverage dependence relationships between sensor readings to facilitate rapid detection of erroneous sensor readings and rapid response times when accurate sensor readings indicate events that call for remedial action. When a target sensor reports a questionable sensor reading, systems described herein can query additional sensors whose readings share expected dependence relationships with readings from the target sensor. Systems described herein determine whether the questionable sensor reading is trustworthy based on the expected dependence relationships and additional sensor readings received from the additional sensors. If the questionable sensor reading is trustworthy, the system can trigger appropriate remedial action and increase query rates for the additional sensors (and the target sensor) to monitor an underlying event more closely. If the questionable reading is not trustworthy, the system can trigger appropriate remedial action (e.g., by deactivating the sensor or sending an alert to an administrator).