Context-Aware Sensor Selection for Data Completeness
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Ensuring data completeness across various data gathering contexts is challenging due to varying standards and sensor relevance, leading to inconsistent and incomplete data sets.
Innovation Solution
Utilizing Machine Learning (ML) models to identify relevant data streams and sensor attributes, reposition sensors as needed, and initiate corrective actions to ensure data completeness by assessing and correcting sensor positions, data quality, and format, and initiating maintenance when necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple sensors are deployed to monitor various activities in a Data Gathering Zone, then data coverage and monitoring capability are improved, but data completeness for specific activities deteriorates due to sensor irrelevance and varying standards
Solution Approach 1:
The patent segments the monitoring system into activity-specific sensor groups. Each activity type (planting, harvesting, etc.) has its own designated sensors that are activated only when that activity is detected. This segmentation ensures that only relevant sensors contribute to data completeness for each specific activity, resolving the contradiction between having many sensors for coverage versus needing only relevant sensors for completeness.
Solution Approach 2:
The system dynamically adjusts which sensors are active based on the current activity being performed in the Data Gathering Zone. The activity recognition module identifies the current activity and dynamically enables/disables sensor groups accordingly. This dynamic adaptation allows the system to maintain high data completeness for the current activity while still having the capability to monitor multiple different activities over time.
2Area of stationary object
If sensors are positioned to maximize general monitoring coverage, then area coverage is improved, but data quality for specific activities deteriorates due to suboptimal positioning
Solution Approach 1:
The patent implements local quality by having different sensors optimized for different local conditions and activity types. Each sensor is positioned and configured to provide optimal data quality for its specific activity, rather than attempting to provide general coverage. The system selects and positions sensors based on the specific activity requirements, ensuring high measurement precision for each activity type in its local context.
3Adaptability or versatility
If manual configuration is used to determine relevant sensors and data attributes for each activity, then data completeness standards can be customized, but system complexity and configuration time increase
Solution Approach 1:
The system performs self-service by automatically determining which sensors and data attributes are relevant for each activity type. The activity recognition module and data stream identification module work together to automatically configure the monitoring system based on the detected activity, eliminating the need for manual configuration. This self-service approach maintains adaptability to different activities while significantly reducing system complexity and configuration time.
Solution Approach 2:
The patent implements preliminary action by pre-configuring activity profiles that define which sensors and data attributes are relevant for each activity type. These profiles are established in advance and stored in the system. When an activity is detected, the system simply retrieves and applies the corresponding pre-configured profile, enabling quick adaptation to different activities without complex real-time configuration decisions.
4Speed
If all sensors continuously monitor all activities, then response time to activity changes is improved, but energy consumption increases
Solution Approach 1:
The system uses periodic action by continuously monitoring for activity changes and periodically adjusting sensor activation based on the current activity state. Rather than having all sensors run continuously, the system periodically checks for activity transitions and activates/deactivates sensor groups accordingly. This maintains fast response to activity changes while significantly reducing overall energy consumption by keeping sensors inactive when not needed.
Data Source
AI summary
A computer collects data relevant to a primary activity occurring in a data gathering zone. The computer receives from an Activity Indication Source (AIS), an indication of a primary activity associated with a Data Gathering Zone (DGZ). The computer determines, uses a first Machine Learning (ML) model to identify data streams relevant for data gathering activities, a Target Set of Data Streams (TSDS) associated with the primary activity. The computer identifies within a set of sensors having a monitoring arrangement with the DGZ, a set of preferred sensors to provide the TSDS. The computer determines using a second Machine Learning (ML) model trained to identify Preferred Data Stream Attributes (PDSAs) for data gathering activities, a target set of PDSAs. The computer retrieves, a data stream relevant to the TSDS. The computer compares attributes of the data stream to the PDSAs and initiates corrective action as necessary.