Edge Network Event Mapping for Granular IoT Device Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies have limitations in practical utility, sophistication, granularity, and accuracy in managing devices over networks, particularly in 'Internet of Things' applications, failing to leverage network connectivity for efficient collection and utilization of device-related information.
Innovation Solution
Instruments devices with multimodal sensors to collect data, processes it, and uses a network management engine with an event database and machine learning to match sensor data with events, enabling precise event detection and actionable responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional network management approaches are used to collect device data, then device connectivity is maintained, but measurement precision and granularity of device information are insufficient
Solution Approach 1:
The patent segments device monitoring into multiple hierarchical levels: individual sensor data points, sensor groupings by function, device-level aggregation, and network-level consolidation. This segmentation enables precise capture of granular device information while organizing it into manageable units for analysis and action.
Solution Approach 2:
The patent adds temporal and contextual dimensions to device data collection. By incorporating time-series sensor readings, environmental context, and device state information, the system transforms simple presence/absence data into multi-dimensional operational profiles that enable precise event detection and granular device management.
2Measurement precision
If more sensors and monitoring capabilities are added to devices, then event detection accuracy improves, but device complexity increases
Solution Approach 1:
The patent introduces event databases and network management engines as intermediaries between sensors and decision-making systems. These intermediaries aggregate, filter, and contextualize sensor data, reducing the burden on individual devices while improving overall event detection accuracy through centralized intelligence.
Solution Approach 2:
Devices are equipped with local event databases that enable them to autonomously detect and respond to events without constant network connectivity or centralized control. This self-service capability allows devices to maintain high detection accuracy while reducing system complexity by distributing intelligence to the edge.
3Reliability
If continuous monitoring of all device parameters is implemented, then event detection capability improves, but energy consumption increases
Solution Approach 1:
The patent implements periodic sampling of sensor data with variable intervals based on device state and event criticality. Rather than continuous monitoring, the system adjusts sampling frequency dynamically, maintaining reliable event detection for critical parameters while reducing power consumption during stable operational states.
Solution Approach 2:
The patent applies partial monitoring strategies where only critical device parameters and high-priority sensors are continuously monitored, while less critical parameters are sampled periodically or only when triggered by contextual events. This selective approach maintains sufficient event detection capability while significantly reducing overall energy consumption.
4Measurement precision
If centralized network management is used to process all device data, then event matching accuracy improves, but network bandwidth requirements and processing time increase
Solution Approach 1:
The patent segments event processing into local and centralized components. Individual devices perform local event detection using their own sensor data and local event databases, while the centralized network management engine handles cross-device pattern recognition and complex event correlation. This segmentation reduces network bandwidth requirements and processing time by distributing computational load.
Solution Approach 2:
The patent pre-loads event databases with anticipated event patterns, thresholds, and response protocols before they are needed. By having event matching criteria prepared in advance, the system can rapidly compare incoming sensor data against pre-defined patterns, improving event matching accuracy while minimizing real-time processing time and network communication overhead.
Data Source
AI summary
Network management systems and methods are provided. A system is provided that includes an event database that provides a mapping of multimodal sensor data to events of devices coupled to a network, such as operational or behavioral events. A network management engine obtains a set of multimodal sensor data relating to a device, which may include sensor measurement or output data relating to each of multiple device operation or behavior parameters. The engine determines, utilizing the mapping, a match of the set of multimodal sensor data to a specific event associated with the device. Based at least in part on the determined match, the engine causes generation or updating of event data associated with the specific event, the data being accessible by network management software of the network management system.