Virtual Sensor Assembly for Low-Bandwidth IoT Event Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current smart environment and IoT systems rely on special-purpose sensors that are limited in scope and interoperability, leading to fragmented data and high upgrade costs, failing to achieve a holistic smart home experience due to bandwidth and processing speed issues.
Innovation Solution
A ubiquitous sensing system that uses one or more sensors capable of direct or indirect event detection, with a sensor assembly that can communicate with a server for processing, employing machine learning to characterize sensor data and create virtual sensors that can detect events from combined data streams without requiring direct instrumentation of objects or people.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional special-purpose sensors are used to measure individual environmental aspects, then each sensor can be optimized for its specific function, but the system becomes fragmented with limited interoperability and comprehensive monitoring capability
Solution Approach 1:
The patent applies universality by creating a unified sensing system where a single sensor assembly with multiple sensors can detect various environmental facets (sound, vibration, temperature, humidity, light, motion). The system serves multiple functions through one integrated platform, enabling comprehensive environmental monitoring without requiring separate specialized sensors for each parameter, thus improving interoperability while maintaining detection accuracy.
2Loss of information
If all sensed data is transferred to a backend server for processing, then centralized processing can provide comprehensive data analysis, but bandwidth usage and processing speed are compromised
Solution Approach 1:
The patent segments the data processing task by performing feature extraction and event detection locally at the sensor assembly before transmitting data to the server. The microcontroller extracts features from raw sensor data and identifies events locally, transmitting only processed information rather than raw data streams. This segmentation reduces bandwidth consumption while maintaining complete data analysis capability through distributed processing architecture.
Solution Approach 2:
The patent extracts and processes critical features and event information at the local sensor assembly level before transmission to the backend server. By taking out the feature extraction and event detection functions from the server and placing them at the sensor assembly, the system reduces the amount of data transmitted while preserving essential information for comprehensive environmental monitoring.
3Adaptability or versatility
If multiple sensors are deployed to sense diverse environmental facets, then comprehensive monitoring is achieved, but device complexity and upgrade costs increase
Solution Approach 1:
The patent merges multiple sensors (microphone, vibration sensor, temperature sensor, humidity sensor, light sensor, motion sensor) into a single integrated sensor assembly that can detect diverse environmental facets. By combining these sensors in one unit with shared processing capabilities, the system achieves comprehensive environmental monitoring while reducing overall system complexity compared to deploying separate specialized sensor systems for each parameter.
4Productivity
If machine learning algorithms process sensor data locally at the sensor assembly, then processing speed and bandwidth efficiency improve, but computational resources at the sensor assembly must be sufficient
Solution Approach 1:
The patent applies partial action by implementing machine learning-based feature extraction and event detection algorithms at the sensor assembly level to process only the most critical data transformations locally. This partial computational action at the sensor assembly improves processing speed and reduces bandwidth usage, while the complexity is managed by performing only the necessary computational tasks rather than all possible processing operations.
Data Source
AI summary
A sensing system includes a sensor assembly that is communicably connected to a computer system, such as a server or a cloud computing system. The sensor assembly includes a plurality of sensors that sense a variety of different physical phenomena. The sensor assembly featurizes the raw sensor data and transmits the featurized data to the computer system. Through machine learning, the computer system then trains a classifier to serve as a virtual sensor for an event that is correlated to the data from one or more sensor streams within the featurized sensor data. The virtual sensor can then subscribe to the relevant sensor feeds from the sensor assembly and monitor for subsequent occurrences of the event. Higher order virtual sensors can receive the outputs from lower order virtual sensors to infer nonbinary details about the environment in which the sensor assemblies are located.