Sensor Fusion for Occupancy Estimation via Synthetic Variables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods lack an efficient way to estimate and measure interest, activity, and occupancy levels at physical locations, which are crucial for optimizing space usage, advertising effectiveness, and operational management in commercial and industrial settings.
Innovation Solution
A system and method utilizing a plurality of sensors to collect and process physical parameters such as temperature, humidity, sound, and motion data, generating synthetic variables, and employing machine learning models to calculate and display interest, activity, and occupancy indicators.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to measure interest, activity, and occupancy at physical locations, then the system is simple to implement, but the measurement precision and accuracy of interest levels are insufficient
Solution Approach 1:
The patent segments the measurement system into multiple independent sensor components, each measuring specific physical parameters (temperature, humidity, pressure, sound, motion, etc.). This segmentation allows for precise measurement of individual parameters while maintaining system modularity, resolving the contradiction between measurement precision and device complexity.
Solution Approach 2:
The patent introduces synthetic variables as intermediary constructs that bridge raw sensor data and high-level indicators. These synthetic variables process and combine multiple physical parameters to represent complex concepts like 'interest level,' enabling precise measurement without directly complicating the sensor network architecture.
2Measurement precision
If multiple sensors are deployed to capture comprehensive physical parameters, then the measurement accuracy improves, but the device complexity and data processing requirements increase
Solution Approach 1:
The patent implements a universal data processing framework that handles multiple physical parameters through consistent synthetic variable construction and machine learning model application. This multi-functional approach allows the same processing architecture to accommodate various sensor types and parameters, improving measurement accuracy while controlling system complexity through standardization.
Solution Approach 2:
The patent transforms raw physical parameters into synthetic variables through mathematical processing and combination. This parameter transformation converts diverse sensor readings into unified indicators that feed into machine learning models, enabling accurate measurement of abstract concepts like 'interest level' without proportionally increasing system complexity.
3Measurement precision
If machine learning models are used to generate indicators from sensor data, then the estimation accuracy of interest and occupancy levels improves, but the computational requirements and processing time increase
Solution Approach 1:
The patent performs preliminary data processing by constructing synthetic variables from raw sensor data before feeding them into machine learning models. This pre-processing step organizes and condenses information in advance, reducing the computational burden during real-time inference and minimizing processing time while maintaining estimation accuracy.
Solution Approach 2:
The patent selectively processes only the most relevant physical parameters and synthetic variables needed for specific indicators, rather than processing all available data uniformly. This partial processing approach reduces computational requirements and processing time while maintaining sufficient accuracy for practical decision-making purposes.
Data Source
AI summary
Techniques for determining levels of interest, activity, or occupancy at a physical location can include receiving data corresponding to physical parameters sensed by a plurality of sensors at the physical location. The physical parameters can include temperature, humidity, pressure, sound, distance to an object, visible light, infra-red light, motion of objects, acceleration, magnetic field, vibration, and radio signals. Synthetic variables can be generated based on the received data and can represent a processed or combined value for its corresponding physical parameters. The physical parameters and synthetic variables can be stored in a memory device. One or more indicators for a level of: (i) interest, (ii) activity, or (iii) occupancy at the physical location can be generated based on the received data and the one or more synthetic variables by utilizing a machine learning model and output to a user computing device for display in a user interface.