Occupancy tracking using sound recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing HVAC systems cannot accurately determine the number of people present in a space or identify individuals, limiting their ability to provide personalized temperature settings and efficient energy management.
Innovation Solution
An occupancy tracking system that uses audio signal processing, device detection, and wireless signal strength analysis to predict the number of people in a space, distinguishing between human voices and electronic device sounds, and adjusts HVAC settings accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If proximity sensors or motion detection sensors are used to determine occupancy, then the system can detect whether a space is occupied, but it cannot determine the number of people present or identify individuals
Solution Approach 1:
The patent replaces mechanical/proximity sensors with acoustic field-based detection. The system uses microphones to capture audio signals and processes these signals to extract occupancy information, including the number of people and their identities, thereby substituting mechanical detection with acoustic field analysis.
Solution Approach 2:
The system changes the detection parameter from binary occupancy status to continuous acoustic signal analysis. By analyzing audio signal characteristics such as frequency, amplitude, and temporal patterns, the system can determine the number of people and identify individuals, transforming the detection parameter from simple presence/absence to detailed occupancy characteristics.
2Measurement precision
If audio signals are captured to determine the number of people, then the system can identify occupancy count, but it may also capture sounds from electronic devices causing false detections
Solution Approach 1:
The system extracts and isolates human voice signals from the mixed audio environment. By using signal processing techniques to separate human speech characteristics from electronic device sounds, the system can accurately determine occupancy count without being affected by television, radio, or other electronic device noises.
Solution Approach 2:
The patent introduces machine learning models as an intermediary between raw audio signals and occupancy determination. These models are trained to recognize patterns specific to human voices and can distinguish them from electronic device sounds, acting as a mediator that filters out harmful interference while preserving accurate occupancy information.
3Measurement precision
If machine learning models are used to filter electronic noise and identify voices, then the system can improve occupancy detection accuracy, but it increases system complexity
Solution Approach 1:
The system employs machine learning models that are trained on large datasets to automatically learn and adapt to different acoustic environments and voice patterns. Once trained, these models self-service by autonomously filtering electronic noise and identifying human voices without requiring manual intervention or complex real-time processing, thereby reducing operational complexity while maintaining high detection accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables precise control of HVAC systems based on the number of people present, improving energy efficiency and personal comfort by using machine learning models to filter out electronic noise and identify user devices.
Implementation Method 1
a process for determining the number of people that are present within a space based on voices that are heard within the space. Unlike existing HVAC systems that rely on proximity sensors and motion sensors, the occupancy tracking system captures and processes audio signals from within a space
Implementation Method 2
The disclosed system also includes a process for determining the number of people that are present within a space based on a signal strength of a network connection between a thermostat and an access point
Data Source
AI summary
An occupancy tracking device configured to receive a plurality of sound samples over a predetermined time period. The device is further configured to compute an audio signature for each sound sample. The audio signature includes a numerical value that uniquely identifies characteristics of an audio signal. The device is further configured to populate entries in the voice data log for the sound samples, to identify one or more clusters based on an audio signature that is associated with the populated entries, and to determine a number of clusters that are identified. The device is further configured to determine a predicted occupancy level based on the number of clusters that are identified and to control a Heating, Ventilation, and Air Conditioning (HVAC) system based on the predicted occupancy level.