Distributed Machine Learning for Thermostat Energy Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current thermostat systems fail to optimally adjust heating or cooling based on user inputs and time-temperature data, leading to inefficiencies in energy usage.
Innovation Solution
A computer-implemented system using machine learning models, both on a cloud server and locally, to predict thermal characteristics of a room, adjusting thermostat settings for energy savings while maintaining comfort by analyzing historical and real-time data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If traditional thermostat control is used based on time and temperature, then the system is simple to operate, but energy efficiency is insufficient
Solution Approach 1:
The patent segments the machine learning model into two parts: a cloud-based model that processes historical data and generates predictions, and a local model that executes predictions and collects real-time data. This segmentation allows the system to achieve high energy efficiency through intelligent predictions while keeping local device complexity low, as the thermostat only needs to run lightweight inference rather than full model training.
Solution Approach 2:
The patent introduces a cloud server as an intermediary between the thermostat and the machine learning model. The cloud server handles complex model training and updates, while the thermostat receives pre-processed predictions and sends basic telemetry data. This intermediary approach enables energy-efficient local operation without requiring the thermostat to have high computational power.
2Use of energy by moving object
If machine learning models are implemented locally on the thermostat, then energy optimization improves, but device complexity increases
Solution Approach 1:
The patent segments the machine learning model into cloud-based training components and local execution components. The thermostat only runs lightweight inference using pre-trained models, which maintains ease of operation, while the cloud handles complex model updates that improve energy optimization over time.
Solution Approach 2:
The system implements self-service through automated model training and updates. The cloud server automatically retrains models using accumulated historical data and pushes updated model parameters to thermostats, eliminating the need for user intervention. This maintains ease of operation while continuously improving energy optimization capabilities.
3Measurement precision
If user inputs are continuously monitored and analyzed, then thermostat adjustment accuracy improves, but data processing requirements increase
Solution Approach 1:
The patent extracts the heavy data processing and model training functions from the thermostat and relocates them to the cloud server. The thermostat only performs lightweight data collection and model inference, while the cloud handles complex analytics on user inputs, historical patterns, and environmental data. This extraction enables high adjustment accuracy without increasing local processing power requirements.
Data Source
AI summary
Current thermostat data including room temperature readings, thermostat control commands, and user input temperature set point changes is collected. The current thermostat data is sent from a thermostat to a cloud server that creates and continually updates a first server machine learning model trained using previous and the current thermostat data, a second thermostat machine learning model is created for sending control commands to the thermostat using the previous and the current thermostat data, and the second thermostat machine learning model is updated. The second thermostat machine learning model is updated by sending updated model parameters from the first server machine learning model to the second thermostat machine learning model. One or more new thermostat commands determined from the current thermostat data are received from the second thermostat machine learning model and the one or more new thermostat commands are executed.