Machine learning systems for modeling and balancing the activity of air quality devices in industrial applications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial facilities face challenges in effectively managing air quality due to large open areas and hazardous processes, which can lead to environmental and health issues, and existing air handling systems in residential and commercial settings are not optimized for industrial needs, lacking holistic control over diverse air quality devices.
Innovation Solution
A machine learning-based air quality control system that uses a wireless mesh network of sensors and devices to collect data on environmental conditions, analyze it using machine learning algorithms, and generate instructions to optimize air handling unit operations across multiple types of equipment for improved air quality and safety, integrating with existing hardware and networked devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional air handling systems with individual thermostats or on/off switches are used in industrial facilities, then each device can be controlled independently, but the system lacks holistic control over diverse air quality devices and cannot effectively manage air quality across large open areas
Solution Approach 1:
The patent combines multiple individual air handling devices into a unified networked system where devices communicate with each other and a central controller. This merging enables holistic control while preserving individual device controllability through the network architecture that coordinates actions across diverse equipment types.
Solution Approach 2:
The control system is designed to work with multiple types of air quality devices (HVAC units, MAUs, baghouses, exhaust units) through a universal communication protocol and interface. This multi-functionality allows the system to provide holistic control across diverse equipment while maintaining the ability to control each device type according to its specific requirements.
2Ease of manufacture
If residential or commercial HVAC systems are used in industrial facilities, then equipment is readily available and easier to install, but the systems are not optimized for industrial needs with large open areas and hazardous processes
Solution Approach 1:
The system segments the air quality management function into independent controllable units that can be individually selected and installed based on specific industrial needs. Each air handling device operates as a separate controllable entity within the network, allowing flexible configuration for large open areas and hazardous processes while maintaining ease of installation of individual components.
Solution Approach 2:
The system transforms static, fixed-function residential/commercial HVAC units into dynamic, adaptable air quality devices through network connectivity and intelligent control algorithms. The devices can dynamically adjust their operation based on real-time environmental conditions, process requirements, and coordination with other devices, thereby achieving industrial-grade reliability while using more accessible equipment.
3Productivity
If machine learning algorithms are used to optimize air handling unit operations, then energy efficiency and air quality management are improved, but system complexity and computational requirements increase
Solution Approach 1:
The patent introduces a network controller as an intermediary that handles the computational complexity of machine learning algorithms separately from the air handling devices. The controller collects data from sensors and devices, runs the optimization algorithms, and sends control commands back to the devices. This mediator approach enables advanced energy efficiency and air quality management while keeping individual device complexity low.
Data Source
AI summary
An indoor air quality control system may be implemented to control a plurality of air handling units within an industrial facility in a concerted effort to effect an overall air quality goal. A remote server analyzes sensor data, historical data, and other environmental data (e.g., predicted weather data), and uses one or more machine learning algorithms to model the behavior of air within the facility. The sensed air quality data is considered holistically to understand the overall condition of the facility and the gradient of air flows and/or contaminant flows within the 3-dimensional space. Air handling models are applied to current sensor data to generate instructions to selectively turn on/off or otherwise control components of various air handling equipment to reach an optimized air quality result. Decisions on how to control the facility are based on environmental health and safety considerations.