Proactive building air quality management
An air quality management system comprises a plurality of air quality sensors to sense air quality within a building, a plurality of air cleaning devices, and a computer system in communication with the plurality of air quality sensors and the plurality of air cleaning devices. The plurality of air quality sensors is located at a particular location within the building. The computer system determines a correlational model of air quality for the building that indicates a correlational relationship between the sensed air quality, a spatial parameter, a temporal parameter, and operation of the air cleaning devices. The computer system controls the plurality of air cleaning devices to implement an air quality control policy based on one or more air quality management parameters.
1. An air quality control system for a building, the air quality control system comprising:
a plurality of air quality sensors to sense air quality and located at different locations within the building;
one or more occupancy sensors;
a plurality of air quality affecting devices comprising:
a first set of one of more air quality affecting devices that positively affect air quality; and
a second set of one or more air quality affecting devices that negatively affect air quality;
a plurality of device sensors, each of the plurality of device sensors associated with one or more of the plurality of air quality affecting devices; and
a computer system in communication with the plurality of air quality sensors and the plurality of air quality affecting devices, the computer system comprising:
a processor; and
a memory coupled to the processor, the memory storing instructions that, when executed by the processor, cause the computer system to, on an ongoing basis:
receive air quality measurements from the plurality of air quality sensors;
determine device states of the plurality of air quality affecting devices from sensor data from the plurality of device sensors;
determine spatial locations of the plurality of air quality sensors based on received signal strengths of wireless signals from the plurality of air quality sensors;
determine, based on the one or more occupancy sensors, locations of occupants in the building;
maintain a spatial model that indicates spatial relationships between the plurality of air quality sensors, the plurality of air-quality affecting devices, and occupants of the building;
predict a future air quality of the building based upon the received air quality measurements and the determined spatial relationships by, through machine learning, determining a correlation of the device states of the plurality of air quality affecting devices to future air quality; and
control the air quality affecting devices to affect air quality in the building according to the predicted future air quality.
2. The air quality control system of claim 1 , wherein the first set of one or more air quality affecting devices that positively affect air quality comprises an air cleaning device.
3. The air quality control system of claim 2 , wherein the first set of one or more plurality of air quality affecting devices that negatively affect air quality comprises an air contaminating device.
4. The air quality control system of claim 1 , wherein the plurality of device sensors comprise a device sensor that is configured to receive a first device state of an air quality affecting device from a device controller associated with the air quality affecting device, wherein the first device state comprises on/off state, power level, and/or operational parameters.
5. The air quality control system of claim 1 , wherein the plurality of device sensors comprise a device sensor that is integrated into an air quality affecting device.
6. The air quality control system of claim 1 , wherein the plurality of device sensors comprise a device sensor that is configured to sense a first device state of an air quality affecting device based on measurements of the air quality affecting device, wherein the first device state comprises on/off state, power level, and/or operational parameters.
7. The air quality control system of claim 1 , wherein the computer uses a recurrent neural network trained to predict the future air quality of the building based upon the received air quality measurements and the determined spatial relationships as inputs thereto.
8. The air quality control system of claim 1 , wherein:
the first set of one or more air quality affecting devices that negatively affect air quality comprises an air quality affecting device selected from the group consisting of a window, a fireplace, a kitchen device, an appliance, a vacuum, and a door; and
the first set of one or more air quality affecting devices that positively affect air quality comprises an air quality affecting device selected from the group consisting of an air purifier, an air filter, a fan, and an HVAC system.
9. A back-end computer system for controlling air quality of a building, the back-end computer system comprising:
a processor; and
a memory coupled to the processor, the memory storing instructions that, when executed by the processor, cause the computer system to, on an ongoing basis:
receive air quality measurements from a plurality of air quality sensors associated with the building;
receive occupancy information from one or more occupancy sensors associated with the building;
determine device states of a plurality of air quality affecting devices associated with the building from sensor data from a plurality of device sensors associated with a plurality of air quality affecting devices, wherein the plurality of air quality affecting devices comprises:
a first set of one of more air quality affecting devices that positively affect air quality; and
a second set of one or more air quality affecting devices that negatively affect air quality;
determine spatial locations of the plurality of air quality sensors based on received signal strengths of wireless signals from the plurality of air quality sensors;
determine, based on the one or more occupancy sensors, locations of occupants in the building;
maintain a spatial model that indicates spatial relationships between the plurality of air quality sensors, the plurality of air-quality affecting devices, and occupants of the building;
predict a future air quality of the building based upon the received air quality measurements and the determined spatial relationships by, through machine learning, determining a correlation of the device states of the plurality of air quality affecting devices to future air quality; and
control the air quality affecting devices to affect air quality in the building according to the predicted future air quality.
10. A computer-implemented method for controlling air quality of a building, the method comprising:
receiving, by a computer system, air quality measurements from a plurality of air quality sensors associated with the building;
receiving, by one or more occupancy sensor, occupancy information associated with the building;
determining, by the computer system, device states of a plurality of air quality affecting devices associated with the building from sensor data from a plurality of device sensors associated with a plurality of air quality affecting devices, wherein the plurality of air quality affecting devices comprises:
a first set of one of more air quality affecting devices that positively affect air quality; and
a second set of one or more air quality affecting devices that negatively affect air quality;
determine spatial locations of the plurality of air quality sensors based on received signal strengths of wireless signals from the plurality of air quality sensors;
determine, based on the one or more occupancy sensors, locations of occupants in the building;
maintaining, by the computer system, a spatial model that indicates spatial relationships between the plurality of air quality sensors, the plurality of air-quality affecting devices, and occupants of the building;
predicting, by the computer system, a future air quality of the building based upon the received air quality measurements and the determined spatial relationships by, through machine learning, determining a correlation of the device states of the plurality of air quality affecting devices to future air quality; and
controlling, by the computer system, the air quality affecting devices to affect air quality in the building according to the predicted future air quality.
11. The computer-implemented method of claim 10 , wherein the first set of one or more plurality of air quality affecting devices that positively affect air quality comprises an air cleaning device.
12. The computer-implemented method of claim 11 , wherein the first set of one or more plurality of air quality affecting devices that negatively affect air quality comprises an air contaminating device.
13. The computer-implemented method of claim 10 , wherein the plurality of device sensors comprise a device sensor that is configured to receive a first device state of an air quality affecting device from a device controller associated with the air quality affecting device, wherein the first device state comprises on/off state, power level, and/or operational parameters.
14. The computer-implemented method of claim 10 , wherein the plurality of device sensors comprise a device sensor that is integrated into an air quality affecting device.
15. The computer-implemented method of claim 10 , wherein the plurality of device sensors comprise a device sensor that is configured to sense a first device state of an air quality affecting device based on measurements of the air quality affecting device, wherein the first device state comprises on/off state, power level, and/or operational parameters.
16. The computer-implemented method of claim 10 , wherein predicting the future air quality through machine-learning system comprises using a recurrent neural network that is trained to predict the future air quality of the building based upon the received air quality measurements and the determined spatial relationships as inputs thereto.
17. The computer-implemented method of claim 10 , wherein:
the first set of one or more air quality affecting devices that negatively affect air quality comprises an air quality affecting device selected from the group consisting of a window, a fireplace, a kitchen device, an appliance, a vacuum, and a door; and
the first set of one or more air quality affecting devices that positively affect air quality comprises an air quality affecting device selected from the group consisting of an air purifier, an air filter, a fan, and an HVAC system.
18. The computer-implemented method of claim 10 , further comprising:
sensing, by the plurality of air quality sensors, the air quality measurements, wherein each of the plurality of air quality sensors is located at a different location within the building and each of the plurality of air quality sensors is in communication with the computer system;
sensing, by the plurality of device sensors, the states of the plurality of air quality affecting devices, wherein each of the plurality of devices sensors are in communication with the computer system; and
sensing, by the computer system, locations of the occupants of the building via one or more motion sensors.