System and method for detecting room occupancy

Abstract

Embodiments of the present disclosure describe systems and methods for detecting room occupancy. The method includes receiving sensor readings indicative of corresponding occupancy parameters from a plurality of sensors located within the room. The plurality of sensors are configured to measure the corresponding occupancy parameters. The method also includes determining one or more of an occupancy status of the room and a confidence value associated with the determined occupancy status based on the received sensor readings. The occupancy status indicates one of a positive status indicating that the room is occupied and a negative status indicating that the room is not occupied. The method also includes, in response to determining the occupancy state, triggering a control action associated with the room based on the determined occupancy state.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 513,286, filed on July 12, 2023, which is incorporated herein by reference in its entirety. Technical Field

[0002] The present disclosure relates generally to detection systems, and more particularly to systems and methods for detecting room occupancy. Background Art

[0003] Typically, rooms at a facility that provides lodging are allocated to occupants who wish to stay at the facility. The occupants may not spend a majority of their time in the room, and the room may remain unoccupied for a considerable duration during the period in which the occupants stay at the facility. When the room is unoccupied, equipment within the room, such as lights and heating, cooling, and air conditioning (HVAC) units, may continue to operate, resulting in energy consumption.

[0004] In addition, conventional techniques for detecting the presence of an occupant in a room include single sensor-based systems and timer-based systems. For example, a timer-based system may classify a room as occupied for a duration longer than required. In addition, motion sensors may be used to detect the presence of an occupant in a room, however, motion sensors do not cover every area of ​​a room, resulting in blind spots, which may cause the room to be incorrectly classified as unoccupied. In addition, proximity sensors may be used to detect the presence of an occupant in a room, however, proximity sensors result in unbalanced detection.

[0005] As an example, motion sensor based detection results in a high false negative rate, i.e., a room is detected as unoccupied when it is occupied. As another example, proximity sensor based detection results in a high false positive rate, i.e., a room is detected as occupied when it is not occupied. The false positive rate is further associated with increased energy usage by appliances in the room, resulting in wasteful energy usage.

[0006] Furthermore, systems based on a single sensor are not resilient, as failure of the sensor results in failure of the entire detection system. It would therefore be advantageous to provide a solution that overcomes the above-mentioned problems. Summary of the invention

[0007] This summary is provided to introduce a selection of concepts in a simplified form that are further described in the detailed description of the disclosure. This summary is neither intended to identify key or essential inventive concepts of the disclosure nor to determine the scope of the disclosure.

[0008] Disclosed herein is a system for detecting room occupancy. The system includes a plurality of sensors located in the room and a control unit communicatively connected to the plurality of sensors. The plurality of sensors are configured to measure corresponding occupancy parameters. The control unit includes one or more processors, and the one or more processors are configured to receive sensor readings indicating the corresponding occupancy parameters from the plurality of sensors. Each of the received sensor readings includes a measurement sequence of the corresponding occupancy parameter within a predetermined duration. The one or more processors are also configured to determine, based on the received sensor readings, an occupancy state of the room and one or more of a confidence value associated with the determined occupancy state. The occupancy state indicates one of a positive state indicating that the room is occupied and a negative state indicating that the room is not occupied. The one or more processors are also configured to trigger a control action associated with the room based on the determined occupancy state in response to determining the occupancy state.

[0009] In one or more embodiments, each of the received sensor readings comprises a sequence of measurements of the corresponding occupancy parameter over a predetermined time duration.

[0010] In one or more embodiments, to determine one or more of the occupancy state and the confidence value, the one or more processors are further configured to select a neural network (NN) model from a plurality of NN models based on the types of the plurality of sensors and the corresponding occupancy parameters. The one or more processors are further configured to determine the one or more of the occupancy state and the confidence value based on the selected NN model.

[0011] In one or more embodiments, to determine one or more of the occupancy state and the confidence value, the one or more processors are further configured to: for each sensor reading in the received sensor readings, preprocess the measurement sequence to generate a preprocessed measurement sequence. To preprocess the measurement sequence, the one or more processors are configured to resample, filter, and normalize the measurement sequence. The one or more processors are further configured to determine the occupancy state and the confidence value based on the selected NN model and the preprocessed measurement sequence.

[0012] In one or more embodiments, the one or more processors are further configured to determine a first classification probability value and a second classification probability value associated with the occupancy state of the room based on the selected NN model and the preprocessed measurement sequence. The one or more processors are further configured to compare the first classification probability value and the second classification probability value. The one or more processors are further configured to, in response to determining that the first classification probability value is greater than the second classification probability value, determine the occupancy state as the positive state indicating that the room is occupied. The one or more processors are further configured to, in response to determining that the second classification probability value is greater than the first classification probability value, determine the occupancy state as the negative state indicating that the room is not occupied.

[0013] In one or more embodiments, in order to trigger the control action associated with the room, the one or more processors are further configured to control the operation of one or more devices associated with the room. The one or more devices include a heating, ventilation and air conditioning (HVAC) system, a thermostat, and a lighting unit.

[0014] In one or more embodiments, in order to trigger the control action associated with the room, the one or more processors are also configured to cause information related to the occupancy status and the confidence value to be displayed on one or more of a user interface arranged in the room and a user interface associated with the user's handheld device.

[0015] In one or more embodiments, in order to trigger the control action associated with the room, the one or more processors are also configured to send information related to the occupancy status and the confidence value to a room management unit, and the room management unit is configured to generate one or more of a maintenance alarm for the room and a tracking alarm for the room based on the determined occupancy status.

[0016] In one or more embodiments, the occupancy parameters include one or more of: motion, temperature, humidity, pressure, occupant heart rate readings, proximity, noise, particulate matter (PM), volatile organic compounds (VOC), carbon dioxide (CO2), light intensity, lock operation, and plug power consumption.

[0017] In one or more embodiments, the one or more processors are further configured to receive user input from a user indicating a request to determine the occupancy state of the room. The one or more processors are further configured to determine one or more of the occupancy state and the confidence value in response to receiving the user input.

[0018] Also disclosed herein is a method for detecting room occupancy. The method includes receiving sensor readings indicating corresponding occupancy parameters from a plurality of sensors located within the room. The plurality of sensors are configured to measure the corresponding occupancy parameters. Each of the received sensor readings includes a sequence of measurements of the corresponding occupancy parameters within a predetermined duration. In addition, the method includes determining an occupancy state of the room and one or more of a confidence value associated with the determined occupancy state based on the received sensor readings, wherein the occupancy state indicates one of a positive state indicating that the room is occupied and a negative state indicating that the room is not occupied. In addition, the method includes, in response to determining the occupancy state, triggering a control action associated with the room based on the determined occupancy state.

[0019] In one or more embodiments, each of the received sensor readings comprises a sequence of measurements of the corresponding occupancy parameter over a predetermined time duration.

[0020] In one or more embodiments, determining one or more of the occupancy state and the confidence value includes selecting a neural network (NN) model from a plurality of NN models based on the types of the plurality of sensors and the corresponding occupancy parameters. In addition, the method includes determining the one or more of the occupancy state and the confidence value based on the selected NN model.

[0021] In one or more embodiments, determining one or more of the occupancy state and the confidence value further comprises, for each of the received sensor readings, preprocessing the measurement sequence to generate a preprocessed measurement sequence. Preprocessing the measurement sequence comprises at least one of resampling, filtering, and normalizing the measurement sequence. In addition, the method comprises determining the occupancy state and the confidence value based on the selected NN model and the filtered measurement sequence.

[0022] In one or more embodiments, the method further comprises determining a first classification probability value and a second classification probability value associated with the occupancy state of the room based on the selected NN model and the filtered measurement sequence. In addition, the method comprises comparing the first classification probability value and the second classification probability value. In addition, the method comprises, in response to determining that the first classification probability value is greater than the second classification probability value, determining the occupancy state as the positive state indicating that the room is occupied. In addition, the method comprises, in response to determining that the second classification probability value is greater than the first classification probability value, determining the occupancy state as the negative state indicating that the room is not occupied.

[0023] In one or more embodiments, triggering the control action associated with the room includes controlling the operation of one or more devices associated with the room, wherein the one or more devices include a heating, ventilation, and air conditioning (HVAC) system, a thermostat, and a lighting unit.

[0024] In one or more embodiments, triggering the control action associated with the room includes causing information related to the occupancy status and the confidence value to be displayed on one or more of a user interface arranged in the room and a user interface associated with a user's handheld device.

[0025] In one or more embodiments, triggering the control action associated with the room includes sending information related to the occupancy status and the confidence value to a room management unit, and the room management unit is configured to generate one or more of a maintenance alarm for the room and a tracking alarm for the room based on the determined occupancy status.

[0026] In one or more embodiments, the occupancy parameters include one or more of: motion, temperature, humidity, pressure, occupant heart rate readings, proximity, noise, particulate matter (PM), volatile organic compounds (VOC), carbon dioxide (CO2), light intensity, lock operation, and plug power consumption.

[0027] In one or more embodiments, the method further comprises receiving user input from a user indicating a request to determine the occupancy state of the room. In addition, the method comprises determining one or more of the occupancy state and the confidence value in response to receiving the user input.

[0028] In order to further illustrate the advantages and features of the method, system and apparatus, a more specific description of the method, system and apparatus will be presented by reference to its specific embodiments shown in the accompanying drawings. It is understood that these drawings only depict typical embodiments of the present disclosure and are therefore not to be considered as limiting the scope thereof. The present disclosure will be described and explained with additional specificity and detail using the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] These and other features, aspects and advantages of the present disclosure will become better understood when the following detailed description is read with reference to the accompanying drawings, in which like reference numerals refer to like parts throughout the drawings, and in which:

[0030] Figure 1 A system environment for detecting room occupancy according to one or more embodiments of the present disclosure is shown;

[0031] Figure 2 A schematic block diagram of a system for detecting room occupancy according to one or more embodiments of the present disclosure is shown;

[0032] Figure 3A-3B A process flow diagram depicting a method for detecting room occupancy according to one or more embodiments of the present disclosure is shown; and

[0033] Figure 4 An additional process flow diagram depicting a method for training a neural network (NN) model to detect room occupancy is shown in accordance with one or more embodiments of the present disclosure.

[0034] Furthermore, those skilled in the art will appreciate that the elements in the drawings are shown for simplicity and may not necessarily be drawn to scale. For example, a flow chart illustrates the method in terms of the most significant steps involved to help improve understanding of various aspects of the present disclosure. Furthermore, in terms of the construction of the device, one or more components of the device may have been represented in the drawings by conventional symbols, and the drawings may show only those specific details relevant to understanding the embodiments of the present disclosure so as not to obscure the drawings with details that would be readily apparent to one of ordinary skill in the art having the benefit of the description herein. DETAILED DESCRIPTION

[0035] For the purpose of promoting an understanding of the principles of the present disclosure, reference will now be made to various embodiments, and specific language will be used to describe these embodiments. It will be understood, however, that no limitation of the scope of the present disclosure is intended thereby, and such changes and further modifications in the illustrated systems, and such further applications of the principles of the present disclosure as illustrated therein are deemed to be normally within the scope of those skilled in the art to which the present disclosure relates.

[0036] Those skilled in the art will understand that the foregoing general description and the following detailed description are illustrative of the present disclosure and are not intended to limit the present disclosure.

[0037] Throughout this specification, references to "on one hand," "on the other hand," or similar language mean that a particular feature, structure, or characteristic described in conjunction with the embodiment is included in at least one embodiment of the present disclosure. Thus, throughout this specification, appearances of the phrases "in an embodiment," "in another embodiment," "some embodiments," "one or more embodiments," and similar language may, but do not necessarily, all refer to the same embodiment.

[0038] The terms "comprises", "includes" or any other variations thereof are intended to cover a non-exclusive inclusion such that a process or method that includes a series of steps includes not only those steps but may also include other steps not expressly listed or inherent to such process or method. Similarly, without more limitations, one or more devices or subsystems or elements or structures or components followed by "comprises..." do not exclude the presence of other devices or other subsystems or other elements or other structures or other components or additional devices or additional subsystems or additional elements or additional structures or additional components.

[0039] In addition to overcoming challenges associated with detecting the presence of occupants in a room, the present disclosure provides a system that allows for automatic control of appliances in the room, such as HVAC systems and lighting units, which results in energy and cost savings. Furthermore, based on accurate detection of the presence of occupants in the room, staff efficiency is improved. Thus, a high confidence occupancy detection system is provided that results in energy and cost savings.

[0040] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.

[0041] Figure 1 A system environment 100 for detecting room occupancy is shown. The system environment 100 may include an indoor area / environment 110 including a plurality of rooms 120a-120n (hereinafter interchangeably referred to as "120"). The indoor area 110 may be associated with a facility that provides accommodation to users, such as, but not limited to, a hotel, a hostel, etc. The plurality of rooms 120a-120n may include rooms that have been assigned to users and rooms that are available for assignment to users seeking accommodation.

[0042] The plurality of rooms 120a-120n may be associated with a plurality of sensors 122a-122n, as each room 120 of the plurality of rooms 120a-120n may include one or more of the plurality of sensors 122a-122n located within the room 120. The plurality of sensors 122a-122n may be configured to measure corresponding occupancy parameters associated with the room 120. In some embodiments, each of the plurality of sensors 122a-122n may be configured to measure a corresponding occupancy parameter, the corresponding occupancy parameter indicating occupancy of the room 120. Specifically, based on the corresponding occupancy parameters sensed by the plurality of sensors, the occupancy of the room 120 may be determined, as will be described in further detail below. Those skilled in the art will appreciate that the details explained with respect to one room 120 may equally apply to each of the plurality of rooms 120a-120n without departing from the scope of the present disclosure.

[0043] In some embodiments, the corresponding occupancy parameters include multiple parameters such as, but not limited to, motion, temperature, humidity, pressure, occupant heartbeat readings, proximity, noise, particulate matter (PM) 2.5, PM 10, volatile organic compounds (VOC), carbon dioxide (CO2), light intensity, lock operation, and plug power consumption. The plurality of sensors 122a-122n may include sensors configured to sense one or more of the plurality of occupancy parameters and generate sensor readings indicative of the sensed occupancy parameters.

[0044] In some embodiments, the plurality of sensors 122a-122n may include a passive infrared (PIR) sensor for detecting motion of any occupant in the room 120. In some embodiments, the plurality of sensors 122a-122n may include a temperature sensor, a humidity sensor, and a pressure sensor for measuring temperature, humidity, and pressure, respectively, within the room 120. In some embodiments, the plurality of sensors 122a-122n may include a proximity sensor for sensing proximity of any occupant relative to the proximity sensor. In some embodiments, the plurality of sensors 122a-122n may include an indoor air quality (IAQ) sensor for measuring a plurality of parameters, such as, but not limited to, PM 2.5, PM 10, VOC, and CO2. In some embodiments, the plurality of sensors 122a-122n may include a noise sensor and a light sensor for measuring the amount of noise and the amount of light intensity, respectively, within the room 120. In some embodiments, the plurality of sensors 122a-122n may include a current transformer (CT) sensor for measuring power plug consumption. In some embodiments, the plurality of sensors 122a - 122n may include motion sensors and / or radio frequency identification (RFID) tags that are used to detect locking (and unlocking) operations of doors associated with the room 120 .

[0045] In some embodiments, one or more of the plurality of sensors 122a-122n may be sensors independently arranged at corresponding positions in the room 120. In some embodiments, one or more of the plurality of sensors 122a-122n may be integrated with a detection unit arranged in the room 120. In some embodiments, the temperature sensor, the humidity sensor, and the proximity sensor may be integrated with a thermostat arranged in the room 120. In some embodiments, the temperature sensor, the humidity sensor, the PIR sensor, the noise sensor, and the IAQ sensor may be integrated with a smoke detector arranged in the room 120. In some embodiments, the IAQ sensor may be integrated with an IAQ monitoring unit arranged in the room 120.

[0046] In some embodiments, each room 120 of the plurality of rooms 120a-120n may be associated with a heating, ventilation, and air conditioning (HVAC) system 124, a thermostat 126, and a lighting unit 128. The HVAC system 124 may be configured to facilitate circulation of fresh air in the room, and further facilitate heating and cooling of the room 120. The thermostat 126 may be configured to provide instructions to the HVAC system 124 regarding desired heating and cooling set points in the room 120. The lighting unit 128 may include various electrical components within the room 120, such as electrical outlets, lamps, table lamps, and other electrical devices.

[0047] The system environment also includes a control device 130 that is communicatively coupled to the plurality of sensors 122a-122n via a communication network 140, as discussed below throughout the present disclosure. In some embodiments, the HVAC system 124, the thermostat 126, and the lighting unit 128 may be communicatively coupled to the control device 130 via the communication network 140. The control device 130 may be configured to detect occupancy of the room 120 based on occupancy parameters sensed by the plurality of sensors 122a-122n associated with the room 120. The control device 130 and the plurality of sensors 122a-122n may form a system 150 for detecting occupancy of the room 120. In some embodiments, the functionality of the system 150 including the control device 130 and the plurality of sensors 120a-122n may be provided by means of a single device.

[0048] In some embodiments, the control device 130 may be a server-based device remote from the indoor area 110. In some embodiments, the control device 130 may be a cloud-based device. In some embodiments, the control device 130 may be disposed within the indoor area 110. For example, the control device 130 may be integrated with the thermostat 126, and the functionality of the control device 130 may be provided by the thermostat 126, and thus, in such embodiments, the thermostat 126 may form part of the system 150. In some embodiments, one or more components of the control device 130 may be disposed on the cloud, and one or more components of the control device 130 may be disposed on the server.

[0049] In some embodiments, the system environment 100 may include a room management unit 160 in communication with the system 150 (particularly with the control device 130 of the system 150). The room management unit 160 may be configured to manage the allocation status of the plurality of rooms 120a-120n and to manage the maintenance of the plurality of rooms 120a-120n. In some embodiments, the room management unit 160 may be associated with a front desk system that may be configured to manage the maintenance and allocation of the plurality of rooms 120a-120n. In some embodiments, the room management unit 160 may be implemented on hardware on-site (i.e., in the indoor area 110, such as a receiving space within the indoor area 110). In some embodiments, the room management unit 160 may be implemented remotely, such as via a cloud-based network and / or a server-based system. In some embodiments, the functionality associated with the room management unit 160 may be provided via a handheld device associated with one or more staff members.

[0050] In some embodiments, the room management unit 160 may be configured to track the allocation status of the plurality of rooms 120a-120n and generate tracking alerts for the plurality of rooms 120a-120n. In some embodiments, the room management unit 160 may be configured to communicate the allocation status of the plurality of rooms 120a-120n to the control device 130. In some embodiments, the room management unit 160 may be configured to generate maintenance alerts for the plurality of rooms 120a-120n. In some embodiments, the room management unit 160 may be configured to communicate the maintenance alerts and tracking alerts to one or more staff members to facilitate the management of the components (HVAC system 124, thermostat 126, and lighting unit 128) within the plurality of rooms 120a-120n and the maintenance of the plurality of rooms 120a-120n, as discussed in detail below throughout this disclosure.

[0051] Figure 2 A schematic block diagram of a system 150 for detecting occupancy of a plurality of rooms 120a - 120n in an indoor area 110 is shown.

[0052] like Figure 2 As shown in FIG, system 150 includes at least control device 130 and a plurality of sensors 122a-122n associated with room 120. It should be appreciated that although Figure 2 A single room 120 is depicted in the figure, but the control device 130 can communicate with multiple sensors 122a-122n associated with other rooms in the multiple rooms 120a-120n, and the details provided about the multiple sensors 122a-122n of the room 120 are also applicable to the multiple sensors 122a-122n of the other rooms.

[0053] In one or more embodiments, the control device 130 may include one or more processors 202 , a memory 204 , one or more modules 206 , and a communication interface 208 .

[0054] The one or more processors 202 may be configured to communicate with the memory 204 to store sensor readings generated by the plurality of sensors 122a-122n. In some embodiments, the memory 204 may include a sensor data unit 204a for storing sensor readings generated by the plurality of sensors 122a-122n. The sensor readings may include real-time sensor readings and historical sensor readings. In one or more embodiments, the one or more processors 202 may be one or more microprocessors or microcontrollers. The one or more processors 202 may include one or more processors, which may include one or more general-purpose processors (e.g., central processing unit (CPU), application processor (AP), etc.), pure graphics processing units (e.g., graphics processing unit (GPU), visual processing unit (VPU)), and / or artificial intelligence (AI) dedicated processors (e.g., neural processing unit (NPU)).

[0055] In some embodiments, the memory 204 may store data and instructions executable by the processor(s) 202 to perform method steps for detecting occupancy of the room 120, as discussed herein throughout the present disclosure. The memory 204 may also include, but is not limited to, non-transitory computer-readable storage media, such as various types of volatile and non-volatile storage media, including, but not limited to, random access memory, read-only memory, programmable read-only memory, electrically programmable read-only memory, electrically erasable read-only memory, flash memory, tape or disk, optical media, etc. In addition, the non-transitory computer-readable storage media of the memory 204 may include executable instructions in the form of modules 206 and a database for storing data. The module 206 may include a set of instructions that may be executed to cause the one or more processors 202 to perform any one or more of the methods for detecting occupancy of the plurality of rooms 120a-120n based on sensor readings, as discussed herein throughout the present disclosure. Specifically, the one or more modules 206 may be configured to use the data stored in the database of the memory 204 to perform the steps of the present disclosure for detecting occupancy of the plurality of rooms 120a-120n based on sensor readings. In another embodiment, module 206 may be one or more hardware units that may be located external to memory 204. In one embodiment, memory 204 may communicate via a bus within processor(s) 202.

[0056] In one or more embodiments, the communication interface 208 may include a transmitter and a receiver configured to communicate with the plurality of sensors 122a-122n via the communication network 140. In some embodiments, the communication interface 208 may be configured to communicate with one or more of the HVAC system 124, the thermostat 126, and the lighting unit 128 via the communication network 140. The communication via the communication network 140 may be based on a wireless communication protocol. The communication interface 208 may be configured to communicate internally between internal hardware components and to communicate with external devices (e.g., the plurality of sensors 122a-122n) via one or more networks (e.g., radio technologies). The communication interface 208 may include electronic circuitry specific to a standard that may enable wireless communication.

[0057] refer to Figure 1 and Figure 2 , the one or more processors 202 may be configured to receive sensor readings from a plurality of sensors 122a-122n associated with the room 120. The sensor readings may indicate corresponding occupancy parameters associated with the room 120. In some embodiments, the one or more processors 202 may be configured to receive sensor readings from at least two of the plurality of sensors 122a-122n. In some embodiments, the one or more processors 202 may be configured to receive sensor readings from a sensor combination of the plurality of sensors 122a-122n. For example, the one or more processors 202 may be configured to receive sensor readings from a combination of an IAQ sensor and a PIR sensor, a combination of an IAQ sensor and a noise sensor, a combination of an IAQ sensor, a PIR sensor, and a noise sensor, a combination of an IAQ sensor, a PIR sensor, and a proximity sensor, a combination of an IAQ sensor, a noise sensor, and a proximity sensor, and the like.

[0058] In some embodiments, the sensor readings from each of the plurality of sensors 122a-122n may include a sequence of measurements of the corresponding occupancy parameter, which is sensed over a predetermined duration. The sequence of measurements from each of the plurality of sensors 122a-122n may include real-time measurements of the corresponding occupancy parameter and historical measurements of the corresponding occupancy parameter. In some embodiments, the predetermined duration may include any amount of duration, such as, but not limited to, 30 minutes, 45 minutes, 1 hour, 2 hours, etc. In some embodiments, the sequence of measurements may be in the form of time series data indicating measurements of the corresponding occupancy parameter over the predetermined duration.

[0059] The one or more processors 202 may be configured to receive sensor readings from the plurality of sensors 122a-122n and determine an occupancy state of the room 120 and / or a confidence value associated with the occupancy state of the room 120 based on the received sensor readings. The occupancy state of the room 120 indicates whether the room 120 is occupied or unoccupied. The occupancy state of the room 120 may indicate one of a positive state or a negative state. A positive state may indicate that the room 120 is occupied, while a negative state may indicate that the room 120 is unoccupied.

[0060] In some embodiments, one or more processors 202 may be configured to use multiple neural network (NN) models to detect occupancy states and / or confidence values. The multiple NN models may include trained machine learning (ML) models that process sensor readings to detect occupancy states and / or confidence values. In some embodiments, the multiple NN models may include supervised classification networks, such as, but not limited to, pattern recognition networks (PRNs), long short-term memory (LSTMs), and the like.

[0061] In some embodiments, a plurality of NN models may be stored in the NN unit 204b of the memory 204. In some embodiments, each of the plurality of NN models may be associated with a corresponding combination of sensors of the plurality of sensors 122a-122n. Specifically, each NN model may be associated with the types of the plurality of sensors 122a-122n and the corresponding occupancy parameters, because each NN model may be trained to detect occupancy states and confidence values ​​for the relevant types and combinations of the plurality of sensors 122a-122n and the corresponding occupancy parameters. As an example, a first NN model may be associated with a combination of an IAQ sensor and a PIR sensor, the first NN model being configured to detect occupancy states and confidence values ​​based on sensor readings from the IAQ sensor and the PIR sensor. As another example, a second NN model may be associated with a combination of an IAQ sensor, a PIR sensor, a noise sensor, and a proximity sensor, the second NN model being configured to detect occupancy states and confidence values ​​based on sensor readings from the IAQ sensor, the PIR sensor, the noise sensor, and the proximity sensor. Thus, each of the plurality of NN models may be associated with a particular type and combination of the plurality of sensors 122a-122n and corresponding occupancy parameters.

[0062] In some embodiments, the one or more processors 202 may be configured to select an NN model from a plurality of NN models based on the type or combination of the plurality of sensors 122a-122n that send sensor readings of corresponding occupancy parameters. The one or more processors 202 may be configured to determine the occupancy state and confidence value of the room 120 based on the selected NN model. In some embodiments, the selected NN model may be the best NN model for detecting the occupancy state of the room 120, which is predetermined during the training of the plurality of NN models. Therefore, the relevant NN model (e.g., the best NN model) may be used to accurately determine the occupancy state of the room based on the plurality of sensors that are being used in the room and provide sensor readings of corresponding occupancy parameters associated with the room. In addition, in the event that a particular sensor in the room fails, a different NN model may be used to accurately determine the occupancy state based on other working sensors in the room. In addition, in the event that the measurement sequence received from the sensor combination associated with the best NN model includes missing data, then a different NN model may be selected to determine the occupancy state. As an example, the best NN model may be the first NN model associated with a combination of CO2, noise, and motion sensors. However, for a predetermined duration, the data sequence from the CO2 sensor may be lost. As a result, a second NN model associated only with the noise and motion sensors may be selected to detect the occupancy status of the room.

[0063] In some embodiments, the one or more processors 202 may be configured to process each of the sensor readings received from the active (working) sensors among the plurality of sensors within the room 120. The one or more processors 202 may be configured to receive a sequence of measurements sensed within a predetermined duration from the active sensors at a time instance at which occupancy of the room 120 is to be detected. The sequence of measurements includes a real-time sensor reading of a corresponding occupancy parameter at the time instance and a historical sensor reading of the corresponding occupancy parameter sensed within the predetermined duration.

[0064] In some embodiments, one or more processors 202 may be configured to pre-process the received measurement sequence before providing the measurement sequence as input to the selected NN model. In some embodiments, the pre-processing of the received measurement sequence includes at least one of normalization, resampling and filtering of the measurement sequence. In some embodiments, one or more processors 202 may be configured to resample each measurement sequence in the received measurement sequence based on a resolution parameter, i.e., based on the desired resolution, so as to generate a resampled measurement sequence. In some embodiments, the resampled measurement sequence may include a uniformly spaced data sequence associated with the received measurement sequence. In some embodiments, sensor readings from IAQ sensors and noise sensors may be resampled based on interpolation techniques. In some embodiments, sensor readings from PIR sensors may be resampled using a zero-order hold technique. In some embodiments, the resolution parameter may be any desired resolution, such as but not limited to 5 seconds, 30 seconds, 1 minute, etc.

[0065] In some embodiments, one or more processors 202 may be configured to filter the resampled measurement sequence to generate a filtered measurement sequence. The filtered measurement sequence may correspond to a preprocessed measurement sequence determined by resampling and filtering the measurement sequence. In some embodiments, one or more processors 202 may be configured to remove noise from the resampled measurement sequence and / or remove outliers from the resampled measurement sequence. In some embodiments, one or more processors 202 may be configured to smooth the resampled measurement sequence based on one or more filtering techniques (e.g., but not limited to Gaussian filtering, Savitzky-Golay filtering, etc.). In some embodiments, the measurement sequence may be received at a desired resolution, and resampling may not be required. In such an embodiment, one or more processors 202 may be configured to filter the measurement sequence to generate a filtered measurement sequence, which may correspond to the preprocessed measurement sequence. In some embodiments, one or more processors 202 may be configured to normalize each measurement sequence in the received measurement sequence to generate a normalized measurement sequence, which may correspond to the preprocessed measurement sequence.

[0066] In some embodiments, one or more processors 202 may be configured to pass the filtered data sequence to the selected NN model, and determine the occupancy state and confidence value based on the selected NN model and the filtered measurement sequence. One or more processors 202 may be configured to determine a first classification probability value and a second classification probability value associated with the occupancy state of the room 120 based on the selected NN model and the filtered measurement sequence. In some embodiments, one or more processors 202 may be configured to determine a characteristic sequence of the measurement sequence, such as a sequence of standard deviations, derivatives, moving averages, etc. The determined characteristic sequence may be provided to the selected NN model, and a first classification probability value and a second classification probability value associated with the occupancy state of the room 120 may be determined based on the selected NN model and the determined characteristic sequence. One or more processors 202 may be configured to compare the first classification probability value and the second classification probability value.

[0067] Based on the comparison of the first classification probability value and the second classification probability value, the one or more processors 202 may be configured to determine an occupancy state and a confidence value for the determined occupancy state. In some embodiments, the one or more processors 202 may be configured to, in response to determining that the first classification probability value is greater than the second classification probability value, determine the occupancy state as a positive state indicating that the room 120 is occupied. In some embodiments, the first classification probability value may be a confidence value associated with the determined occupancy state (positive state). In some embodiments, the one or more processors 202 may be configured to, in response to determining that the second classification probability value is greater than the first classification probability value, determine the occupancy state as a negative state indicating that the room 120 is not occupied. In some embodiments, the second classification probability value may be a confidence value associated with the determined occupancy state (negative state).

[0068] In some embodiments, the one or more processors 202 may be configured to store the occupancy state and the confidence value of the determined occupancy state in the memory 204 (e.g., in the sensor data unit 204a of the memory 204). In some embodiments, the one or more processors 202 may be configured to determine the occupancy state and the confidence value of the determined occupancy state at a predetermined frequency. The predetermined frequency may be any desired frequency, such as but not limited to 30 seconds, 1 minute, 2 minutes, etc. In some embodiments, the one or more processors 202 may be configured to receive a user input indicating a request to determine the occupancy state, and in response to receiving the user input, the one or more processors 202 may be configured to determine the occupancy state and the confidence value of the determined occupancy state. In some embodiments, the user input may be provided by a staff member via a user device that communicates with the control device 130. In some embodiments, the user input may be provided via a web-based application or a mobile application on the user device.

[0069] In some embodiments, the one or more processors 202 may be configured to receive a room allocation status of the room 120 from the room management unit 160, and determine an occupancy status of the room 120 based on the received room allocation status. The room allocation status may include, for example, information about whether the room 120 is allocated to an occupant, whether the room 120 is available for allocation, and a check-in / check-out time associated with the room 120. The one or more processors 202 may determine the occupancy status taking into account the room allocation status. As an example, the one or more processors 202 may determine the occupancy status of the room 120 in conjunction with the plurality of sensors 122a-122n only when the room allocation status indicates that the room 120 is allocated to an occupant. As another example, the one or more processors 202 may determine the occupancy status of the room 120 in conjunction with the plurality of sensors 122a-122n only when the room allocation status indicates that an occupant may be in the room 120 based on the check-in and check-out times associated with the room 120.

[0070] In some embodiments, the one or more processors 202 may be configured to receive sensor readings from different sensor combinations of the plurality of sensors 122a-122n based on one or more external parameters, such as a particular time of day. As an example, the one or more processors 202 may be configured to receive sensor readings from a first sensor combination during morning hours and to receive sensor readings from a second sensor combination during evening hours.

[0071] In some embodiments, the one or more processors 202 may be configured to train multiple NN models based on the type or combination of multiple sensors 122a-122n that send sensor readings corresponding to occupancy parameters. Each NN model in the multiple NN models may be associated with a combination of multiple sensors 122a-122n, so that in the event of failure of one or more sensors in the room 120, a relevant NN model may be selected based on active sensors within the room 120 to determine the occupancy state of the room 120. Based on the training, the best NN network may be initially selected from the multiple NN models to detect the occupancy state of the room 120. In some cases, a different NN model may be selected from the multiple NN models as the best NN network, such as when there is a sensor failure and / or loss of information associated with one or more sensors. As an example, initially, a NN model based on a combination of a noise sensor, a PIR sensor, and an IAQ sensor may be used as the best NN model to determine the occupancy state of the room 120. In the event of a failure of the noise sensor, or in the event that the noise sensor may provide incomplete readings, the PIR sensor and the IAQ sensor may still provide corresponding sensor readings. Then, another NN model trained to consider sensor readings from a combination of the PIR sensor and the IAQ sensor can be selected as the best NN model to determine the occupancy state. Thus, a reliable system for determining the occupancy state of the room 120 can be achieved despite the failure of one or more sensors.

[0072] In some embodiments, one or more processors 202 may be configured to train multiple NN models for corresponding sensor types and combinations. In some embodiments, multiple NN models may be trained based on supervised learning using historical sensor readings from a lab-based setting. In some embodiments, multiple NN models may be trained based on supervised learning using historical sensor readings and actual occupancy data associated with room 120 and / or other rooms. Thus, multiple NN models may be trained based on at least one of sensor readings from a lab-based setting or sensor readings from one or more rooms. In some embodiments, multiple NN models may initially be trained based on sensor readings from a lab-based setting, and once deployed in a room (e.g., room 120), may be further trained based on sensor readings from the room and / or other different rooms.

[0073] For each NN model, the one or more processors 202 may be configured to obtain historical sensor readings for the corresponding sensor type and combination. In some embodiments, the historical sensor readings may be retrieved from the sensor data unit 204a of the memory 204. The historical sensor readings may be for a specified time period. In some embodiments, the historical sensor readings may include a measurement sequence for a specified time period.

[0074] In some embodiments, one or more processors 202 may be configured to obtain actual occupancy data for a specified time period, the actual occupancy data indicating the actual occupancy state of the room 120 at each time instance in the specified time period. The actual occupancy data may be used to construct the training output of the NN model to be trained. In some embodiments, the NN model to be trained may have two output labels—a label [1, 0] indicating that the room is occupied, and a label [0, 1] indicating that the room is not occupied.

[0075] In some embodiments, one or more processors 202 may be configured to provide historical sensor readings as training inputs to the NN model to be trained. In some embodiments, one or more processors 202 may be configured to resample, smooth, and normalize the historical sensor readings before providing them as training inputs to the NN model in training. In some embodiments, one or more processors 202 may be configured to convert the historical sensor readings into a sequence set based on a windowing technique. As an example, a window size in minutes and a step size in minutes may be determined. In some embodiments, when the NN model in training is deployed for real-time use, the window size may have the same value as the predetermined duration of the measurement sequence to be received. In some embodiments, when the NN model in training is deployed for real-time use, the resolution of the historical sensor readings to be provided as input to the NN model in training may have the same value as the resolution parameter associated with the measurement sequence to be received.

[0076] In some embodiments, one or more processors 202 can be configured to train NN models based on training inputs (historical sensor readings) and training outputs (actual occupancy data). Similarly, each of the multiple NN models can be trained based on historical sensor readings from sensors of corresponding types and combinations. Thus, a set of fail-safe NN models can be trained and stored in NN unit 204b of memory 204, so that in real-time use, readings from any combination of sensors active within room 120 can be used to determine the occupancy of room 120.

[0077] In some embodiments, based on the training, the best NN model can be determined by one or more processors 202 for the room 120, and the best NN model can be selected by one or more processors 202 to detect the occupancy state of the room 120. The best NN model can refer to the NN model that provides the most accurate detection of the occupancy state of the room 120 based on the training. In the event that one or more sensors in the sensor combination associated with the best NN model fail, the one or more processors 202 can select another NN model to detect the occupancy state of the room 120. In some embodiments, the training of multiple NN models can include receiving feedback from one or more staff members associated with the room 120, the feedback being related to the occupancy state of the room 120. Based on the feedback, the best NN model can be updated from the multiple NN models. For example, the first NN model can initially be selected as the best NN model for detecting the occupancy state, however, based on the feedback, it can be determined that the second NN model provides better accuracy, and therefore the second NN model can be determined as the best NN model for detecting the occupancy state of the room 120.

[0078] In some embodiments, the one or more processors 202 may be configured to, in response to determining an occupancy state, trigger a control action associated with the room 120 based on the determined occupancy state. In some embodiments, the one or more processors 202 may be configured to control the operation of the HVAC system 124, the thermostat 126, and the lighting unit 128 associated with the room 120 based on the determined occupancy state of the room 120. In some embodiments, when it is determined that the occupancy state of the room is a negative state (i.e., the room 120 is not occupied), the one or more processors 202 may be configured to send a control signal to the thermostat 126 and / or the HVAC system 124 to adjust the temperature set point of the room 120, e.g., to back off the temperature set point of the room 120. Thus, efficient temperature backoff control is achieved because the temperature can be backed off for the room 120 when the room 120 is not occupied, thereby resulting in energy savings for the room 120. In some embodiments, the one or more processors 202 may be configured to send control signals to the lighting units 128 (either directly to the lighting units 128 or via a lighting control unit within the room 120) to adjust the operation of the lighting units 128. For example, when the room 120 is not occupied, the lighting units 128 may be turned off, thereby resulting in further energy savings for the room 120.

[0079] In some embodiments, the one or more processors 202 may be configured to cause information related to the determined occupancy status and confidence value to be displayed on a user interface of one or more user devices. In some embodiments, the information related to the determined occupancy status and confidence value may be displayed on a user interface located within the room 120 (e.g., a user interface associated with the thermostat 126). In some embodiments, the information related to the determined occupancy status and confidence value may be displayed on a user interface associated with a user's handheld device (e.g., a handheld device of one or more staff members). Thus, the staff can monitor the occupancy status of the room 120, thereby facilitating the staff to plan housekeeping and maintenance tasks for the room 120.

[0080] In some embodiments, the one or more processors 202 may be configured to send information related to the determined occupancy status and confidence value to the room management unit 160. The room management unit 160 may be configured to generate maintenance alerts and / or tracking alerts for the room 120 based on the occupancy status. In some embodiments, the maintenance alert may indicate maintenance requirements for the room 120, and the tracking alert may indicate a change in the occupancy status of the room 120. In some embodiments, the information related to the determined occupancy status and confidence value may be displayed on a user interface associated with the room management unit 160. The room management unit 160 may also be configured to generate maintenance and tracking alerts for one or more staff members, thereby allowing the staff members to take advantage of time when the room 120 is not occupied to perform housekeeping and maintenance tasks.

[0081] Figure 3A A process flow describing a method 300 for detecting room occupancy is shown. The method 300 may be described with reference to the room 120, and those skilled in the art will appreciate that the method may be performed for each of the plurality of rooms 120a-120n in the indoor area 110.

[0082] At step 302, the method 300 includes receiving sensor readings indicating corresponding occupancy parameters from a plurality of sensors 122a-122n located within the room 120. The plurality of sensors 122a-122n may be configured to measure the corresponding occupancy parameters. In some embodiments, the sensor readings may be received from at least two of the plurality of sensors 122a-122n. In some embodiments, the sensor readings may be received from a combination of sensors in the plurality of sensors 122a-122n. In some embodiments, the received sensor readings include a sequence of measurements of the corresponding occupancy parameters sensed over a predetermined duration.

[0083] At step 304, method 300 includes determining an occupancy state of room 120 based on the received sensor readings and one or more of a confidence value associated with the determined occupancy state. The occupancy state may indicate one of a positive state indicating that the room is occupied and a negative state indicating that the room is not occupied.

[0084] In some embodiments, the occupancy state and the confidence value may be determined based on a NN model selected from a plurality of NN models. The NN model may be selected based on a type or combination of the plurality of sensors 122a-122n associated with the received sensor readings. In some embodiments, the plurality of NN models may be trained based on historical sensor readings associated with the plurality of sensors 122a-122n. In some embodiments, each of the plurality of NN models may be associated with a corresponding type or combination of the plurality of sensors 122a-122n, and the plurality of NN models may be trained based on historical sensor readings from the plurality of sensors 122a-122n of the corresponding type.

[0085] At step 306, the method 300 includes, in response to determining the occupancy status, triggering a control action associated with the room 120 based on the determined occupancy status. In some embodiments, triggering the control action may include controlling the operation of one or more of the thermostat 126, the HVAC system 124, and the lighting unit 128 associated with the room 120. In some embodiments, triggering the control action may include causing information related to the occupancy status and the confidence value to be displayed on a user interface disposed within the room and / or a user interface associated with a handheld device of a user. In some embodiments, triggering the control action may include sending the information related to the occupancy status and the confidence value to the room management unit 160, which may generate a maintenance and / or tracking alert for the room 120.

[0086] In some embodiments, to determine the occupancy state and the confidence value, the method 300 may include sub-steps 304a-304g for determining the occupancy state and the confidence value, such as Figure 3B At step 304a, the method 300 includes selecting a NN model from a plurality of NN models based on the types of the plurality of sensors 122a-122n and the corresponding occupancy parameters.

[0087] At step 304b, method 300 includes, for each received sensor reading, resampling a measurement sequence associated with the received sensor reading based on the resolution parameter to generate a resampled measurement sequence.

[0088] At step 304c, the method 300 includes filtering the resampled measurement sequence to generate a filtered measurement sequence.

[0089] At step 304d, the method 300 includes determining a first classification probability value and a second classification probability value associated with the occupancy state of the room 120 based on the selected NN model and the filtered measurement sequence.

[0090] At step 304e, method 300 includes comparing the first classification probability value and the second classification probability value.

[0091] At step 304f, the method 300 includes, in response to determining that the first classification probability value is greater than the second classification probability value, determining the occupancy state as a positive state indicating that the room 120 is occupied.

[0092] At step 304g, the method 300 includes, in response to determining that the second classification probability value is greater than the first classification probability value, determining the occupancy state as a negative state indicating that the room 120 is unoccupied.

[0093] Figure 4 A process flow describing a method 400 for training a NN model among a plurality of NN models to determine an occupancy state of a room 120 and a confidence value for the occupancy state is shown.

[0094] At step 402, the method 400 includes retrieving historical sensor readings of the plurality of sensors 122a-122n from the memory 204. In some embodiments, the historical sensor readings may include time series measurements of the plurality of sensors 122a-122n. In some embodiments, the historical sensor readings may be for a particular time period. In some embodiments, the NN model may be associated with a type or combination of the plurality of sensors 122a-122n, and the NN model may be trained based on the historical sensor readings of the plurality of sensors 122a-122n of the associated type.

[0095] At step 404, method 400 includes generating training inputs for the NN model under training based on processing the historical sensor readings. In some embodiments, the historical sensor readings may include time series measurements or measurement sequences, and the time series measurements may be resampled, smoothed, and normalized. In some embodiments, windowing techniques may be used to convert the time series measurements into a sequence set, which forms the training input for the NN model under training.

[0096] At step 406, the method 400 includes retrieving actual occupancy data associated with the room 120. In some embodiments, the actual occupancy data may be stored in the memory 204. In some embodiments, the actual occupancy data may be related to a particular time period for retrieving historical sensor readings of the plurality of sensors 122a-122n.

[0097] At step 408 , method 400 includes training the NN model based on the set of sequences as training input and the actual occupancy data as training output.

[0098] In some embodiments, the method 400 may be performed on each of a plurality of NN models to train the plurality of NN models based on historical sensor readings of a plurality of sensors 122a - 122n of related types and combinations.

[0099] Although shown and described in a particular order Figure 3A , Figure 3B and Figure 4 However, according to various embodiments of the present disclosure, these steps may appear in a variation of this order. Figure 3A , Figure 3B and Figure 4 A detailed description of the various steps involved is included in the Figure 1-Figure 2 The relevant description is omitted in this article for the sake of brevity.

[0100] In some embodiments, the one or more processors 202 may be configured to execute instructions included in a computer program product. The computer program product may be embodied on a non-transitory computer readable medium. The computer program product may include instructions that, when executed by the one or more processors 202, cause the one or more processors 202 to perform a reference Figure 3A , Figure 3B and Figure 4 Detailed description of the method steps.

[0101] The present disclosure provides methods and systems for accurate and high confidence occupancy detection in a room. Accurate occupancy detection allows automatic control of appliances in the room, resulting in energy and cost savings. For example, the temperature can be efficiently controlled in the room during unoccupied hours, resulting in energy savings. Accurate occupancy detection is critical for efficient temperature control.

[0102] Furthermore, occupancy detection with high confidence allows for improved worker efficiency, as well as providing insight into room occupancy trends. For example, workers can rely on accurate room occupancy status to efficiently plan housekeeping tasks. Furthermore, workers do not need to manually check if a room is occupied, and can utilize the time when the room is unoccupied for maintenance and housekeeping.

[0103] Thus, a high confidence occupancy detection system is provided at a low cost. The detection system does not affect guest privacy. Furthermore, the detection system can be easily installed and does not require regular maintenance.

[0104] As will be apparent to those skilled in the art, various working modifications may be made to the methods disclosed herein in order to implement the inventive concepts as taught herein.

[0105] In addition, the actions of any flowchart do not need to be implemented in the order shown; nor do all actions need to be performed. Likewise, those actions that are not dependent on other actions can be performed in parallel with other actions.

[0106] The accompanying drawings and the foregoing description give examples of embodiments. It will be appreciated by those skilled in the art that one or more of the described elements may be well combined into a single functional element. Alternatively, some elements may be divided into multiple functional elements. Elements from one embodiment may be added to another embodiment. For example, the order of the processes described herein may be changed and are not limited to the methods described herein.

[0107] Benefits, other advantages, and solutions to problems have been described above with respect to specific embodiments. However, the benefits, advantages, solutions to problems, and any (one or more) components that may cause any benefit, advantage, or solution to appear or become more significant should not be construed as key, required, or essential features or components of any or all claims.

[0108] Although specific language has been used to describe the subject matter, any resulting limitations are not intended. As will be apparent to those skilled in the art, various work modifications may be made to the method to implement the inventive concept as taught herein. The accompanying drawings and the foregoing description provide examples of embodiments. Those skilled in the art will appreciate that one or more of the described elements may be well combined into a single functional element. Alternatively, some elements may be divided into a plurality of functional elements. Elements from one embodiment may be added to another embodiment.

Claims

1. A system for detecting room occupancy, the system comprising: a plurality of sensors located within the room, the plurality of sensors configured to measure corresponding occupancy parameters; as well as A control unit communicatively connected to the plurality of sensors, the control unit comprising one or more processors, the one or more processors being configured to: receiving sensor readings indicative of the corresponding occupancy parameter from the plurality of sensors, wherein each of the received sensor readings comprises a sequence of measurements of the corresponding occupancy parameter over a predetermined time duration; determining one or more of an occupancy state of the room and a confidence value associated with the determined occupancy state based on the received sensor readings, wherein the occupancy state indicates one of a positive state indicating that the room is occupied and a negative state indicating that the room is unoccupied; and In response to determining the occupancy status, a control action associated with the room is triggered based on the determined occupancy status.

2. The system according to claim 1, wherein: To determine one or more of the occupancy state and the confidence value, the one or more processors are configured to: selecting a neural network (NN) model from a plurality of NN models based on the types of the plurality of sensors and the corresponding occupancy parameters; as well as The one or more of the occupancy state and the confidence value are determined based on the selected NN model.

3. The system according to claim 2, wherein: To determine one or more of the occupancy state and the confidence value, the one or more processors are further configured to: For each of the received sensor readings: preprocessing the measurement sequence to generate a preprocessed measurement sequence, wherein to preprocess the measurement sequence, the one or more processors are configured to at least one of resample, filter, and normalize the measurement sequence; and Based on the selected NN model and the preprocessed measurement sequence, the occupancy state and the confidence value are determined.

4. The system according to claim 3, wherein: The one or more processors are further configured to: determining, based on the selected NN model and the preprocessed measurement sequence, a first classification probability value and a second classification probability value associated with the occupancy state of the room; comparing the first classification probability value and the second classification probability value; In response to determining that the first classification probability value is greater than the second classification probability value, determining the occupancy state as the affirmative state indicating that the room is occupied; as well as In response to determining that the second classification probability value is greater than the first classification probability value, the occupancy state is determined to be the negative state indicating that the room is unoccupied.

5. The system according to claim 1, wherein: In order to trigger the control action associated with the room, the one or more processors are configured to control the operation of one or more devices associated with the room, wherein the one or more devices include a heating, ventilation and air conditioning (HVAC) system, a thermostat and a lighting unit.

6. The system according to claim 1, wherein: In order to trigger the control action associated with the room, the one or more processors are configured to cause information related to the occupancy status and the confidence value to be displayed on one or more of a user interface arranged in the room and a user interface associated with a user's handheld device.

7. The system according to claim 1, wherein: In order to trigger the control action associated with the room, the one or more processors are configured to send information related to the occupancy status and the confidence value to a room management unit, and the room management unit is configured to generate one or more of a maintenance alarm for the room and a tracking alarm for the room based on the determined occupancy status.

8. The system according to claim 1, wherein: The occupancy parameters include one or more of: motion, occupant heartbeat reading, temperature, humidity, pressure, proximity, noise, particulate matter (PM), volatile organic compounds (VOC), carbon dioxide (CO2), light intensity, lock operation, and plug power consumption.

9. The system according to claim 1, wherein: The one or more processors are configured to: receiving user input from a user indicating a request to determine the occupancy status of the room; and In response to receiving the user input, one or more of the occupancy state and the confidence value are determined.

10. A method for detecting room occupancy, the method comprising: receiving sensor readings indicative of a corresponding occupancy parameter from a plurality of sensors located within the room, wherein the plurality of sensors are configured to measure the corresponding occupancy parameter, and wherein each of the received sensor readings comprises a sequence of measurements of the corresponding occupancy parameter over a predetermined duration; and determining one or more of an occupancy state of the room and a confidence value associated with the determined occupancy state based on the received sensor readings, wherein the occupancy state indicates one of a positive state indicating that the room is occupied and a negative state indicating that the room is unoccupied; and In response to determining the occupancy status, a control action associated with the room is triggered based on the determined occupancy status.

11. The method according to claim 10, wherein: Determining one or more of the occupancy state and the confidence value comprises: selecting a neural network (NN) model from a plurality of NN models based on the types of the plurality of sensors and the corresponding occupancy parameters; and The one or more of the occupancy state and the confidence value are determined based on the selected NN model.

12. The method according to claim 11, wherein: Determining one or more of the occupancy state and the confidence value further comprises: For each of the received sensor readings: preprocessing the measurement sequence to generate a preprocessed measurement sequence, wherein preprocessing the measurement sequence comprises at least one of resampling, filtering, and normalizing the measurement sequence; and Based on the selected NN model and the preprocessed measurement sequence, the occupancy state and the confidence value are determined.

13. The method according to claim 12, further comprising: determining, based on the selected NN model and the preprocessed measurement sequence, a first classification probability value and a second classification probability value associated with the occupancy state of the room; comparing the first classification probability value and the second classification probability value; In response to determining that the first classification probability value is greater than the second classification probability value, determining the occupancy state as the affirmative state indicating that the room is occupied; as well as In response to determining that the second classification probability value is greater than the first classification probability value, the occupancy state is determined to be the negative state indicating that the room is unoccupied.

14. The method according to claim 10, wherein: Triggering the control action associated with the room includes controlling operation of one or more devices associated with the room, wherein the one or more devices include a heating, ventilation, and air conditioning (HVAC) system, a thermostat, and a lighting unit.

15. The method according to claim 10, wherein: Triggering the control action associated with the room includes causing information related to the occupancy status and the confidence value to be displayed on one or more of a user interface disposed within the room and a user interface associated with a handheld device of a user.

16. The method according to claim 10, wherein: Triggering the control action associated with the room includes sending information related to the occupancy status and the confidence value to a room management unit, and the room management unit is configured to generate one or more of a maintenance alarm for the room and a tracking alarm for the room based on the determined occupancy status.

17. The method according to claim 10, wherein: The occupancy parameters include one or more of: motion, temperature, humidity, pressure, occupant heartbeat reading, proximity, noise, particulate matter (PM), volatile organic compounds (VOC), carbon dioxide (CO2), light intensity, lock operation, and plug power consumption.

18. The method according to claim 10, further comprising: receiving a user input from a user indicating a request to determine the occupancy status of the room; as well as In response to receiving the user input, one or more of the occupancy state and the confidence value are determined.

Images