>FIG. 6B, the control unit 104 identifies the unknown burning material i.e., the source of smoke/fire as wood. Furthermore, as the identified burning material i.e., wood falls in the database of alarming category, the control unit 104 triggers the alarm.

In another example, referring to FIGS. 7A and 7B, if CO₂and CO gases are detected in the smoke caused by an unknown burning material and the concentration of the detected CO₂and CO gases matches that of smoke generated while cooking as shown in FIG. 7A, and further if the temporal characteristic of fire caused by the unknown burning material also matches with the known temporal characteristics of fire caused during cooking as shown in FIG. 7B, the control unit 104 identifies the unknown burning material i.e., the source of smoke/fire as a cooking source. Furthermore, as the identified burning material i.e., wood falls in the database of the non-alarming (safe) category, the alarm system is not triggered by the control unit 104, thereby preventing false alarm generation.

In one or more embodiments, the system 100 and device 200 include the multi-type wave sensor that remains in communication with the control unit 104. The sensors are configured to capture multi-dimensional and temporal characteristics of the fire and smoke caused by unknown burning material or gases present in pollution. In such embodiments, the control unit 104 matches the captured multi-dimensional and temporal characteristics of the fire and smoke caused by the unknown burning material or pollution-causing material with the database 124 comprising a set of known multi-dimensional and temporal characteristics associated with smoke and fire caused by the plurality of known burning materials. Accordingly, the control unit 104 identifies and further classifies the unknown burning/polluting material as one of the known burning/polluting materials. Further, the control unit 104 also determines if the identified burning/polluting material falls within an alarming category or a non-alarming category as already explained above in detail.

In one or more embodiments, the system 100 or device 200 is implemented in an aspirating fire detection device 300 as shown in FIG. 3, where the aspirating device 300 receives the smoke present at the AOI or receives pollution from outside and facilitates the flow of the received smoke or pollution through the device 200 to enable identification and discrimination of the smoke and the corresponding burning material by the device 200. The detailed operation of the device 200 has been described later in conjunction with FIG. 3. In other embodiments, the device 200 or the components of the system 300 are directly installed at a predefined position in the AOI as shown in FIG. 4, where the device 200 or system 100 identifies and discriminates the smoke and the corresponding burning material present at the AOI.

Referring back to FIG. 1, the functional modules of the system 100 and control unit 104 are disclosed. Although the invention has been explained considering that system 100 is implemented by the control unit 104, it may be understood that system 100 may also be implemented in a variety of computing systems, such as a laptop computer, a desktop computer, a notebook, a workstation, a server, a network server, a cloud-based environment and the like. The control unit 104 comprises one or more processor(s) 106 (first processor) operatively coupled to a memory 108 (first memory). The processor 106 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, logic circuitries, and/or any devices that manipulate data based on operational instructions. Among other capabilities, processor 106 is configured to fetch and execute computer-readable instructions stored in the memory 108. The memory 108 may store one or more computer-readable instructions or routines, which may be fetched and executed to create or share the data units over a network service. The memory 108 may comprise any non-transitory storage device including, for example, volatile memory such as RAM, or non-volatile memory such as EPROM, flash memory, and the like.

The control unit 104 also comprises an interface(s) 110 that may comprise a variety of interfaces, for example, interfaces for data input and output devices, referred to as I/O devices, storage devices, and the like. The control unit 104 further includes a communication unit 112 such as WiFi module, transceiver, Bluetooth module, cellular connection module such as 2G, 3G, 4G, and 5G, and the like to facilitate communication of the control unit 104 with the sensors 102-A, 102-B, a fire system or alarm system, and mobile devices associated with users of the invention, through the network. The interface(s) 110 may also provide a communication pathway for one or more internal components or units of the control unit 104. Examples of such internal components include, but are not limited to, processing engine(s) 114 and database 124.

The processing engine(s) 114 is implemented as a combination of hardware and programming (for example, programmable instructions) to implement one or more functionalities of the processing engine(s) 114. In the examples described herein, such combinations of hardware and programming may be implemented in several different ways. For example, the programming for the processing engine(s) 114 may be processor-executable instructions stored on a non-transitory machine-readable storage medium and the hardware for the processing engine(s) 114 may comprise a processing resource (for example, one or more processors), to execute such instructions. In the present examples, the machine-readable storage medium may store instructions that, when executed by the processing resource, implement the processing engine(s). In such examples, the control unit 104 comprises the machine-readable storage medium storing the instructions and the processing resource to execute the instructions, or the machine-readable storage medium may be separate but accessible to server 102 and the processing resource. In other examples, the processing engine(s) may be implemented by an electronic circuitry. The database may comprise data that is either stored or generated as a result of functionalities implemented by any of the components of the processing engine(s) 114 or system 100.

In one or more embodiments, the processing engine(s) 114 includes a gas detection unit 116, a temporal characteristic unit 118, a Hidden Markov Models (HMM) unit or Redundant Neural Networks (RNN) unit 120, an alarm unit 122, and other units (s). Other unit(s) can supplement the functionalities of the processing engine or the control unit 104. The database 124 is pre-stored with the list of known burning (combustible) materials or fire sources, along with the corresponding gas data and temporal characteristic data of fire caused by all these known burning materials.

In one or more embodiments, the gas detection unit 116 enables the control unit 104 to allow the first sensors 102-A to detect gas present in the smoke caused by an unknown burning material and receive the detected gas data from the first sensors 102-A. The gas detection unit 116 also enables the control unit 104 to allow the first sensors 102-A to detect gas present in pollution at an AOI or outside and receive the detected gas data from the first sensors 102-A. The gas detection unit 116 further enables the control unit 104 to match the concentration of gases detected in the smoke or pollution with the database 124 comprising the concentration of known gases associated with smoke generated by the plurality of known burning materials or by the pollution-causing materials.

In one or more embodiments, the temporal characteristic unit 118 enables the control unit 104 to allow the second sensors 102-B to capture the temporal characteristic of fire caused by the unknown burning material and receive the captured temporal characteristic data from the second sensors 102-B. The temporal characteristic unit 118 further enables the control unit 104 to match the temporal characteristic of the fire caused by the unknown burning material with the database 124 comprising the temporal characteristic of fire caused by the plurality of known burning materials.

In one or more embodiments, temporal characteristic unit 118 may include a machine learning system configured to identify the burning material based on the temporal characteristics using one or more temporal models. For example, in one or more embodiments, the HMM/RNN unit 120 may be configured to enable the temporal characteristic unit to detect and classify the temporal characteristic of fire by using any of the temporal pattern recognition technologies, such as Hidden Markov Models (HMM) or Redundant Neural Networks (RNN), however, other temporal pattern recognition technologies may also be employed. For example, the machine learning models may include decision trees, support vector machines, regression analysis, Bayesian networks, random forest learning, dimensionality reduction algorithms, boosting algorithms, artificial neural networks (e.g., fully connected neural networks, deep convolutional neural networks, or recurrent neural networks) and/or other machine learning models. In one or more embodiments, the machine learning system may be configured to train the one or mode temporal models to identify the burning material. In one or more embodiments, the machine learning system may use one or more of supervised learning, semi-supervised, unsupervised learning, reinforcement learning, and/or other machine learning techniques.

Accordingly, the gas detection unit 116 and the temporal characteristic unit 118 enable the control unit 104 to identify and classify the unknown material as one of the known burning materials based on matching of the gas detected by the first sensors 102-A and the temporal characteristics captured by the second sensors 102-B, with the database 124 of known materials. In addition, the gas detection unit 116 and the temporal characteristic unit 118 enable the control unit 104 to match the identified burning material with the database 124 comprising the set of known safe burning materials and set of known unsafe burning materials, which enables the control unit 104 to discriminate the identified burning material into any of the safe category and the alarming category. Further, the gas detection unit 116 and the temporal characteristic unit 118 enable the control unit 104 to match the identified pollution causing material with the database 124 comprising the set of known safe polluting materials and set of known unsafe polluting materials, which enables the control unit 104 to discriminate the identified polluting material into any of the safe category and the alarming category. Furthermore, the gas detection unit 116 and the temporal characteristic unit 118 enable the control unit 104 to discriminate the identified unsafe burning/polluting material into an organic source material, a liquid source material, and a synthetic fire source material.

In one or more embodiments, when multi-type wave sensors are used in the system or device, the control unit 104 enables the multi-type wave sensors to capture multi-dimensional and temporal characteristics of the fire and smoke caused by the unknown burning/polluting material. Further, the gas detection unit 116 and the temporal characteristic unit 118 enable the control unit 104 to match the captured multi-dimensional and temporal characteristics of the fire and smoke with the database 124 comprising a set of known multi-dimensional and temporal characteristics associated with smoke and fire caused by the plurality of known burning/polluting materials, and correspondingly identify and classify the burning/polluting material as one of the known burning/polluting materials.

In one or more embodiments, the process of decision-making by the control unit 104 to identify and discriminate fire/smoke causing material, can be based on the decision-making theory of distributed systems, in particular voting algorithms. In such voting algorithms, a consensus is determined by the interaction of the different nodes of the system 100. For instance, a node may be the first sensor 102-A, the second sensor 102-B, the temporal ML engine (HMMRNN unit), a traditional smoke sensor or a gas sensor, and the like.

Each node, based on the result of its own processing and the result of other (some or all) nodes, emits a vote about the question to be answered. For example, the question can be “has a specific type of substance (say X) detected in the air?” or “is there a burning fire that needs to be signaled?” or even “the probability of detecting the specific material X is Y” or “the sensor that has a bigger probability of being active is Z”, and the likes. The votes of all nodes are transmitted to the control unit 104. Further, using one of a variety of voting algorithms (exact/inexact/approval/and the likes) a consensus is met, and accordingly, a decision is made. This kind of decision-making may be safer and may help minimize the number of false positives and false negatives.

Further, in order to minimize the number of false negatives, which is a safety requirement, a (negative) veto may be employed in the algorithm in one or more embodiments. For example, the system 100 may at least incorporate the sensor 102, so that it can be used as a baseline for safety. For instance, even if the voting algorithm consensus suggests that there is no fire and the sensor 102 decides that there is a fire, a fire event (e.g., a fire alarm) is triggered by the control unit 104. This may help lower the number of false negatives, thereby meeting another safety requirement.

Accordingly, the control unit 104 can identify and classify the burning material as one of the known burning materials based on a voting consensus between one or more nodes including the smoke sensors or gas sensor, and the HMM/RNN unit associated with the control unit 104.

In one or more embodiments, the alarm unit 122 enables the control unit 104 to actuate the alarm system of the device or a fire alarm system associated with the AOI when the burning/polluting material is identified to be in the alarming category. Further, the alarm unit 122 also enables the control unit 104 to actuate the alarm system of the device or the fire alarm system when the concentration of the detected gas in the smoke caused by the burning/polluting material exceeds a first threshold value, and/or a level of the temporal characteristics of the fire caused by the burning/polluting material exceeds a second threshold value.

Referring back to FIG. 2, an exemplary block diagram of the device 200 is provided. The device 200 includes a processor 202 (second processor), a memory 204 (second memory), the first sensors 102-A, the second sensors 102-B, an alarm 206, and a communication unit (such as a transceiver) 208, being configured within a single housing that can be easily positioned at a predefined location. The processor 202 is coupled to the memory 204, the sensors 102-A, 102-B, the transceiver 208, and the alarm 206. The processor 202 and memory 204 collectively form the control unit 104, where the processor 202 includes suitable logic, circuitry, and/or interfaces that are operable to execute one or more instructions stored in the memory to perform a predetermined operation. The memory 204 may be operable to store one or more instructions. The processor 202 may be implemented using the processors known in the art. Some of the commonly known memory implementations include, but are not limited to, a random access memory (RAM), a read only memory (ROM), a hard disk drive (HDD), and a secure digital (SD) card. Further, the memory includes the one or more instructions that are executable by the processor to perform specific operations. It is apparent to a person having ordinary skills in the art that the one or more instructions stored in the memory enable the hardware of the device to perform the predetermined operations.

In some embodiments, the first sensors 102-A of the device 200 are operable to detect one or more gas present in the smoke generated by burning/polluting material. The first sensors 102-A include gas sensors including but are not limited to a CO₂sensor, CO sensor, and the likes, which are already known in the art. The second sensors 102-B of the device 200 are operable to capture temporal characteristics of fire caused by the burning material. The second sensors 102-B include but are not limited to temperature sensors, light sensors, and the like, which are already known in the art. The data captured by the sensors 102-A, 102-B are then transmitted to the processor or control unit 104 for further processing as already explained in detail in FIG. 1.

In one or more embodiments, the sensors are multi-type wave sensors that are operable to capture multi-dimensional and temporal characteristics of the fire and smoke caused by the burning/polluting material. Those skilled in the art would appreciate that the use of a multi-type wave sensor allows the detection of reflected light waves of different wavelengths while detecting the gases and gas concentration in the smoke/pollution. As a result, the individual performance of the sensors is improved and the overall smoke/fire discrimination capability of the invention is also improved.

The communication unit 208 (transceiver) transmits captured data to the mobile devices of the users or to the fire detection and alarm system. Examples of the transceiver 208 may include but are not limited to, an antenna, an Ethernet port, an USB port, or any other port that can be configured to transmit the captured data.

The transceiver 208 transmits data in accordance with the various communication protocols, such as TCP/IP, UDP, and 2G, 3G, or 4G communication protocols. The alarm 206 of the device 200 includes an LED, a siren, and/or a buzzer. The device 200 upon detection of an actual fire, triggers the alarm of the device or the fire detection and alarm system to alert about the fire/smoke. The device 200 further includes a power source such as a battery that is within the housing of the device 200 to supply electrical power to the components of the device. Besides, the device 200 may also be electrically connected to a battery of the fire detection and alarm system already installed at the AOI.

In one or more embodiments, the device 200 includes a display panel 210 indicating the status of the device 200. If the source of the smoke (i.e., the burning material) is identified to be in the non-alarming (safe) category, the device 200 does not activate the alarm but is acknowledged as a textual message displayed on the display 210. Further, if the burning material is identified to be in the alarming category, the device 200 activates the alarm 206 so that the occupants of the AOI are alerted to the presence of the fire and the burning/polluting material. The display 210 also acknowledges the identified burning material and any further indication of what the identified burning material or fire source is.

Referring back to FIG. 3, the invention being implemented in an aspirating fire detection device 300 is illustrated. The aspirating device 300 receives the smoke present at an AOI or environment 302 and facilitates the flow of the received smoke through the sensors 102-A, 102-B to enable identification and discrimination of the smoke and the corresponding burning material. The aspirating device 300 includes a central detection unit 304 having a hollow housing 306 enclosing the sensors 102-A, 102-B, which is fluidically connected to a sampling pipe 308. The sampling pipe 308 is exposed to the AOI or environment 302 via a plurality of sampling holes 310 in the sample pipe 308, such that the central detection unit 304 is in fluid communication with the AOI. The central detection unit 304 is in wired and/or wireless communication with the control unit 104. The control unit 104 is also in wired and/or wireless communication with an alarm system 316 and a fire mitigation device 318, both of which are present/installed at a desired location in the AOI or environment 302.

The central detection unit 304 includes an inlet 312 that is connected to the sampling pipe 308. The central detection unit 304 accommodates the device 200 (not shown) or sensors 102-A, 102-B and an outlet 314. Air is aspirated into the central detection unit 304 using an aspirator (not shown), which draws smoke along with the air and pollutants from the AOI/environment 302 through the sampling pipe 308 and into the central detection unit 304 via the inlet 312. The sampled air along with smoke present at the AOI 302 is passed to the first sensors 102-A. The sampled air and smoke are then exhausted from the central detection unit 304 via the outlet 314. The flow of air and smoke through the central detection unit 304 is indicated by the arrows as shown in FIG. 3.

The aspirated air and smoke are allowed to flow into the central detection unit 304 upon the rotation of an impeller of the aspirator. The aspirated air and smoke are then passed to the first sensors 102-A, before being exhausted from the housing via the outlet 314, which allows the first sensors 102-A to detect the concentration of gases present in the received smoke or pollution. Further, the second sensors 102-B are housed in the housing 306 of the central detection unit 304 such that the second sensors 102-B can capture the temporal characteristic of the fire caused by the burning/polluting material in the AOI/environment 302 without any interruption. The control unit 104 identifies and discriminates the smoke and the corresponding burning/polluting material present at the AOI as explained above in detail in