DISPENSING DEVICE FOR MONITORING A PIPE SYSTEM AND SYSTEM FOR MONITORING A PIPE SYSTEM

Abstract

Sensing device for monitoring a pipe system (10), wherein the sensing device comprises: at least one sensor (102, 104) designed to detect at least one operating characteristic of a pipe section (12) of the pipe system (10), wherein the at least one sensor (102, 104) comprises a transducer (130) designed such that an audio signal emanating from the pipe section (12) causes the transducer (130) to generate a voltage signal indicative of the audio signal; a wake-up circuit (170) which is operatively connected to the converter (130) and is designed to generate a wake-up signal in response to the voltage signal generated by the converter (130) exceeding a predetermined threshold; and a processor (172) designed to identify an operating condition of the pipe system (10) with reference to the at least one operating characteristic sensed in response to the reception of the wake-up signal from at least one sensor (102, 104). a memory (174) that stores: first audio data indicative of the audio signal emanating from the pipeline section (12); and second audio data comprising an acoustic signature corresponding to a particular event occurring in at least one of the pipeline section (12) and an area (14) surrounding the pipeline section (12), wherein the particular event corresponds to an operating condition of the pipe system (10); wherein the processor (172) is designed to identify the operating condition of the pipe system (10) by comparing the audio signal from the first audio data and the acoustic signature of the second audio data, and to determine that the audio signal is indicative of the particular event; and wherein the at least one sensor (102, 104) further comprises a microphone (164) designed to detect a further audio signal emanating from an area (14) surrounding the pipe section (12); and The processor (172) is further designed to apply a noise reduction algorithm to the audio signal emanating from the pipe section (12) with reference to the further audio signal.

Description

field of technology

[0001] The present invention relates to a sensing device for monitoring a pipe system and a system for monitoring a pipe system.

[0002] WO 2017 / 175 136 A1 discloses a sensing device comprising at least one sensor, a wake-up circuit, and a processor. The at least one sensor is designed to sense the operating characteristic of a section of the pipe system and includes a transducer configured such that an audio signal emanating from the pipe section causes the transducer to generate a voltage signal indicative of the audio signal. The wake-up circuit is operatively connected to the transducer and generates a wake-up signal in response to the voltage signal exceeding a predetermined threshold. The processor is operatively connected to the wake-up circuit and the at least one sensor and is designed to identify an operating condition of the pipe system based on the operating characteristic sensed by the at least one sensor in response to receiving the wake-up signal.

[0003] DE 38 88 919 T2 discloses an ultrasonic sensor and a control system and a device for remotely detecting leaks in pressure systems by detecting ultrasonic vibrations.

[0004] US Patent 9,664,589 B2 discloses a system for the acoustic detection of fluid leaks, wherein a processing component is in communication with a plurality of sensor units, which is configured to perform an analysis of combinations of acoustic signature data acquired by the plurality of sensor units, wherein the analysis has a sensitivity of the system leak detection, wherein the analysis performed by the processing component comprises a) identifying an acoustic frequency range that is active during a fluid flow event from one sensor in the plurality of sensors, and b) adjusting other sensors in the plurality of sensors to focus on a subset of the acoustic frequency range.

[0005] WO 2017 / 214 729 A1 discloses a method for estimating the flow rate in a pipeline based on the pipeline's acoustic behavior. Initial acoustic data are measured from the pipeline. Subsequently, the fluid flow rate in the pipeline is estimated. This estimate is based on the initial acoustic data and a correlation established between subsequent acoustic data and corresponding flow data from an experimental pipeline. The correlation is generated using a machine learning process (which may involve the use of an artificial neural network such as an autoencoder). The subsequent acoustic data and the corresponding flow data serve as inputs for the machine learning process. background

[0006] Piping systems are typically installed underground, in walls or ceilings, or in other locations where the pipes are not fully visible. As a result, if a leak occurs, it may remain undetected for some time, and its exact location can be difficult to pinpoint. Leaks in a piping system not only release fluid into the environment but also allow contaminants to enter the system. Promptly detecting and locating a leak can help minimize these risks.

[0007] Leak detection techniques have been developed that utilize pressure sensors. Fluids generally move within a pipe system under pressure. A leak can cause a pressure drop in the system, which can be detected by a pressure sensor. In some cases, the pressure drop caused by a leak may be indistinguishable from normal pressure head losses in a system, resulting in a delay before the leak is detected.

[0008] Audio leak detection techniques have also been developed. In one example, a pressure drop in a pipe caused by a leak induces a pressure fluctuation that sends an audio signal, detectable by a vibration sensor, upstream and downstream from the leak. Correlating vibration detections at multiple locations in a system allows the location of the leak to be determined. In another example, ultrasonic signals are transmitted through targeted sections of a pipe, using the effect of the fluid in the pipe on the signal to detect a leak. These conventional systems have a number of disadvantages, as discussed below.

[0009] Vibrations at various frequencies are generated by a number of sources, resulting in frequent false positive leak detections in such systems. In one conventional system, leak detection is only performed during "quiet hours" when external noise, such as from traffic or construction work, is minimized. The disadvantage is that a leak may exist for a considerable time before these quiet hours begin and detection starts.

[0010] Conventional systems are generally either continuously operating or only check for leaks at preset intervals. Continuously operating systems require a significant amount of power, which can drain a sensor battery or necessitate an expensive and complex wired power connection. Interval systems can delay the detection of a leak occurring between scheduled leak detections.

[0011] Furthermore, there are cases where a leak might not generate an audio signal detectable by a sensor, leading to false negative results. The leak might be outside the sensor's range, the signal might be too weak to be detected, or the audio signal might be masked by ambient noise.

[0012] Furthermore, systems that utilize multiple sensors across a pipe network generally require a centrally located node or server to process sensor data and locate leaks. A central node not only increases the complexity and cost of a system, but also acts as a single point of failure. Additionally, the requirement for all sensors to communicate with a central node increases the complexity and cost of communication within the system.

[0013] Therefore, a leak detection system that reduces false readings would be useful. A system that can be used in a wide range of ambient noise conditions would also be useful. A system that does not require a central hub would also be useful. A system capable of detecting leaks under a variety of circumstances would also be useful. Brief description

[0014] To monitor a pipe system without downtime between monitoring intervals, a monitoring system is designed to wake up in response to audio signals emanating from sections of the pipe system.

[0015] The invention provides a sensing device for monitoring a pipe system according to claim 1 and a system for monitoring a pipe system according to claim 7.

[0016] In some embodiments, the sensing device further comprises a mounting mechanism designed for attaching the sensing device to the pipe section. The transducer is positioned on the mounting mechanism such that, when the sensing device is mounted to the pipe section via the mounting mechanism, the transducer is in direct contact with a surface of the pipe section.

[0017] In some embodiments, the at least one sensor further comprises a temperature sensor designed to sense the temperature of the pipe section and a humidity sensor designed to sense humidity in an area around the pipe section. The temperature sensor is positioned on the mounting mechanism such that it is in direct contact with the surface of the pipe section when the sensing device is mounted on the pipe section via the mounting mechanism.

[0018] In some embodiments, the sensing device further comprises a memory that stores a first machine learning algorithm and initial data. The first machine learning algorithm can be executed by the processor to determine, over time, a normal operating condition of the pipe system with reference to the at least one sensor. The initial data are indicative of the normal operating condition of the pipe system. The processor is further designed to identify a fault in the pipe system by determining that the identified operating condition deviates from the normal operating condition.

[0019] According to the invention, the sensing device comprises a memory that stores first audio data and second audio data. The first audio data are indicative of the audio signal emanating from the pipe section. The second audio data comprise an acoustic signature corresponding to a specific event occurring in at least one area within the pipe section and in a region surrounding the pipe section. This specific event corresponds to an operating condition of the pipe system. The processor is designed to identify the operating condition of the pipe system by comparing the audio signal from the first audio data with the acoustic signature of the second audio data, and to determine that the audio signal is indicative of the specific event.

[0020] According to the invention, the at least one sensor comprises a microphone designed to detect a further audio signal emanating from an area surrounding the pipe section. The processor is further designed to apply a noise reduction algorithm to the audio signal emanating from the pipe section, taking into account the further audio signal.

[0021] In some embodiments, the memory also stores a second machine learning algorithm that can be executed by the processor to identify a correspondence between a relevant operating condition of the pipe system and a relevant specific event.

[0022] In some embodiments, the first sensing device further comprises a communication module operatively connected to the processor, which can be operated at least by sending and receiving information relating to the operating condition of the pipe system to and from an external device.

[0023] In some embodiments, the system further includes an actuator. The actuator comprises a valve element and a second communication module. The valve element is operable to selectively restrict and allow the flow through a section of the pipe system. The information transmitted by the first communication module includes an activation instruction, and the second communication module is designed to receive the activation instruction and activate the valve element in response to the received activation instruction.

[0024] In some embodiments, the system further comprises a remote computing device. The remote computing device includes a third communication module, an output device, and an additional processor. The third communication module is configured for at least one of the following: sending information to and receiving information from the sensor. The additional processor is operatively connected to the third communication module and the output device and is configured to output information received from the third communication module regarding the operating condition of the pipe system via the output device.

[0025] In some embodiments, the additional processor is designed to send further information regarding the operating conditions of the pipe system to the sensor. The memory stores initial data corresponding to the operating conditions of the pipe system. Upon receiving this additional information from the remote computing device, the processor is designed to update the initial data based on this additional information.

[0026] In some embodiments, the additional information sent by the remote computing device includes an activation instruction. The second communication module is designed to receive the activation instruction and, in response to the received activation instruction, to activate the valve element. Brief description of the drawings Fig. Figure 1 shows a schematic representation of a system with a sensing device for a pipe system. Fig. Figure 2 shows a side view of the first sensor of the system's sensing device. Fig. 1. Fig. Figure 3 shows a perspective front view of the first sensor of the sensing device. Fig. 1. Fig. Figure 4 shows a bottom view of the first sensor of the sensing device. Fig. 1. Fig. Figure 5 shows a top view of the first sensor of the sensing device of Fig. 1. Fig. Figure 6 shows a schematic representation of a sensing device according to an embodiment of the present invention. Fig. Figure 7 shows a flowchart of an exemplary process for operating the sensing device of Fig. 1. Fig. Figure 8 shows a flowchart of another exemplary process for operating the sensing device of Fig. 1. Fig. Figure 9 shows a flowchart of another exemplary process for operating the sensing device of Fig. 1. Fig. Figure 10 shows a flowchart of another exemplary process for operating the sensing device of Fig. 1. Fig. Figure 11 shows a flowchart of an exemplary process for operating the sensing device of Fig. 1. Fig. Figure 12 shows another flowchart of an exemplary process for operating the sensing device of Fig. 1. Detailed description

[0027] To provide an understanding of the principles of the embodiments described herein, reference is now made to the drawings and descriptions in the following written description. This reference is not intended to limit the scope of protection of the invention. This disclosure also includes any changes and modifications of the embodiments shown and further applications of the principles of the described embodiments, as would normally be apparent to a person skilled in the art in the field to which this document relates.

[0028] Fig. Figure 1 shows a schematic representation of an exemplary embodiment of a system 100 with a sensing device for leak detection in a pipe system 10. The system 100 comprises a first sensor device 102, a second sensor device 104, an actuator 106 and a remote computing device 108. The elements of the system 100 are interconnected via a data network 125.

[0029] The first sensor 102 is mounted on a section of a pipe 12 of the piping system 10. As explained in more detail below, the first sensor 102 is designed to detect operating conditions of the section of pipe 12 and of an area 14 around the section of pipe 12, and the system 100 is designed to infer a status of the section of pipe 12 and / or the piping system 10 with reference to the operating conditions detected by the first sensor 102 and to selectively perform an operation with reference to the inferred status. For illustrative purposes, in some examples, a detected operating condition includes a temperature, pressure, humidity, and / or acoustic measurement of the pipe section 12 and / or the area 14. In some examples, an inferred status of the piping system 10 includes a fault in the pipe section 12 and / or the piping system 10.In some examples, an operation performed in response to an inferred status includes reporting a fault to the remote computing device 108, activating the first sensor 102 and / or the second sensor 104 and / or actuating the actuator 106.

[0030] In this example, the second sensor 104 comprises similar components to the first sensor 102 and is similarly designed. The second sensor 104 is mounted on a different section of the pipe 12' and, like the first sensor 102, is designed to detect operating conditions of the section of pipe 12' and an area 14' around the section of pipe 12'. The system 100 is further designed to infer a status of the pipe system 10 with reference to the operating conditions detected by the second sensor 104 and to selectively perform an operation with reference to the inferred status. In some examples, the system 100 is further designed to infer a status of the pipe system 10 with reference to the operating conditions detected by several sensors in combination.Referencing multiple sensors allows System 100 to draw conclusions about the operational status of pipe system 10, which might otherwise be undetectable or incorrect. While System 100 in this example includes two sensors, 102 and 104, System 100 in other examples includes a different number of sensors. Some examples include only one sensor, and others include more than two.

[0031] The actuator 106 is positioned on a section of pipe 20 of the pipe system and comprises a communication module 110 and a valve element 112. The communication module 110 is designed to receive an activation instruction from other devices, such as the first sensor 102, the second sensor 104, and the remote computing device 108, and to activate the valve element 112 in response to the received activation instruction. In some examples, the communication module 110 is designed to act as a relay and transmit communications from one device to another, such as from the first sensor 102 to the second sensor 104 or to the remote computing device 108, or vice versa.

[0032] The valve element 112 can be selectively operated to allow and restrict flow through the pipe section 20. Any acceptable type of valve element can be used. In some examples, the actuator 106 is assigned to the first sensor 102, with the pipe section 12 for the first sensor 102 being located near the pipe section 20, so that the actuator 106 is configured to allow and restrict flow through the pipe section 12. In some examples, the system 100 includes an assigned actuator for each sensor. In some examples, the system 100 is configured to dynamically assign an actuator to a sensor based on information about the pipe system and a detected disturbance.

[0033] In various examples, the remote computing device 108 is any acceptable computing device, such as a personal computer, a tablet computer, a mobile phone, or the like. In some embodiments, the remote computing device 108 comprises a server designed to communicate with one or more client devices. The remote computing device 108 comprises a processor 130, a memory 132, a communication module 134, an input device 136, and an output device 138.

[0034] The processor 130 is operatively connected to the memory 132, the communication device 134, the input device 136, and the output device 138, and is designed to execute programming instructions stored in the memory 132. The memory 132 is designed to store data relating to the pipe system 10, such as historical usage information, sensor data, audio data such as acoustic profiles for ambient noise sources, and other data.

[0035] The communication device 134 is designed to send and receive transmissions from other components of the system 100, such as the first sensor 102, the second sensor 104, the actuator 106, and other devices remote from the system 100, via the data network 125. In various examples, the data network 125 enables the transmission of data and information via WiFi, BTE, LoRa, a lightweight messaging protocol such as MQTT, or any other acceptable communication protocol. In some examples, the data network 125 includes a connection via the internet. In some examples, the communication device 134 provides other devices with access to the memory 132 and / or retrieves data from other devices and stores the retrieved data in the memory 132.

[0036] Input device 136 includes any acceptable device capable of receiving input from a user. Examples include a keyboard, a mouse, a touchscreen, and the like. Output device 138 includes any acceptable device capable of generating output for the user. Examples include a visual display, an audio device, an indicator light, and the like.

[0037] While the system 100 in Fig. While one example comprises a single remote computing device (108), other examples include any number of remote computing devices. In some examples, the system (100) does not include any remote computing device.

[0038] Fig. Figures 2-5 each show a side view, perspective front view, bottom view and top view of the first sensor 102. Fig. 1. As in Fig. As shown in Figure 2, the first sensor 102 comprises a base section 120 and an electronics section 160. The base section 120 includes a base plate 122 and a fixing mechanism 124. The base plate 120 supports the electronics section 160. The fixing mechanism 124 is for fixing the first sensor 102 onto the tube section 12 ( Fig. 1) designed.

[0039] In this example, the fixing mechanism 124 comprises a first clamp 126 and a second clamp 128. Other embodiments use different numbers of clamps and other fixing mechanisms. In this example, the clamps 126 and 128 are dimensioned to correspond to a size of the pipe section 12. In some examples, the clamps and / or other fixing mechanisms are designed to be adjustable to accommodate pipe sections of different sizes.

[0040] Referring to the underside view of the first sensor 102 in Fig. 4 The first clamp 126 includes an acoustic transducer 130. The transducer 130 is positioned such that when the clamp 126 engages with the pipe section 12 to mount the first sensor 102, the transducer 130 engages with the pipe section 12 and is designed to detect an audio signal from a surface 16 of the pipe section 12. In some embodiments, the clamp 126 is designed to exert a force when attached to the pipe section 12, which serves to maintain a firm contact between the transducer 130 and the surface 16 of the pipe section 12.

[0041] The transducer 130 is further designed to generate a voltage indicative of the signal, where a voltage level corresponds to an audio level (dB) of the signal. In various embodiments, the transducer 130 is analog, digital, or a combination thereof. In some embodiments, the transducer 130 comprises one or more vibration sensors, piezoelectric contact sensors, and audio contact sensors. In this example, the transducer 130 is designed to be powered by vibrations, so that the voltage generated by the transducer 130 is induced by the signal being received by the transducer 130. Because the transducer 130 is powered by induced energy, it does not require a power supply from the first sensor 102 to function.

[0042] In some examples, the first sensor 102 further includes an analog-to-digital converter (not shown) designed to convert an analog signal from the converter 130 into a digital signal. In various examples, the converter is included in the converter 130, the clamp 126, the electronics section 160, or is located at any other acceptable location on the first sensor 102. In several other examples, the converter 130 is located at any other acceptable location on the sensor 102 such that the converter 130 maintains surface contact with the surface 16 of the pipe section 12 when the first sensor 102 is mounted thereon.

[0043] The clamp 128 includes a temperature sensor 132. The temperature sensor 132 is positioned so that it is in direct contact with the surface 16 of the pipe section 12 when the clamp 128 is engaged with the pipe section 12, and is designed to measure an operating temperature of the pipe section 12. Any acceptable temperature sensor may be used. In some examples, the temperature sensor 132 is a contact temperature sensor, and the clamp 128 is designed to exert a force when fixed to the pipe section 12, which serves to maintain a firm contact between the temperature sensor 132 and the surface 16 of the pipe section 12. In other examples, the temperature sensor 132 is included on the clamp 126 or at any other location on the first sensor 102.

[0044] Electronics section 160 ( Fig. 2) comprises a body 162 which contains several electronic components, such as a power supply designed to store energy and deliver the energy to the sensor 102 (not shown), for example a battery.

[0045] Fig. Figure 6 shows a schematic representation of a sensing device according to an embodiment of the present invention, in particular a schematic representation for the electronics section 160 of the first sensor 102. As shown in Fig. As shown in Figure 6, the electronics section 160 further comprises a microphone 164, a humidity sensor 166, an LED status indicator 168, a wake-up circuit 170, a processor 172, a memory 174, a communication module 176 and an electronic circuit board 182.

[0046] The microphone 164, the humidity sensor 166 and the LED status indicator 168 are mounted on a top 170 of the body 162 ( Fig. 5) to be exposed to the area 14 surrounding the pipe section 12. The microphone 164 is designed to detect audio signals from the area 14 surrounding the pipe section 12. In this embodiment, the microphone 164 is designed to detect audio signals originating from outside the pipe section 12. The humidity sensor 166 is designed to detect ambient humidity in the area 14 surrounding the pipe section 12. The display 168 is designed to provide a visual signal to the user, such as whether the sensor 102 is powered, processing information, communicating with the remote computing device 108, or is in an alarm state.

[0047] Referring to Fig. In example 6, the processor 172, the memory 174, and the communication module 176 are mounted together on the electronic circuit board 182, which is housed in the body 162. In other examples, different components are mounted on one or more electronic circuit boards or are mounted elsewhere in the body 162 or elsewhere in the sensor 102.

[0048] The temperature sensor 132, the microphone 164, the humidity sensor 166, the display 168, the memory 174, and the communication module 176 are each operatively connected to the processor 172. These components are also operatively connected to the power supply (not shown).

[0049] The processor 172 is designed to selectively operate the first sensor 102 between a standby mode (i.e., a power-saving mode) and an active mode (i.e., a full-power mode). In standby mode, the processor 172 is designed to minimize power consumption, for example, to maximize battery life. While most of the components of the first sensor 102 remain in a low-power mode or are without power, the processor 172 is, in some examples, designed to query the non-acoustic sensors, i.e., the temperature sensor 132 and the humidity sensor 166, at regular, predetermined intervals.In some examples, the processor 172 is designed to operate the first sensor 102 in standby mode by default and to switch the operation of the first sensor 102 to active mode only in response to a query, a signal from another device, or an instruction from the wake-up circuit 170, as explained in more detail below. In active mode, the processor 172 is designed to determine an operating state of the pipeline section 12 with reference to the temperature sensor 132, the humidity sensor 166, the transducer 130, and the microphone 164, and to perform an operation based on this determination, as explained in more detail below.

[0050] The wake-up circuit 170 is operatively connected between the transducer 130 and the processor 172 and is designed to send an instruction to the processor 172 that causes the processor 172 to switch the first sensor 102 into active mode in response to the voltage from the transducer 130 exceeding a predetermined threshold. In other words, since the voltage generated by the transducer 130 is based on the audio level (dB) of an audio signal emanating from the pipe section 12, an audio signal sufficient to induce a voltage in the transducer above the predetermined threshold causes the wake-up circuit 170 to instruct the processor 172 to switch to active mode.

[0051] The storage unit 174 stores historical information about the pipe system 10, such as water consumption, temperature, humidity, acoustic characteristics, etc. The storage unit 174 also stores predetermined operating ranges for the pipe system, such as a temperature operating range, humidity threshold, and pressure operating range.

[0052] In this embodiment, the memory 174 further stores audio data comprising features of acoustic signatures corresponding to specific events. Examples of a specific event include events indicative of disturbance, such as water dripping, spraying, or splashing sounds, as well as events not indicative of disturbance, such as ambient noise, machine noise, and noise emanating from devices utilizing the pipe system 10, such as appliances like a washing machine, shower, toilet, dishwasher, etc., or other devices like pumps, filters, boilers, etc. An acoustic signature is an audio signature associated with or generated by a specific event. Features of an acoustic signature include any acceptable descriptive aspect of the audio signal. In some embodiments, features are Mel frequency cepstrum coefficients (“MFCCs”) extracted from the audio signal.

[0053] In some embodiments, memory 174 also stores one or more machine learning algorithms. For example, in some embodiments, memory 174 includes a first machine learning algorithm that can be executed to determine operating characteristics and flow patterns of the pipe system 10 based on signals from the first sensor 102 and / or other devices in the system 100, such as the second sensor 104, actuator 106, and remote computing device 108. In some embodiments, the first machine learning algorithm operates with reference to historical usage information and / or audio data acquired by sensor 102 and / or other devices in the system 100. In some embodiments, the first machine learning algorithm can be executed to assign the actuator 106 to a disturbance detected in the pipe system 10.In other words, the first machine learning algorithm is executable to determine that a disturbance in the pipe system 10 is located such that activating the actuator 106 weakens or reduces the effect of the disturbance.

[0054] In some embodiments, the memory 174 includes a second machine learning algorithm that can be executed to associate an audio signal received via the transducer 130 and / or the microphone 164 with a specific event. In some embodiments, the second machine learning algorithm is designed to operate with reference to user instructions, such as the user identifying the specific event. In some embodiments, the second machine learning algorithm is designed to identify a received audio signal as a disturbance event or a non-disturbance event over a period of time associated with the audio signal, taking into account the operating conditions of the pipeline section 12 and / or the pipe system 10.

[0055] Memory 174 also stores pre-trained and / or learned classification parameters that can be used with one or more machine learning algorithms. In various embodiments, the parameters and / or the one or more machine learning algorithms are preloaded into memory 174, accumulated and updated over time, and / or a combination thereof. In some embodiments, the one or more machine learning algorithms comprise at least one from a deep learning algorithm, a nearest-neighbor algorithm, a support-vector algorithm, a convolutional network, and any other acceptable machine learning technique. Memory 174 further stores a noise suppression algorithm that can be executed to isolate audio signals emanating from the pipe section 12 from ambient noise in the area 14 surrounding the pipe section 12.

[0056] The communication module 176 is operable to enable communication between the first sensor 102 and one or more of the second sensor 104, the actuator 106, the remote computing device 108, and other devices. In some embodiments, the communication module 176 is designed to send and receive transmissions over a data network such as the Internet. In some examples, the communication module 176 transmits data via WiFi, BTE, LoRa, a lightweight messaging protocol such as MQTT, or any other acceptable communication protocol. In some embodiments, the communication module 176 provides other devices with access to the memory 174 and / or retrieves data from other devices and stores the retrieved data in the memory 174.In some examples, the communication module 176 and the communication device 134 are designed to work together, so that the memory 174 and the memory 132 operate as a networked memory. As explained in more detail below, in various examples, communication includes one or more indicative information for the operating relationship between the pipeline section 12 and the piping system 10, historical information about the piping system 10, fault indications in the pipeline section 12 and / or the piping system 10, instructions, and other data.

[0057] Fig. Figure 7 shows a flowchart of an example process for operating the first sensor 102 in standby mode. At block 702, the processor 172 operates the first sensor 102 in standby mode. At block 704, the processor determines the operating state of the pipe section 12. In some examples, the processor 172 queries the temperature sensor 132 and the humidity sensor 166 at regular intervals and makes the determination based on the query. In some examples, determining that the pipe section 12 is not operating normally involves receiving an indication from the temperature sensor 132 that the pipe section 12 is operating outside its operating temperature range, i.e., below a freezing point or above a temperature / pressure limit. Operation outside the operating temperature range could cause the pipe section 12 to burst due to freezing or overpressure caused by heat.In some examples, determining that pipe section 12 is not operating normally involves receiving a signal from humidity sensor 166 indicating that the ambient humidity in area 14 around pipe section 12 is above the predetermined humidity threshold. Excessive humidity in the area around pipe 108 may indicate a leak in a section of the pipe that is not actively monitored. In some examples, the determination is based on a machine learning algorithm, such as the second machine learning algorithm described above.

[0058] In block 706, in response to determining that pipe section 12 is in a normal operating state, processor 172 stores information regarding the normal operating state in memory 174. In some examples, processor 172 periodically operates the communication module 176 to send data relating to the historical use and operation of pipe section 12 to remote computing device 108 and / or other devices, and to receive information regarding the historical use and operation of the pipe system 10. In some examples, such data includes audio data relating to potential sources of ambient noise in and around the pipe system 10. In some examples, processor 172 uses the data regarding the historical use and operation of pipe section 12 to update the one or more machine learning algorithms stored in memory 174.

[0059] In block 708, the processor 172 is designed to switch the first sensor 102 into active mode in response to a detection that the pipeline section 12 is not operating in a normal state.

[0060] Fig. Figure 8 shows a flowchart of another exemplary process for operating the first sensor 102 in standby mode. At block 802, the processor 172 operates the first sensor 102 in standby mode. At block 804, an audio signal in the pipe section 12 causes the transducer 130 to generate a voltage signal above the predetermined threshold for the wake-up circuit 170. At block 806, in response to the voltage signal exceeding the predetermined threshold, the wake-up circuit 170 transmits a wake-up instruction to the processor 172, and at block 808, the processor 172 switches the operation of the first sensor 102 to active mode.

[0061] Fig. Figure 9 shows a flowchart of another exemplary process for operating the first sensor 102 in standby mode. In block 902, processor 172 operates the first sensor 102 in standby mode. In block 904, processor 108 receives a transmission via communication module 176 containing a wake-up instruction for sensor 102. In some examples, the wake-up instruction is received from remote computer 108 as a result of user interaction with the remote computer. In other examples, the wake-up instruction is received from remote computer 108 in response to a detection of a fault in the pipe system 10. In some examples, the wake-up instruction is received from the second sensor 104, such as in response to the detection of a fault in the pipe system 10 by the second sensor 104.At block 906, in response to the wake-up instruction, processor 108 initiates the operation of the first sensor 102 into active mode.

[0062] Fig. Figure 10 shows a flowchart of an example process for operating the first sensor 102 in active mode. At block 1002, the processor 172 operates the first sensor 102 in active mode. At block 1004, the processor 172 determines, with reference to readings from one or more of the transducer 130, the temperature sensor 132, the microphone 164, and the humidity sensor 166, whether the pipeline section 12 and / or the pipe system 10 is operating outside of its normal operating state.

[0063] In block 1006, the processor 172 is designed to transfer the sensor's operation to standby mode in response to a determination that the pipeline section 12 and the pipe system 10 are operating normally. In some embodiments, the processor 172 additionally uses the communication module 176 to send a no-fault indication to the remote computing device 108 and / or another device in the system 100, such as a device from which the first sensor 102 received a wake-up instruction. In some examples, the processor 172 additionally stores information from one or more transducers 130, temperature sensors 132, microphones 164, and humidity sensors 166 in memory 174 and / or transmits the information to another device in the system 100. In some examples, the processor 172 uses the information to update the one or more machine learning algorithms in memory 174.

[0064] In block 1008, the processor 172 is designed to identify a fault in response to a detection that the piping section 12 and / or the piping system 10 are not operating in normal conditions.In various embodiments, the identification is based on one or more of the following: (i) a signal received from another device in the system 100, such as the second sensor 104 or the remote computing device 108; (ii) a signal from the temperature sensor 132 indicating that the pipeline section 12 is operating outside the operating temperature range; (iii) a signal from the humidity sensor 166 indicating that the ambient humidity in the area 14 around the pipeline section 12 is above the predetermined humidity threshold; and (iv) a determination that an acoustic signal in the audio data from one or more transducers 130 and microphones 164 is indicative of a disturbance, as will be explained in more detail below.

[0065] In various examples, the identified disturbance comprises one or more temperature disturbances, humidity disturbances, flow disturbances, pressure disturbances, leakage disturbances, or any other type of disturbance that affects the operation of the piping system 10. In some examples, identifying a disturbance includes identifying the location of the disturbance within the piping system 10. In some embodiments, the location of the disturbance is identified with reference to information from one or more transducers 130, temperature sensors 132, microphones 164, and humidity sensors 166.

[0066] At block 1010, processor 172 performs an operation to mitigate the identified disturbance. In some examples, the operation involves identifying an actuator 106 assigned to pipe section 12 and sending an activation instruction to the assigned actuator 106 via communication module 176. In one example, the identified disturbance is a leakage disturbance indicative of a leak in the piping system 10 near pipe section 12, and actuator 106 is assigned to pipe section 12 because it is located where it is designed to interrupt flow through pipe section 12 when activated.

[0067] In some examples, the operation involves transmitting a wake-up instruction to the second sensor 104. In one example, in some cases, identifying the location of a fault within the pipe system 10 based solely on the first sensor 102 may be inaccurate, and the first sensor 102 is designed to activate the second sensor 104 to work in conjunction with the second sensor 104 to identify a fault location.

[0068] In some examples, the operation involves transmitting a notification via the communication module 176 to the remote computing device 108. In one example, the first sensor 102 transmits a message to a user's mobile phone 108 containing information regarding the identified disturbance.

[0069] In some examples, the operation includes an additional operation by another device in the system 100. In some embodiments, the remote computing device 108 is configured to receive an instruction from a user to activate the actuator 106. In one example, the first sensor 102 transmits a notification of a fault to the remote computing device 108. A user operating the remote computing device 108 views the notification and issues an instruction to the remote computing device 108 to activate the actuator 106 and / or the second sensor 104.

[0070] In some examples, as described above, a fault is identified based on an assessment that an acoustic signal in the audio data from one or more of the transducer 130 and the microphone 164 is indicative of a fault. Unlike the temperature sensor 132 and humidity sensor 166, which indicate a fault based on whether the pipeline section 12 is operating within predetermined operating ranges, in some embodiments the transducer 130 and the microphone 164 do not indicate a fault simply based on whether a received audio signal is within a predetermined range. Instead, the processor 172 is designed to process audio signals received from the transducer 130 and / or the microphone 164 to infer the operating state of the pipeline section 12 and / or the piping system 10, and to identify a fault based on this inference.

[0071] In some examples, the processor 172 separates audio data for an audio signal received by the converter 130 and / or microphone 164 into segments and extracts features such as MFCCs from each segment. The processor 172 compares the extracted features with features corresponding to audio signatures of special events stored in memory 174 and, based on the comparison, determines that the audio signal is indicative of the special event.

[0072] In some examples, the processor 172 is designed to apply one or more machine learning algorithms to the audio signal to determine the operating characteristics of pipeline section 12 and / or piping system 10. In one example, the processor 172 uses the first machine learning algorithm to determine the operating characteristics of pipeline section 12 and / or piping system 10. The processor 172 determines whether the operating characteristics of pipeline section 12 and / or piping system 10 deviate by more than a predetermined threshold from historical information about the operating characteristics for pipeline section 12 and / or piping system 10, and identifies a fault based on this determination.

[0073] In some embodiments, the processor 172 is designed to apply one or more machine learning algorithms to the audio signal to detect the occurrence of a special event in the pipeline section 12 and / or pipe system 10. In one example, the processor 172 uses the second machine learning algorithm to determine that the audio signal is indicative of the occurrence of a special event. The processor 172 extracts features from the audio signal, compares the extracted features with features for special events stored in memory 174, and determines that a special event corresponds to the audio signal.

[0074] In some embodiments, the processor 172 is designed to apply the second machine learning algorithm to the audio signal and to determine that the audio signal is not indicative of a special event stored in memory. Fig. Figure 11 shows an exemplary embodiment of a process 1100 for using the system 100 in a flowchart. At block 1102, the first sensor 102 receives an audio signal from the pipeline section 12. At block 1104, the processor 172 determines that the audio signal is not indicative of a special event stored in memory 174. In other words, the processor 172 determines that an acoustic signal detected by at least one transducer 130 and microphone 164 does not match any of the classified acoustic signatures stored in memory 174.

[0075] At block 1106, processor 172 transmits a notification to remote computing device 108 indicating the detection of an unidentified acoustic signature. In some examples, notification 1102 includes audio data relating to the unidentified acoustic signature and / or operational characteristics of pipeline section 12 and / or pipe system 10 within a timeframe close to the detection point. At block 1108, system 100 receives an indication that the unknown acoustic signature originates from a source that does not affect the infrastructure health of pipe system 10. In various examples, the indication is received by a user via input device 136 of remote computing device 108 using the first machine learning algorithm or any other acceptable process.Examples of sources of an acoustic signal that do not affect the infrastructure health of pipe system 10 include a device or machine, ambient noise from traffic or construction work, or the like.

[0076] In some examples, the display includes a user-entered name for the unidentified acoustic signature, such as "washing machine" or similar, which identifies the source of the acoustic signature. At block 1110, the remote computing device 108 transmits the display to the first sensor 102. At block 1112, the processor 172 updates the memory 174 based on the display to identify the unidentified acoustic signature. In some embodiments, the processor 172 updates the second machine learning algorithm based on the display.

[0077] Fig.Figure 12 shows a flowchart of another exemplary process 1200 for using the system 100. At block 1202, the first sensor 102 receives an audio signal from the pipeline section 12. At block 1204, the processor 172 determines that the audio signal is indicative of a special event stored in memory 174, which corresponds to a fault in the pipe system 10.

[0078] In block 1206, processor 172 transmits a notification to remote computing device 108 indicating a fault detection. In some embodiments, the notification 1102 includes audio data relating to the unidentified acoustic signature and / or operational characteristics of pipeline section 12 and / or piping system 10 during a period close to the detection time. In some examples, the notification includes a request to the user to inspect a section of piping system 10, such as pipeline section 12, or another location associated with the detected fault.

[0079] In block 1208, the actuator 106 is activated in response to the detection to restrict the flow in pipeline section 20 and reduce the disturbance. In some embodiments, the activation is caused by an instruction transmitted from the remote computing device 108 in response to an instruction entered by the user. In some examples, the remote computing device 108 determines that the actuator 106 should be activated in response to the detected disturbance. In other examples, the processor 172 determines that the actuator 106 should be activated in response to the detected disturbance.

[0080] At block 1210, processor 172 updates memory 174 based on the detected disturbance. In some examples, processor 172 updates one or more of the first and second machine learning algorithms based on the detected disturbance.

[0081] In some embodiments, the processor 172 is configured to use the audio signals acquired by the microphone 164 as a noise reference for the audio signal acquired by the transducer 130. For example, in one embodiment, the processor 172 is configured to apply a noise reduction algorithm to an audio signal acquired by the transducer 130, with reference to another audio signal acquired by the microphone 164, in order to isolate noise emanating from the pipe section 12 from noise with an external source. Isolating noise emanating from the pipe section 12 reduces the risk of false positive leak detection caused by ambient noise. In some embodiments, the noise reduction algorithm includes removing audio signals acquired by the microphone 164 from audio signals simultaneously acquired by the transducer 130.

[0082] In some embodiments, the processor 172 applies the second machine learning algorithm to audio signals acquired via the microphone 164 to classify features of ambient noise in the vicinity of the pipe section 12. By subtracting different features or sets of features from the audio signal acquired via the transducer 130, the processor 172 is able to isolate noise emanating from the pipe section 12 even when the signals simultaneously acquired via the microphone 164 are not noise-suppression adapted.

[0083] In some of the embodiments described above, the processor 172 is integrated into the sensor 102. Such embodiments are an example of "edge computing," where computer operations are performed at the point of use rather than at a central location. In some embodiments, the processor 172 of a first sensor 102 is configured to operate the communication module 176 to communicate with processors located in other devices, such as the second sensor 104, so that the sensors 102 are configured to operate together as a distributed mesh network. In some embodiments, the system 100 additionally includes a cloud-based platform or a cloud-based node that aggregates data from and to multiple devices and performs further data analysis.

[0084] This disclosure is not limited to the features discussed in relation to any single embodiment.

[0085] In some designs, an acoustic transducer is attached to a tube via a built-in clamp.

[0086] In some embodiments, a circuit is designed to monitor the acoustic transducer and to initiate a wake-up sequence based on the monitoring of the acoustic transducer.

[0087] In some designs, a microphone is designed to capture an ambient audio signal.

[0088] In some embodiments, the ambient audio signal is used to cancel out non-pipe acoustic noises.

[0089] In some embodiments, the ambient audio signal is used to capture surrounding environmental noise for infrastructure monitoring.

[0090] In some embodiments, a temperature sensor is attached to a pipe via a built-in clamp and is designed to monitor the temperature of the pipe.

[0091] In some embodiments, a moisture sensor is used to detect moisture originating from unmonitored pipe sections, from outside the pipe system, or from other sources.

[0092] In some examples, a procedure for detecting leaks and / or deviations from a historical water flow or historical use in a pipe system involves the use of at least one machine learning algorithm.

[0093] In some examples, the procedure involves storing learned and / or pre-learned features for at least one machine learning algorithm.

[0094] In some examples, the procedure includes alerting a user in the event of a leak or malfunction.

[0095] In some examples, the procedure involves actuating an actuator in response to the detection of the leak or malfunction.

[0096] In some examples, the procedure involves storing historical information about the pipe system in a cloud-based system.

[0097] In some examples, the procedure involves performing a data analysis of the historical information and / or signals from devices attached to the pipe system, such as the acoustic transducer.

Claims

[1] Sensing device for monitoring a pipe system (10), the sensing device comprising: at least one sensor (102, 104) designed to detect at least one operating characteristic of a pipe section (12) of the pipe system (10), wherein the at least one sensor (102, 104) comprises a transducer (130) designed such that an audio signal emanating from the pipe section (12) causes the transducer (130) to generate a voltage signal indicative of the audio signal; a wake-up circuit (170) which is operatively connected to the converter (130) and is designed to generate a wake-up signal in response to the voltage signal generated by the converter (130) exceeding a predetermined threshold; and a processor (172) designed to identify an operating condition of the pipe system (10) with reference to the at least one operating characteristic sensed in response to the reception of the wake-up signal from at least one sensor (102, 104). a memory (174) that stores: first audio data indicative of the audio signal emanating from the pipeline section (12); and second audio data comprising an acoustic signature corresponding to a particular event occurring in at least one of the pipeline section (12) and an area (14) surrounding the pipeline section (12), wherein the particular event corresponds to an operating condition of the pipe system (10); wherein the processor (172) is designed to identify the operating condition of the pipe system (10) by comparing the audio signal from the first audio data and the acoustic signature of the second audio data, and to determine that the audio signal is indicative of the particular event; and wherein the at least one sensor (102, 104) further comprises a microphone (164) designed to detect a further audio signal emanating from an area (14) surrounding the pipe section (12); and The processor (172) is further designed to apply a noise reduction algorithm to the audio signal emanating from the pipe section (12) with reference to the further audio signal. [2] Sensing device according to claim 1, further comprising: a fastening mechanism (126, 128) designed for mounting the sensing device on the pipe section (12), wherein the transducer (130) is positioned on the mounting mechanism (126, 128) such that the transducer (130) is in direct contact with a surface of the pipe section (12) when the sensing device is mounted on the pipe section (12) via the mounting mechanism (126, 128). [3] Sensing device according to claim 2, wherein: which further includes at least one sensor (102, 104): a temperature sensor (132) designed to sense the temperature of the pipe section (12); and a moisture sensor (166) designed to detect moisture in an area (14) around the pipe section (12); and the temperature sensor (132) is positioned on the mounting mechanism (126, 128) such that the temperature sensor (132) is in direct contact with the surface of the pipe section (12) when the sensing device is mounted on the pipe section (12) via the mounting mechanism (126, 128). [4] Sensing device according to claim 1, further comprising: the memory (174) that stores: a first machine learning algorithm that can be executed by the processor (172) to determine a normal operating condition of the pipe system (10) over time with reference to at least one sensor (102, 104); and first data that are indicative of the normal operating condition of the pipe system (10); wherein the processor (172) is further designed to identify a fault in the pipe system (10) by determining that the identified operating condition deviates from the normal operating condition. [5] Sensing device according to claim 4, wherein the memory (174) further stores a second machine learning algorithm which can be executed by the processor (172) to identify a correspondence between a relevant operating condition of the pipe system (10) and a relevant particular event. [6] Sensing device according to claim 1, further comprising: a first communication module (176) connected to the processor (172), which can be operated for at least one sending and receiving of information relating to the operating condition of the pipe system (10). [7] System (100) for monitoring a pipe system, comprising: a sensing device according to claim 1, wherein the processor (172) of the sensing device is designed to: to operate the sensor (102, 104) selectively in a standby mode and in an active mode; when operating the sensor (102, 104) in standby mode, to transition the operation of the sensor (102, 104) to active mode in response to the wake-up signal; and When operating the sensor (102, 104) in active mode, the operating condition of the pipe system (10) is determined with reference to the at least one operating characteristic sensed by the at least one sensor (102, 104). [8] System (100) according to claim 7, further comprising: an actuator (106), comprising: a valve element (112) that can be operated to selectively restrict and allow the flow through a section (20) of the pipe system (10); and a second communication module (110), wherein: the information sent by a first communication module (176) includes an activation instruction; the second communication module (110) is designed to receive the activation instruction of the information sent by the first communication module (176) and to activate the valve element (112) in response to the received activation instruction. [9] System (100) according to claim 7 or 8, further comprising: a remote computing device (108), comprising: a third communication module (134) designed to at least one of sending information to and receiving information from the sensor (102, 104); an output device (138); and a further processor (130) designed to output information received from the third communication module (134) relating to the operating condition of the pipe system (10) via the output device (138). [10] System (100) according to claim 9, wherein: the further processor (130) is designed to send further information regarding the operating condition of the pipe system (10) to the sensor (102, 104); the storage unit (174) stores initial data corresponding to the operating conditions of the pipe system; and The processor (172) is designed to update the initial data based on the additional information received from the remote computing device (108). [11] System (100) according to claim 9 or 10, wherein the additional information sent by the remote computing device (108) includes an activation instruction; the second communication module (110) is designed to receive the activation instruction of the information sent by the remote computing device (108) and to activate the valve element (112) in response to the received activation instruction.

Images