Ultrasonic sensor monitoring device, system, and method

Abstract

A sensing assembly configured to monitor flow in a conduit, the sensing assembly comprising: an ultrasonic sensor assembly comprising an ultrasonic sensor, wherein the ultrasonic sensor assembly is configured to physically couple to a conduit, and wherein the ultrasonic sensor is configured to send and receive an ultrasonic signal through the conduit; and a controller electrically coupled to the ultrasonic sensor assembly, wherein the controller is configured to: determine a time of flight (ToF) of the ultrasonic signal, determine a flow rate in the conduit based on the ToF, store flow data comprising the determined flow rate, and identify an individual fixture associated with the flow data using a segmentation machine learning model. In some embodiments, the controller is further configured to identify a leak or failure of an individual fixture based on the flow data.

Description

ULTRASONIC SENSOR MONITORING DEVICE, SYSTEM, AND METHODCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to and the benefit of U.S. Provisional Application No. 63 / 527,279, filed July 17, 2023, and U.S. Provisional Application No. 63 / 527,137, filed July 17, 2023, the entire contents of which are hereby incorporated by reference.FIELD

[0002] The disclosure is directed to an ultrasonic monitoring device and system, in particular for monitoring, measuring and detecting fluid flow, for instance water flow in a building.BACKGROUND

[0003] Many users are currently unaware of their individual water-use habits and the status of the plumbing systems that they use at any given time. To conserve water and safeguard integrity of building structures, attempts have been made to develop systems and methods configured to monitor for and identify plumbing system leaks and to obtain water usage data corresponding to individual water-use fixtures. Detecting and providing data to users regarding their plumbing system leaks and water use efficiency of various water-use fixtures may allow users to take prompt action to repair leaks and to implement water conservation habits.SUMMARY

[0004] Desired are systems and methods configured to determine water use efficiency of various water-use fixtures (e.g. devices and appliances in a home, for instance toilets, faucets, showers, dishwashers, etc.). It would be desirable to have a sensing system which may be coupled to a single main incoming source water pipe for a residence or for a multi-residence building or office building to achieve this.

[0005] Residences and other buildings have incoming source water pipes having a wide variety of sizes and materials. Desired is a universal sensing system which may be easily and non-intrusively coupled to incoming water source pipes of different sizes and materials andconfigured to monitor for and identify a leak in a home or building. Also desired is a universal sensing system configured to identify different water-use devices and appliances and to determine their water use efficiency.

[0006] Accordingly, disclosed is a sensing assembly configured to monitor material flow in a conduit, the sensing assembly comprising an adjustment mechanism; a housing; and an ultrasonic sensor assembly, wherein the ultrasonic sensor assembly is at least partially positioned within the housing and comprises an ultrasonic sensor, the ultrasonic sensor is configured to couple a conduit, and the adjustment mechanism is configured to allow for coupling of the ultrasonic sensor to conduits having a variety of diameters. An adjustment mechanism may also allow for fine adjustment of sensor-to-conduit coupling.

[0007] The sensing assembly may comprise a general C-clamp type shape and require no tools for installation on a conduit. A housing and adjustment mechanism may be configured to properly align a conduit to the assembly, such that optimal ultrasonic signals may be obtained from the ultrasonic sensor. The ultrasonic sensor may comprise a first ultrasonic transducer coupled to a first upstream wedge and a second ultrasonic transducer coupled to a second downstream wedge, wherein the first wedge and the second wedge are configured to be in direct physical contact with a conduit through a wedge lateral contact surface. A wedge may comprise an engineering thermoplastic. Also subject of the disclosure are a wedge and an ultrasonic sensor as described herein.

[0008] A sensing assembly may be configured to recognize and report a system leak or malfunction. A sensing assembly may be configured to recognize and report use of individual water-use devices or appliances, for example a toilet, a shower, a faucet, a dishwasher, a washing machine, etc. A sensing assembly may be configured to determine and report a water leak or failure of a specific (individual) water-use device or appliance. A sensing assembly may be configured to determine and report water use efficiency of individual water-use device or appliance.

[0009] A sensing assembly may be configured to monitor material flow in a conduit, the sensing assembly comprising an adjustment mechanism; a controller, a housing; and an ultrasonic sensor assembly, wherein the ultrasonic sensor assembly is at least partially positioned within the housing and comprises an ultrasonic sensor, the ultrasonic sensorassembly is electrically coupled to the controller, the ultrasonic sensor is configured to couple a conduit, the ultrasonic sensor is configured to send and receive an ultrasonic signal through the conduit, the controller is configured to determine a time-of-flight (ToF) of the ultrasonic signal, and the controller is configured to determine a flow rate in the conduit based on the ToF.

[0010] In some embodiments, a sensing assembly configured to monitor flow in a conduit is provided, the sensing assembly comprising: an ultrasonic sensor assembly comprising an ultrasonic sensor, wherein the ultrasonic sensor assembly is configured to physically couple to the conduit, and wherein the ultrasonic sensor is configured to send and receive an ultrasonic signal through the conduit; and a controller electrically coupled to the ultrasonic sensor assembly, wherein the controller is configured to: determine a time of flight (ToF) of the ultrasonic signal, determine a flow rate in the conduit based on the ToF, store flow data comprising the determined flow rate, and identify an individual fixture associated with the flow data using a segmentation machine learning model.

[0011] In some embodiments, identifying an individual fixture associated with the flow data comprises: sampling the flow data, and separating the sampled data into a plurality of time duration segments, wherein a first time duration segment of the plurality of time duration segments is characterized by not having changes in flow rate, and wherein a second time duration segment of the plurality of time duration segments is characterized by having changes in flow rate.

[0012] In some embodiments, identifying an individual fixture associated with the flow data comprises analyzing one or more of a duration, volume accumulation, direction, starting flow rate, and ending flow rate in one or more of the plurality of time duration segments.

[0013] In some embodiments, identifying an individual fixture associated with the flow data comprises applying one or more rules to the time duration segments using a state machine.

[0014] In some embodiments, determining the flow rate comprises applying a temperature compensation data processing operation to the ToF of the ultrasonic signal.

[0015] In some embodiments, the temperature compensation data processing operation is based on a set of no-flow rate determinations at various temperatures.

[0016] In some embodiments, the flow data comprises one or more of volume data, flow rate moving average data, and a time stamp of the ultrasonic signal.

[0017] In some embodiments, identifying an individual fixture associated with the flow data comprises determining a no-flow rate based on the flow data.

[0018] In some embodiments, the controller is configured to calibrate the ultrasonic sensor assembly using the no-flow rate.

[0019] In some embodiments, the controller is configured to identify a leak or failure of the individual fixture based on the flow data and the no-flow rate.

[0020] In some embodiments, identifying a leak or failure of the individual fixture by the controller comprises identifying a volume above a defined threshold from the volume data.

[0021] In some embodiments, the controller is configured to communicate with a cloud server, and wherein the cloud server is configured to identify a leak or failure of the individual fixture based on the flow data and the no-flow rate.

[0022] In some embodiments, identifying the leak or failure associated with an individual fixture comprises transmitting the volume data and flow rate moving average data to the cloud server and receiving a water use trend analysis from the cloud server.

[0023] In some embodiments, the controller is configured to determine a speed of sound in the conduit based on the ultrasonic signal.

[0024] In some embodiments, the controller is configured to transmit the flow data to the cloud server in accordance with a publishing protocol.

[0025] In some embodiments, the publishing protocol comprises transmitting the flow data upon identifying the volume above the defined threshold.

[0026] In some embodiments, the publishing protocol comprises publishing a flow data sample having an immediately preceding time stamp upon identifying a change in flow rate above a defined threshold.

[0027] In some embodiments, the publishing protocol comprises transmitting flow data from the controller to the cloud server at variable time periods.

[0028] In some embodiments, the publishing protocol accounts for a signal-to-noise ratio(SNR).

[0029] In some embodiments, the sensing assembly is configured to send and receive ultrasonic signals on a millisecond time scale.

[0030] In some embodiments, the sensing assembly is configured to send and receive ultrasonic signals until a moving average catches up to a flow rate in the flow data.

[0031] In some embodiments, the sensing assembly further comprises an LED light indicator configured to indicate a strength of the ultrasonic signal.

[0032] In some embodiments, the ultrasonic sensor assembly is further configured to send a test signal through the conduit and receive a response to the test signal comprising one or more test outputs.

[0033] In some embodiments, the controller is configured to determine a material of the conduit based on the one or more test outputs.

[0034] In some embodiments, the controller is configured to determine a diameter of the conduit based on the one or more test outputs.

[0035] In some embodiments, a sensing assembly configured to monitor flow in a conduit is provided, the sensing assembly comprising an ultrasonic sensor assembly comprising an ultrasonic sensor, wherein the ultrasonic sensor assembly is configured to physically couple to the conduit, and wherein the ultrasonic sensor is configured to send and receive an ultrasonic signal through the conduit; and a controller electrically coupled to the ultrasonic sensor assembly, wherein the controller is configured to: determine a time-of-flight (ToF) of the ultrasonic signal, determine a flow rate and a no-flow rate in the conduit based on the ToF, store flow data comprising the determined flow rate and a no-flow rate, calibrate the ultrasonic sensor using the no-flow rate, and identify a leak or failure of an individual fixture based on the flow data.BRIEF DESCRIPTION OF THE DRAWINGS

[0036] This disclosure is illustrated by way of example and not by way of limitation in the accompanying figures. For simplicity and clarity of illustration, features illustrated in thefigures are not necessarily drawn to scale. For example, the dimensions of some features may be exaggerated relative to other features for clarity. Further, where considered appropriate, reference labels have been repeated among the figures to indicate corresponding or analogous elements.

[0037] FIGS. 1 A-l J show various views of a sensing assembly and sensing assembly components, according to some embodiments.

[0038] FIG. IK provides views of sensing assembly components, according to some embodiments.

[0039] FIG. 2A provides a representation of a segment analysis, according to some embodiments.

[0040] FIG. 2B shows a data processing pipeline, according to some embodiments.

[0041] FIG. 2C illustrates a state machine, according to some embodiments.

[0042] FIG. 3 A shows a rejected time segment flow rate distribution towards a temperature compensation determination, according to some embodiments.

[0043] FIG. 3B shows an accepted time segment flow rate distribution towards a temperature compensation determination, according to some embodiments.

[0044] FIG. 3C provides a graph of a regression analysis to be applied for temperature compensation, according to some embodiments.

[0045] FIG. 4A shows a graph of flow rate according to frequency of a sample, according to some embodiments.

[0046] FIG. 4B shows a graph having flow rate data taken at fixed time interval samples together with moving average data, according to some embodiments.

[0047] FIG. 5A shows a graph of flow rate distribution, according to some embodiments.

[0048] FIG. 6A provides a view of a user dashboard, according to some embodiments.

[0049] FIG. 6B shows a portion of a user dashboard, according to some embodiments.

[0050] FIG. 7A and FIG. 7B provide views of a user interface notification main-page, and notification sub-page, respectively, according to some embodiments.

[0051] FIGS. 8 A- 8D show views of a user interface interactive water level indicator for consumption comparison, according to some embodiments.

[0052] FIG. 9 illustrates the detection of a zero-flow rate, according to some embodiments.

[0053] FIG. 10 illustrates a computer, according to some embodiments.DETAILED DESCRIPTION

[0054] In some embodiments, a sensing assembly is provided that may be configured to monitor material flow, e.g. a water flow rate, in a conduit such as a plumbing fixture. The sensing assembly may include an ultrasonic sensor assembly having an ultrasonic sensor. The ultrasonic sensor assembly may be configured to physically couple to conduits, e.g. pipes, having various shapes and sizes. In some embodiments, the ultrasonic sensor may be configured to send and receive an ultrasonic signal through the conduit. The sensing assembly may also include a controller electrically coupled to the ultrasonic sensor assembly. In some embodiments, the controller may be configured to determine a time-of-flight (ToF) of the ultrasonic signal and determine a flow rate (e.g. in gallons per minute) in the conduit based on the ToF. In some embodiments, the controller may store water flow data comprising the determined flow rate and identify an individual water-use fixture associated with the water flow data. The controller may also transmit water flow data to a cloud server for further analysis and storage. The sensing assembly and / or the cloud server may be configured to provide the water flow data of the individual water-use fixture to the user, so that the user can determine which fixtures are using the most water and adjust their habits accordingly.

[0055] In some embodiments, the controller may be configured to determine a no-flow rate (e.g. a flow rate corresponding to zero water flow) instead of, or in addition to, determining a flow rate. Since the ultrasonic sensor may measure a change in temperature as a change in flow rate, determining a no-flow rate (e.g. a true “zero”) that removes the contribution of temperature can allow for more precise detection of smaller leaks. The controller may be configured to calibrate the ultrasonic sensor using the no-flow rate to make the detection of leaks more precise and accurate. The controller may then be able to analyze water flow data to determine that a small or large leak in the conduit and / or the individualwater use fixture is present. The controller may also be configured to transmit the no-flow rate / water flow data to a cloud server to perform the detection of a leak. The sensing assembly and / or the cloud server may be configured report the presence of the leak to a user.

[0056] In the following description of the various embodiments, it is to be understood that the singular forms “a,” “an,” and “the” used in the following description are intended to include the plural forms as well, unless the context clearly indicates otherwise. It is also to be understood that the term “and / or” as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items. It is further to be understood that the terms “includes, “including,” “comprises,” and / or “comprising,” when used herein, specify the presence of stated features, integers, steps, operations, elements, components, and / or units but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, units, and / or groups thereof.Sensing Assembly

[0057] FIG. 1 A provides a view of sensing assembly 100, according to some embodiments. Sensing assembly 100 comprises housing 101 having ultrasonic sensor assembly 104 positioned therein. Ultrasonic sensor assembly 104 comprises wedges 104w positioned in wedge holder 119. Housing 101 comprises electronic port 118, configured to receive a plug / electric wire to provide electric power to components including a controller (not visible), LED indicator (not visible), and ultrasonic sensor assembly 104. Sensing assembly 100 comprises cradle 115, shown separately in FIG. IB. Cradle 115 is configured to receive and hold conduits of various diameters in physical contact with wedges 104w. Sensing assembly 100 also comprises light pipe 135, which encircles an entire perimeter of an upper edge of housing 101 (edge about the housing interior). Assembly 100 comprises a compact shape and size, allowing it to be coupled to conduits of various diameters in a variety of tight spaces. Light pipe 135 is configured to be visible from essentially any angle, indicating if / when assembly 100 is / has been properly installed on a conduit. Proper installation may be when a conduit is positioned on wedges 104w so that an optimal ultrasonic signal is received by a controller.

[0058] FIG. 1C shows a side view of sensing assembly 100, and FIG. ID shows a crosssection view of sensing assembly 100, according to some embodiments. Conduit CDT iscoupled to ultrasonic sensor assembly 104 in housing 101 and is positioned between housing 101 and cradle 115. Cradle 115 comprises well 115w configured to receive conduit CDT and housing 101 comprises well 101 w also configured to receive conduit CDT. Cradle 115 is configured to be adjustably moved via screw 102 and knob 102k relative to housing 101, such that cradle well 115w and housing well 101 w may be moved along a substantially straight line to and away from (relative to) each other to accommodate conduits of varying diameters. In some embodiments, the knob 102k and screw 102 may together be referred to as an “adjustment mechanism.” An interior of housing 101 comprises rectangular box-shaped opening lOlh, configured to hold ultrasonic sensor assembly 104.

[0059] FIG. IE and FIG. IF provide further views of sensing assembly 100, according to some embodiments. Various dimensions are shown, for example in the position shown, housing 101 and knob 102k together comprise a longest dimension of about 100 mm, a width of about 85 mm, and a depth of about 70 mm. Accordingly, housing 101, including knob 102k, may have a total volume of about 600 cm3. It is seen that cradle well 115w and housing well 101 w are substantially aligned, that is, each having low points substantially positioned on a same line and configured to be adjustably moved relative to each other along a line. FIG. 1G shows sensing assembly 100 coupled to conduits CDT having outer diameters (OD) of about 23 mm on the left, and about 51 mm on the right. Sensing assembly 100 is configured to receive and couple to conduits having minimum OD of about 15 mm (about 0.6 inches) to maximum OD of about 51 mm (about 2.0 inches). In sensing assembly 100, adjustment mechanism 102 / 102k is configured to “pull” cradle 115 towards ultrasonic sensor assembly 104, rather than “push” it towards a ultrasonic sensor assembly.

[0060] FIG. 1H, FIG. II, and FIG. 1 J provide various views of sensing assembly 100, wherein light pipe 135 is visible. Also shown are ultrasonic sensor assembly 104, comprising wedges 104w positioned in wedge holder 119. Also shown is wire portion 118w plugged into electric port 118. Light pipe 135 is a separate part, configured to be positioned about an entire perimeter of an edge of housing 101. Light pipe 135 also functions as a lid positioned over ultrasonic sensor assembly 104. LED lights (not shown) may be mounted on a controller (not shown) positioned at an interior of housing 101. LED lights may be configured to, for example, flash yellow as assembly 100 is positioned on a conduit. As one adjusts assembly 102 / 102k to optimize a conduit position on ultrasonic sensor assembly 104between cradle 115 and housing 101, and a good or optimal ultrasonic signal is detected by the controller, the one or more LED lights may be configured to emit a solid color, for example green. LED lights may be configured to emit light towards light pipe 135, and light may be configured to illuminate an entire or about an entire volume of light pipe 135. Light pipe 135 may be solid or have a hollow volume and may be translucent. Light pipe 135 is configured to provide a view of LED lights from about any view or angle. In an embodiment, light pipe 135 may comprise a thermoplastic, for example an engineering thermoplastic, for instance polycarbonate.

[0061] FIG. IK provides views of ultrasonic sensor assembly 104 and of wedge 104w, according to some embodiments. The top illustration shows ultrasonic sensor assembly 104, having wedges 104w positioned in wedge holder 119. The center illustration shows an exploded view of ultrasonic sensor assembly 104. Wedges have curved contact surface 104c, configured to provide for high surface area contact with a conduit. Wedge holder 119 comprises brackets 119b, configured to receive each wedge 104w. Wedge holder 119 also comprises two pairs of protrusions 119t (one of each pair is visible), configured to couple to blind holes 104h positioned on each side of each wedge 104w. Protrusion 119t / hole 104h couplings are configured to indicate proper attachment of wedges 104w to holder 119 and to allow for a degree of rotational movement of wedges 104w during coupling to a conduit. A degree of rotational movement is configured to compensate for any conduit irregular shapes and to provide for good contact of wedges 104w with a conduit.

[0062] FIG. IK bottom illustrations show perspective views of wedge 104w. Visible is angled surface 104a configured to receive an ultrasonic transducer, opposing angled surface 104b, blind hole 104h, and curved contact surface 104c. Upper flat surface 104u is configured to be seated in brackets 119b of holder 119. Wedge surface 104a comprises extending edge 104p configured to receive an edge of an ultrasonic sensor. Wedge 104w comprises a length LI of about 24.5 mm, a largest width W1 of about 12 mm, a largest height of about 11 mm, and curved contact surface 104c width W2 of about 10 mm. Edge 104p extends about 0.5 mm out from surface 104a. In an embodiment, contact surface 104c may comprise a “rubberized” or elastomeric coating configured to act as a surface contact promoter between surface 104c and a conduit. A surface contact promoter may improvesurface-to-surface contact of surface 104c and a conduit surface and may improve ultrasonic signal strength and quality. For instance, any air gaps in a contact area may be eliminated.

[0063] A thermoplastic may include, for example, an engineering thermoplastic. In some embodiments, a wedge may comprise a thermoplastic selected from one or more of a polyamide, a polyester, a polycarbonate, a polyacetal, acrylonitrile-butadiene-styrene, poly(methyl (meth)acrylate), a polyetheretherketone, a polyetherketoneketone, a polyketone, a polyphenylene sulfide, a polyphenylene oxide, a polysulfone, or polytetrafluoroethylene. Polysulfones include for example one or more of poly(arylene sulfone), poly(bisphenol-A sulfone), polyether sulfone, polyphenhylenesulfone, polysulfone, or VICTREX HTA. Reference to “poly sulfone” may mean one or more of the above-listed poly sulfones.

[0064] In some embodiments, a thermoplastic chosen for fabrication of a wedge may have a speed of sound and / or an acoustic impedance about that of water. This may reduce undesirable effects of refraction of ultrasonic waves as they pass through different mediums. In some embodiments, fabrication of a wedge may be accomplished by molding techniques, for example injection molding. In some embodiments, fabrication of a wedge may be accomplished with 3D printing techniques.

[0065] An ultrasonic sensor assembly may comprise a first ultrasonic transducer coupled to a first upstream wedge, and a second ultrasonic transducer coupled to a second downstream wedge. A first wedge may be positioned on a conduit upstream portion, and a second wedge may be positioned on a conduit downstream portion. The upstream / downstream orientation of the wedges may be reversed as necessary for positioning of a sensing assembly on a material conduit. The first and second wedge assemblies may be symmetrical and interchangeable. When coupled to a conduit, the first and second wedges may be mirror images of each other. The wedges may be in direct physical contact with a material conduit at a wedge lateral contact surface.

[0066] An ultrasonic transducer may be coupled to a first upstream wedge upstream angled surface, and a second ultrasonic transducer may be coupled to a second downstream wedge downstream angled surface as shown in the figures. The wedge angled surfaces may be configured to optimize an angle at which an ultrasonic signal or wave is directed through a conduit and to the other ultrasonic transducer. Likewise, the opposing angled surfaces that donot have the ultrasonic transducers coupled thereto may also be configured to optimize an angle at which an ultrasonic signal or wave is directed through a conduit and to the other ultrasonic transducer.

[0067] In some embodiments, a sensing assembly may comprise a surface contact promoter configured to be in contact with the first wedge, with the second wedge, and with the conduit. A surface contact promoter is configured to increase surface contact between a first wedge, a second wedge, and a conduit. In some embodiments, a surface contact promoter may comprise an elastomeric strip, for example a silicon strip. In some embodiments, a surface contact promoter may comprise a gel. A strip or a gel may comprise an adhesive. In some embodiments, a surface contact promoter may be in direct contact with each of a first and a second wedge and a conduit. In other embodiments, a surface contact promoter may comprise an elastomeric coating on a wedge contact surface. In some embodiments, a wedge having a surface contact promoter may still be considered to be in direct physical contact with a conduit.

[0068] An ultrasonic transducer may be a piezo transducer. Ultrasonic transducers may be in wired or wireless electronic communication with a controller (e.g. a microcontroller or microprocessor) and a power source. The ultrasonic transducers may be in wired or wireless communication with an analog-to-digital converter. In some embodiments, the ultrasonic transducers send and receive, on an alternating basis, ultrasonic waves through a first wedge, through a conduit having a material therein, and through a second wedge. The sonic data may be converted to an electrical signal, which is sent to an analog-to-digital converter. An analog-to-digital converter may convert an electrical signal to digital data. A controller may be configured to receive digital data from an analog-to-digital converter.

[0069] A controller may be configured to determine a time-of-flight (ToF) from sensed data. Based on a ToF measurement, a controller is configured to determine a flow rate from an algorithm. A controller may be configured to evaluate digital flow rate data to make determinations regarding water use of individual appliances or fixtures. A controller may comprise an algorithm configured to recognize and identify water use devices and appliances by their water use “sound signature”, which is reflected in the digital flow rate data. Water use devices and appliances may include one or more of a toilet, a urinal, a faucet, a dishwasher, a washing machine, a shower, a tub spout, an icemaker, a hot water heater, andthe like, including any features having a water valve, for example a sprinkler system or an outdoor hose.

[0070] A controller may be configured to recognize a temperature contribution to the flow rate data. For example, a sensing assembly may comprise one or more temperature sensors in electronic communication with a controller. A controller may comprise an algorithm configured to perform a temperature compensation data processing operation. In some embodiments, the temperature may be continuously monitored by a controller. In some embodiments, a temperature may be determined from monitoring a relative speed of sonic waves through a material.

[0071] A controller may be configured to recognize and determine a normal and abnormal state of a plumbing system. For example, a plumbing system normal state recognition may include determination of “no leak”. A plumbing system abnormal state recognition may include determination of a water leak. A water leak may be a general leak somewhere in a conduit, and / or may be a leak associated with a certain water-use device or appliance. A leak might include a running toilet because a flapper is not properly seated on a valve. In some embodiments, a sensing assembly may be configured to recognize and determine a general leak and / or a leak associated with a specific (individual) water-use device or appliance.

[0072] A controller may be configured to report a normal or abnormal state of a plumbing system. For example, a computing device, such as a smartphone or a laptop computer, may be linked to a controller. A controller may be configured to transmit data to a computing device. A computing device may have a graphical user interface configured to display data, including water-use device and appliance water use data, water efficiency data, efficiency data over time, etc.

[0073] In some embodiments, a controller may be in remote communication with a server, e.g. a cloud server. The controller may be configured to transmit data to / from the cloud server in performing one or more of the functions described herein. In some embodiments, one or more of the fixture detection, leak detection, temperature compensation data processing operations, zeroing functions, and other functions described here and infurther detail below may be performed by a server instead of, or in addition to, the controller performing one or more of these functions.

[0074] In some embodiments, a material conduit may be a water pipe, for example, a main water-source pipe for a residence or a main water-source pipe for a portion of a large apartment building or office building. In some embodiments, a sensing assembly may be configured to be positioned on a conduit with an adjustment mechanism or device, for example a clamp device, a screw, a set screw, a spring, and the like. A sensing assembly may be coupled to conduits of different sizes, for example having diameters of from about 0.5 inches to about 1.5 inches. In some embodiments “diameter” may refer to inner or outer conduit diameter. A sensing assembly may be configured for use on conduits of different materials, for example copper, iron, steel, PVC, or PEX (cross-linked polyethylene).

[0075] In some embodiments, a sensing assembly may be coupled to both an incoming cold water pipe and an incoming hot water pipe. For example, an incoming hot water pipe may be downstream of a hot water heater. In other embodiments, a material conduit may be a gas pipe, for instance a conduit used to carry natural gas. In some embodiments, a material conduit may be a pipe configured to transport crude oil, refine oil, or gasoline. Conduits may include pipes configured to carry or transport a material including one or more of a gas, a liquid (free-flowing or a viscous liquid), a particulate, a liquid suspension, a liquid solution, etc. In some embodiments, a material may include refrigerants, for instance refrigerants employed in HVAC systems.

[0076] In some embodiments, a controller may be electrically coupled to an indicator or display configured to indicate if a sensing assembly in positioned on a conduit in a suitable position to receive sensed information. For example, a controller may be in electronic communication with an LED light indicator that may include one or more LED lights that are built into or attached to the housing of the sensing assembly. In some embodiments, the LED light indicator may be configured to indicate a signal quality as the user installs the sensing assembly. This may allow the sensing assembly to provide the user with feedback during installation of the sensing assembly, without the user having to install the user interface on their mobile device prior to installing the physical sensing assembly to obtain such feedback.

[0077] For example, an LED light indicator may show a first color when no signal is received, a second color when a weak signal is received, and a third color when a strong or optimal signal is received. The user may use this color change as a guide to know when the sensing assembly has been properly positioned during installation. In some embodiments, rather than a change in color, the LED light indicator may be configured to blink, flash, dim, or fade in and out to reflect the quality of signal received. For example, receiving a low- quality signal having a low amplitude received during installation may cause the LED light indicator to dim or fade out to a lower brightness, or fade in an out at an increased rate, which may prompt the user to place the sensing assembly in a more optimal position on the pipe or conduit. Conversely, placing the sensing assembly such that it receives a stronger, higher quality signal may cause the LED light indicator to increase a brightness or cause an LED light to stay on as opposed to fading in and out.

[0078] In some embodiments, the controller may determine a conduit diameter or material in determining a flow rate or other information. A conduit diameter may be measured in any suitable manner (e.g. with a “ruler”) and the data may be inputted to the controller. In some embodiments, a sensing assembly may comprise a proximity sensor configured to determine a position of an adjustment mechanism. A proximity sensor, electrically coupled to the controller, can be configured to relay an adjustment mechanism position to the controller, and the controller can be configured to determine a conduit diameter from the sensed position. Proximity sensors may comprise one or more inductive, capacitive, optical (infrared (IR), photoelectric), magnetic, ultrasonic, or rotary sensors.

[0079] One challenge with clamp-on ultrasonic flow meters is providing accurate and reliable measurements across a wide variety of conduit (e.g. pipe) shapes, sizes, and materials. Existing flow meters may limit the range of compatible pipe conditions and / or may require the user to provide the exact pipe information to calibrate their measurement device. A sensing assembly as described herein may advantageously be calibrated to suit a variety of conduit shapes and sizes without requiring user input, although user input may still be optionally provided. In some embodiments, a sensing assembly may be configured to automatically detect conduit properties and may calibrate one or more parameters of the ultrasonic sensors of the sensing assembly based on the detected conduit properties.

[0080] To detect the properties of a conduit, such as its diameter or material composition, the sensing assembly may be configured to perform different test schedules upon installation by sending test signals through the conduit comprising inputs that may affect how a transducer of the sensing assembly behaves and perceives data in conduits of different sizes. These inputs may include, for example, a frequency of sending the ultrasonic signal, a number of pulses, an amplitude of the signal, or a noise mask of the signal. Exemplary outputs of the received signals may include a time-of-flight, an amplitude of the signal, a frequency of the signal, or a phase shift of the signal. Other related inputs / outputs that may vary in conduits of different sizes and materials may be measured during conduit calibration.

[0081] The controller of the sensing assembly may be configured to match the various outputs from the sensing assembly to a validated library of outputs from sample conduits having different sizes or material properties. The validated library may include, for example, sets of validated outputs associated with sample conduits having different material properties, with the validated outputs corresponding to each of the test schedule inputs. The controller may determine that the sensing assembly is attached to a conduit having a particular property or set of properties based on the closest match between the received outputs and a sample conduit from the sample output library. For example, the controller may receive a measured time-of-flight value from the sensing assembly in response to a particular test schedule and may determine that the measured value has a best match to a validated output corresponding to a copper pipe having a particular size. The controller may be configured to calibrate the sensing assembly based on the properties of the conduit having the best match to the observed outputs.

[0082] In some embodiments, the controller may be configured to determine a similarity percentage between a data value associated with a particular conduit in the known library and the measured outputs received from the sensing assembly in response to the test schedule. In some embodiments, the controller may determine a “match” as being the conduit in the known library having the highest similarity percentage or score relative to the measure data. In some embodiments, the controller may determine a “match” if the similarity percentage is within a preconfigured threshold value. In some embodiments, multiple test schedules may be performed by the sensing assembly, and the closest match may be identified by the conduit in the known library having a highest average similarity percentage. By calibrating the sensingassembly based on the particular properties of the conduit, the accuracy of the flow rate determination, zero determination, temperature contribution, and other measurements to be described in more detail may be improved.

[0083] In some embodiments, a sensing system may be configured to be easily coupled to a conduit in a variety of orientations without the use of any tool. In some embodiments, a sensing system may be “one piece”, meaning the part positioned on the conduit may be a single part that may simply be coupled to a conduit. In some embodiments, a sensing system may comprise a housing and an adjustment mechanism. An ultrasonic sensor may be positioned in a housing. A housing may have an opening through which a conduit may be coupled to an ultrasonic sensor. A housing opening may have angled or curved edges adjacent to the opening, configured for the assembly to receive conduits having different diameters.

[0084] In some embodiments, a first ultrasonic transducer may be coupled to an upstream surface of a first upstream wedge, and a second ultrasonic transducer may be coupled to a downstream surface of a second downstream wedge. In some embodiments, an ultrasonic transducer may be coupled to an angled surface of a wedge. An angled surface may be defined by an angle between a wedge lateral contact surface and the angled surface. In some embodiments, this angle may be between about 40° to about 78°. In some embodiments this angle may be greater than or equal to about 40°, about 42°, about 44°, about 46°, about 48°, about 50°, about 52°, about 54°, or about 56°, to any of about 58°, about 60°, about 62°, about 64°, about 66°, about 68°, about 70°, about 72°, about 74°, or about 76°. In some embodiments, this angle may be less than or equal to about 42°, about 44°, about 46°, about 48°, about 50°, about 52°, about 54°, or about 56°, to any of about 58°, about 60°, about 62°, about 64°, about 66°, about 68°, about 70°, about 72°, about 74°, about 76°, or about 78°.

[0085] In some embodiments, a wedge may comprise a length of about 10 mm to about 46 mm. In some embodiments, a wedge may comprise a length of greater than or equal to about 10 mm, about 12 mm, about 14 mm, about 16 mm, or about 18 mm, to any of about 20 mm, about 22 mm, about 23 mm, about 24 mm, about 25 mm, about 26 mm, about 27 mm, about 28 mm, about 30 mm, about 32 mm, about 34 mm, about 36 mm, about 40 mm, about 42 mm, about 44 mm. In some embodiments, a wedge may comprise a length of less than or equal to about 12 mm, about 14 mm, about 16 mm, or about 18 mm, to any of about 20 mm,about 22 mm, about 23 mm, about 24 mm, about 25 mm, about 26 mm, about 27 mm, about 28 mm, about 30 mm, about 32 mm, about 34 mm, about 36 mm, about 40 mm, about 42 mm, about 44 mm, or about 46 mm.

[0086] In some embodiments, a wedge may comprise a height of about 6 mm to about 30 mm. In some embodiments, a wedge may comprise a height of greater than or equal to about 6 mm, about 8 mm, about 10 mm, about 11 mm, about 12 mm, about 13 mm, about 14 mm, about 15 mm, about 17 mm, about 19 mm, about 21 mm, about 23 mm, about 25 mm, about 27 mm, or about 29 mm. In some embodiments, a wedge may comprise a height of less than or equal to about 11 mm, about 12 mm, about 13 mm, about 14 mm, about 15 mm, about 17 mm, about 19 mm, about 21 mm, about 23 mm, about 25 mm, about 27 mm, about 29 mm, or about 30 mm.

[0087] In some embodiments, a wedge may comprise a curved lateral contact surface, configured to provide for contact with conduits of various sizes and / or diameters. A curved lateral surface may be configured to be in direct physical contact with a conduit.Alternatively, as discussed above, a curved lateral surface may be in direct physical contact with a surface contact promoter, and the surface contact promoter may be in direct physical contact with a conduit. In some embodiments, a lateral contact surface may comprise a width of about 4.0 mm to about 17 mm. In some embodiments, a lateral contact surface may comprise a width of greater than or equal to about 4.0 mm, about 5.0 mm, about 6.0 mm, about 7.0 mm, or about 8.0 mm, to any of about 9.0 mm, about 10.0 mm, about 11.0 mm, about 12.0 mm, about 13.0 mm, about 14.0 mm, about 15.0 mm, or about 16.0 mm. In some embodiments, a lateral contact surface may comprise a width of less than or equal to about 5.0 mm, about 6.0 mm, about 7.0 mm, or about 8.0 mm, to any of about 9.0 mm, about 10.0 mm, about 11.0 mm, about 12.0 mm, about 13.0 mm, about 14.0 mm, about 15.0 mm, about 16.0 mm, or about 17.00 mm.

[0088] In some embodiments, a wedge contact surface may have a lateral trough positioned on either side. A wedge contact surface and the lateral troughs may be substantially parallel with each other. A lateral trough positioned alongside a contact surface may help prevent an ultrasonic sensor from detecting ultrasonic waves that are not aligned with a conduit centerline. In some embodiments, a lateral trough may comprise a width of about 0.3 mm to about 3 mm. In some embodiments a lateral trough may comprise a width ofless than or equal to about 0.5 mm, about 0.7 mm, about 0.9 mm, about 1.1 mm, about 1.3 mm, about 1.5 mm, about 1.7 mm, about 1.9 mm, about 2.1 mm, about 2.3 mm, about 2.5 mm, about 2.7 mm, about 2.9 mm, or about 3 mm. In some embodiments, a lateral trough may comprise a width of greater than or equal to about 0.3 mm, about 0.5 mm, about 0.7 mm, about 0.9 mm, about 1.1 mm, about 1.3 mm, about 1.5 mm, about 1.7 mm, about 1.9 mm, about 2.1 mm, about 2.3 mm, about 2.5 mm, about 2.7 mm, or about 2.9 mm.

[0089] In some embodiments, a lateral trough may comprise a depth of from about 0.4 mm to about 3.2 mm. In some embodiments, a lateral trough may comprise a depth of greater than or equal to about 0.4mm, about 0.6 mm, about 0.8 mm, or about 1.0 mm, about 1.2 mm, about 1.4 mm, about 1.6 mm, about 1.8 mm, about 2.0 mm, about 2.2 mm, about 2.4 mm, about 2.6 mm, about 2.8 mm, or about 3.0 mm. In some embodiments, a lateral trough may comprise a depth of less than or equal to about 0.6 mm, about 0.8 mm, or about 1.0 mm, about 1.2 mm, about 1.4 mm, about 1.6 mm, about 1.8 mm, about 2.0 mm, about 2.2 mm, about 2.4 mm, about 2.6 mm, about 2.8 mm, about 3.0 mm, or about 3.2 mm.

[0090] In some embodiments a first wedge may comprise a downstream angled surface and a second wedge may comprise an upstream angled surface, and the angled surfaces may be substantially mirror images of each other. In some embodiments, a wedge may comprise a channel or trough positioned on an upper portion thereof. A channel positioned on an upper portion may be substantially perpendicular to a lateral contact surface and lateral troughs adjacent the lateral contact surface. Such a channel may be configured to provide for adjustment or movement of a ultrasonic sensor assembly to aid in alignment with a conduit. In some embodiments, an ultrasonic transducer may be directly bonded to a wedge surface. This may be accomplished with an adhesive, ultrasonic welding, etc. Adhesives may include epoxy, acrylic, etc.

[0091] In some implementations, an adjustment mechanism may be configured to adjust, or adjustably move, a cradle towards and away from an ultrasonic sensor assembly positioned in a housing interior. An adjustable movement of a cradle towards and away from an ultrasonic sensor assembly may place a sensing assembly in an optimal position on a conduit. When a sensing assembly is properly positioned on a conduit, a cradle may aid in holding and maintaining the proper position. In some embodiments, an adjustment mechanism may be configured to “push” a cradle configured to receive a conduit towards a housing andultrasonic sensor assembly, or alternatively, may be configured to “pull” a cradle towards a housing and ultrasonic sensor assembly. The adjustment mechanism may be automated or motorized, for example, to automatically adjust tightness of the knob and screw.

[0092] In some embodiments, a housing having an ultrasonic sensor assembly positioned therein, may comprise a first well, or curved portion, configured to receive a conduit and to aid in positioning the conduit on wedges of the ultrasonic sensor assembly. The housing may optionally be configured to adjustment the positions and distances between each of the wedges in an automated fashion, or the positions and distances between the wedges may be adjusted manually. In some embodiments, a cradle may comprise a second well, or curved portion also configured to receive a conduit, and to aid in positioning the conduit on the ultrasonic sensor assembly as the conduit is coupled to the sensing assembly. A first well and a second well may be configured to be adjustably moved relative to each other in order to couple a conduit to a sensing assembly. Such adjustable movement may be linear. A first well and a second well may each comprise a “low point”, that is, a deepest point along a curved well portion. In some embodiments, a first well low point and a second well low point may be positioned on a straight line and configured to be adjustably moved towards and away from each other along the straight line.

[0093] In some embodiments, a housing may comprise a rectangular box-like portion configured to receive an ultrasonic sensor assembly. In some embodiments, a housing may comprise a compact shape, allowing it to be positioned on a conduit in a “tight” space, for example a conduit positioned close to a wall or walls. In some embodiments, a housing, together with any knob portion of an adjustment mechanism, may have a longest dimension of about 70 mm to about 130 mm. In some embodiments, the longest dimension may be greater than or equal to about 70 mm, about 80 mm, or about 90 mm, to any of about 100 mm, about 110 mm, or about 120 mm. In some embodiments, the longest dimension may be less than or equal to about 80 mm, or about 90 mm, to any of about 100 mm, about 110 mm, about 120 mm or about 130 mm.

[0094] In some embodiments, a housing may have a width of about 60 mm to about 110 mm. In some embodiments, a housing may have a width of greater than or equal to about 60 mm, about 65 mm, about 70 mm, or about 75 mm, to any of about 80 mm, about 85 mm, about 90 mm, about 95 mm, or about 100mm. In some embodiments, a housing may have awidth of less than or equal to about 65 mm, about 70 mm, or about 75 mm, to any of about 80 mm, about 85 mm, about 90 mm, about 95 mm, about 100mm, or about 110 mm. In some embodiments, a housing may comprise a depth of about 45 mm to about 95 mm. In some embodiments, a housing may comprise a depth of greater than or equal to about 45 mm, about 50 mm, about 55 mm, or about 60 mm, to any of about 65 mm, about 70 mm, about 75 mm, about 80 mm, about 85 mm, or about 90 mm. In some embodiments, a housing may comprise a depth of less than or equal to about 50 mm, about 55 mm, or about 60 mm, to any of about 65 mm, about 70 mm, about 75 mm, about 80 mm, about 85 mm, about 90 mm, or about 95 mm.

[0095] An overall space, or volume, that a housing may occupy may, for example, be from about 400 cm3 to about 800 cm3. In some embodiments, a housing may occupy a volume of greater than or equal to about 400 cm3, about 450 cm3, about 500 cm3, about 550 cm3, about 600 cm3, about 650 cm3, about 700 cm3, or about 750 cm3. In some embodiments, a housing may occupy a volume of less than or equal to about 450 cm3, about 500 cm3, about 550 cm3, about 600 cm3, about 650 cm3, about 700 cm3, about 750 cm3, or about 800 cm3.

[0096] In some implementations, a sensing assembly may be configured to be adjustably and optimally positioned on, or coupled to, conduits having an outer diameter (OD) of about 15 mm to about 51 mm. In some embodiments, the sensing assembly may be configured to couple to a conduit having an OD of greater than or equal to about 15 mm, about 18 mm, about 21 mm, about 24 mm, about 27 mm, about 30 mm, or about 33 mm, to any of about 36 mm, about 39 mm, about 42 mm, about 45 mm, or about 48 mm. In some embodiments, the sensing assembly may be configured to couple to a conduit having an OD of less than or equal to about 18 mm, about 21 mm, about 24 mm, about 27 mm, about 30 mm, or about 33 mm, to any of about 36 mm, about 39 mm, about 42 mm, about 45 mm, about 48 mm, or about 51 mm.

[0097] In some embodiments, a controller may be positioned at a housing interior. A controller may be electrically coupled to an ultrasonic sensor assembly and may also be electrically coupled to one or more LED indicator lights. One or more LED indicator lights may be configured to indicate a strength of an ultrasonic signal received by the controller from an ultrasonic sensor. For example, an LED indicator light may be configured toindicate to a user if and when a sensing assembly is properly coupled to a conduit such as an optimal ultrasonic signal is received by the controller. For example, an LED indicator light may display or flash a first color as a sensing assembly is being coupled to a conduit, and display or flash a second color when the assembly is properly installed. An LED indicator light may be configured to emit a first or a second color based on instructions from the controller.

[0098] In some embodiments, one or more LED indicator lights may be positioned at a housing exterior and visible to an installer. In other embodiments, one or more LED indicator lights may be positioned at a housing interior and may be configured to emit light from a housing interior towards a conduit and / or towards a cradle. In some embodiments, a housing may comprise a light pipe positioned at least partially about a housing edge, which edge is positioned about a housing interior. A light pipe may be positioned about an entire perimeter of a housing edge. One or more LED lights positioned at a housing interior, may be configured to emit light from the housing interior towards the light pipe, so that the light pipe will be illuminated. In this way, the indicator light may be visible from any angle or about any angle or vantage point.

[0099] In some embodiments, a light pipe may comprise a portion configured to cover a housing interior. In some cases, a light pipe may comprise a portion configured to function as a lid to position an ultrasonic sensor assembly in the housing interior. In some embodiments, one or more LED lights may be configured to illuminate an entire volume of a light pipe. A light pipe may comprise a translucent thermoplastic, for example polycarbonate. A light pipe may be a solid unitary part, or may comprise a hollow portion.

[0100] FIG. 10 depicts parts of a computer, in accordance with various embodiments. As will be appreciated, sensing assembly 100 and / or a controller of sensing assembly 100 can include one or more of the components as will be described with respect to computer 1000. Computer 1000 can be a component of a sensing assembly configured to monitor flow in a conduit, or one or more elements of the sensing assemblies described herein may be in remote communication with one or more computers such as computer 1000.

[0101] Computer 1000 can be a host computer connected to a network. Computer 1000 can be a client computer or a server. As shown in FIG. 10, computer 1000 can be any suitabletype of microprocessor-based device, such as a personal computer, workstation, server, videogame console, or handheld computing device, such as a phone or tablet. The computer can include, for example, one or more of processor 1001, computer input device 1002, output device 1003, storage 1004, and communication device 1005. Computer input device 1002 can generally correspond to those described above and can either be connectable or integrated with the computer.

[0102] Computer input device 1002 can be any suitable device that provides input, such as a touch screen or monitor, keyboard, mouse, or voice-recognition device. Output device 1003 can be any suitable device that provides output, such as a touch screen, monitor, printer, disk drive, or speaker.

[0103] Storage 1004 can be any suitable device that provides storage, such as an electrical, magnetic, or optical memory, including a RAM, cache, hard drive, CD-ROM drive, tape drive, removable storage disk, or other non-transitory computer readable medium. Storage 1004 can include one storage device or more than one storage device. As used herein, the terms storage, memory, and / or storage medium / media may refer to singular and / or plural devices which may store data and / or code / instructions individually, redundantly, and / or in cooperation with one another, for example in a local and / or cloud storage environment. Communication device 1005 can include any suitable device capable of transmitting and receiving signals over a network, such as a network interface chip or card. The components of the computer can be connected in any suitable manner, such as via a physical bus or wirelessly. Storage 1004 can be a non-transitory computer-readable storage medium comprising one or more programs, which, when executed by one or more processors, such as processor 1001, cause the one or more processors to execute methods described herein.

[0104] Software 1006, which can be stored in storage 1004 and executed by processor 1001, can include, for example, the programming that embodies the functionality of the present disclosure (e.g., as embodied in the systems, computers, servers, and / or devices as described above). In some embodiments, software 1006 can be implemented and executed on a combination of servers such as application servers and database servers.

[0105] Software 1006, or part thereof, can also be stored and / or transported within any computer-readable storage medium for use by or in connection with an instruction execution system, apparatus, or device, such as those described above, that can fetch and execute instructions associated with the software from the instruction execution system, apparatus, or device. In the context of this disclosure, a computer-readable storage medium can be any medium, such as storage 1004, that can contain or store programming for use by or in connection with an instruction execution system, apparatus, or device.

[0106] Software 1006 can also be propagated within any transport medium for use by or in connection with an instruction execution system, apparatus, or device, such as those described above, that can fetch and execute instructions associated with the software from the instruction execution system, apparatus, or device. In the context of this disclosure, a transport medium can be any medium that can communicate, propagate, or transport programming for use by or in connection with an instruction execution system, apparatus, or device. The transport-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, or infrared wired or wireless propagation medium.

[0107] Computer 1000 may be connected to a network, which can be any suitable type of interconnected communication system. The network can implement any suitable communications protocol and can be secured by any suitable security protocol. The network can comprise network links of any suitable arrangement that can implement the transmission and reception of network signals, such as wireless network connections, T1 or T3 lines, cable networks, DSL, or telephone lines.

[0108] Computer 1000 can implement any operating system suitable for operating the network. Software 1006 can be written in any suitable programming language, such as C, C++, Java, or Python. In various embodiments, application software embodying the functionality of the present disclosure can be deployed in different configurations, such as in a client / server arrangement or through a web browser as a Web-based application or Web service, for example.Fixture Detection

[0109] Systems and methods of the disclosure may be configured to recognize and identify water-use devices and appliances (e.g. “fixtures”, also referred to herein as “fixturedetection” or “end-point detection”). Fixture detection may allow for a user to determine the efficiency of water use for each individual water-use fixture in their home to identify areas where efficiency could be improved. For example, systems and methods of the disclosure may be configured to recognize and identify a flow rate and volume related to a specific water-use device or appliance. Present systems and methods may be configured to detect water flow rates multiple times per second and to send data to a backend or cloud server over a network according to a publishing protocol. A backend or cloud server may be configured to analyze the data and to identify a specific water-use fixture, e.g. a toilet, a shower, etc., from which the data can be attributed.

[0110] Fixture detection may be difficult because water use may, and typically does, overlap. For example, a faucet may be used while a toilet tank is refilling or while someone is using a shower. Further, water pressure can vary and thereby affect flow rates, as well as other factors. In some embodiments, systems and methods described herein may utilize a machine learning model to identify fixtures by matching entire function signatures to previously obtained training data. In some embodiments, systems and methods described herein may perform fixture detection using a machine learning model trained from generalized training data. In some embodiments, the training data may be obtained from regular use of the state machine in a specific location or residence, allowing the machine learning model to recognize flow rate segments that are specific to a particular residence, thus improving accuracy and precision. Training may typically comprise recording time to a flow rate function signature by marking a start and end of a fixture water usage.[OHl] In some embodiments, fixture detection may be performed on a cloud server by employing a publishing (e.g. compression) protocol and an optional training phase. In some embodiments, a publishing protocol may be implemented on / at a sensing assembly, e.g. at a sensing assembly controller. Published data may be transmitted from a sensing assembly to a cloud server. Fixture detection may then be performed at the cloud server. Fixture detection may comprise detection with untrained data, and the controller may be configured to selftrain based on long-term analysis of collected data. In some embodiments, fixture detection may be carried out in the presence of overlapping water usage and may be improved with training and refinement of the segmentation machine learning model, as will be described.

[0112] In some embodiments, present systems and methods may be configured to detect a fixture (e.g. toilet) with a multi-stage and multi-feature process utilizing time, accumulated volumes, and flow rates. Aided by a publishing (e.g. compression) protocol, sampled data may be separated into time / duration segments of either changes in flow rate or level / flat segments of flow rate. In some embodiments, present systems and methods comprise sampling water flow rates 10 times per second or more, so the publishing protocol can be used to selectively transmit only the most relevant data to the cloud server, as will be described.

[0113] Several properties may be derived from segments which may be used for fixture detection analysis: duration, volume accumulation, direction (up / down), and starting and ending flow rate. In some embodiments, changes in flow rate may be observed by measurements being close together in time, while level segments may have longer durations of time. Detection of fixtures (e.g. toilets) may then be performed by matching segments against either validated, typical segments for that particular fixture, or if in a training cycle, against previously observed properties of a particular fixture in a home, for example. For instance, toilets will generally have a start segment of a specific volume, duration, direction, and flow rate change (e.g. indicating the initiation of a flush), a level segment having a specific flow rate, volume and duration (e.g. indicating that the toilet flush valve has opened), and an end segment having a duration, volume, and flow rate change (e.g. indicating that the toilet flush valve has closed). Once the segments have been associated with, or “matched” as corresponding to a toilet flush, they may be excluded from being used from other detections / matches. Features for toilet segments can be matched heuristically, with predetermined values, or by using Machine Learning (ML) techniques such as gradient boosting to recognize specific toilet flow rate function behaviors. Level segments may be indicative of an accumulated water volume and are also indicative that a segment may belong to a toilet.

[0114] Matching, e.g. detecting, a fixture may be done by analyzing features (e.g. flow rate segments) to observe how closely they match either a trained general model or a trained specific model (e.g. specific to the particular residence / building / fixture). In some embodiments, if at any point a flow rate might fall below a predefined threshold rate — for example, if a toilet end segment no longer changes sufficiently downward — the segments may not be matched to a toilet, and thus a toilet may not be associated with the correspondingwater flow usage data. In some embodiments, present systems and methods allow for a mode where fixture detection is done in a streaming fashion by using optimistic detection. Optimistic detection may involve starting a potential detection whenever a possible toilet start segment has been detected and later relaxing the predefined threshold that is indicative of a toilet start statement. This can help ensure that a relative importance or weight given to a toilet does not guarantee that the fixture detected from the flow rate segments will be a toilet.

[0115] An exemplary segmentation pattern 200 is shown in FIG. 2A. In some embodiments, the pattern of flow rate segments shown in FIG. 2A may be indicative of a toilet. As described here and in further detail below, the sensing assembly may sample flow rates over short time segments by sending and receiving ultrasonic signals in the conduit (e.g. a pipe) at fixed or variable time intervals, or a combination of both. For example, the sensing assembly may be configured to sample the conduit (e.g. send and receive ultrasonic signals) every 100-200 ms in a default state. Upon detecting a positive change (e.g. an increase) in flow rate by the controller, the sensing assembly may be configured to sample the conduit more frequently so as to capture a water use event (e.g. toilet flush) more accurately and to reduce overlap between the detected water use event and background water use / other water use events happening at the same time. Upon detecting a negative change (e.g. decrease) in flow rate, the controller may configure the sensing assembly to reduce the frequency of sampling back to its default state.

[0116] The segmentation machine learning model may be configured to separate these sampled flow rate segments according to whether the segments have a fixed flow rate (indicated as a flat line) or a change in flow rate (indicated as a sloped line). The machine learning model may be configured to associate a particular pattern of segments with a particular water use fixture. The segmented data may also be stored by the controller and / or cloud server and may be used to train the model so that it can be customized to detect the water usage of a particular fixture in a particular home / building.

[0117] In some embodiments, segment 202 as shown in FIG. 2A may indicate the start of a toilet flush. For example, a start of a toilet flush may be marked by a sharp increase in flow rate over a short period of time, as with segment 202. The middle of a toilet flush may be characterized by a brief, flat segment 204, which can indicate a period during the toilet flush in which the flush valve is fully open and the flow rate has stabilized. The end of the toiletflush may be characterized by a third flat segment 206 at a lower flow rate than segment 204, when the flush valve closes. Then, there may be a sloped segment 208 indicating a sharp decrease in flow rate that eventually levels off after the tank has refilled with water and the flush cycle has ended.

[0118] Although the example shown in FIG. 2A shows an exemplary segmentation pattern corresponding to a toilet, a similar segmentation machine learning model can be used to segment water flow data from other fixtures, for example, sinks, showers, and the like, which may have their own distinct, identifiable segmentation patterns. In some embodiments, a process may comprise separating total measured flow rate from specific identified fixtures. This can help to ensure that total volumes are correctly calculated and that the identified fixture volumes are not over-counted. Segments can be short in time, and therefore the chance of an overlap between fixtures during a start or end segment can be minimized. In some embodiments, overlapping water use events may be permitted and individually detected. For example, a shower may be used at a same time a toilet is being refilled. Segments in the middle may not match a water use event and may be excluded. Volumes accumulated may be estimated based on the level segments found. For example, volumes of segments may be calculated by taking the integral of the flow rate graph between a start point and an end point of a flow rate segment, while accumulated volumes may be calculated over time by adding up the volumes of each segment between a time period of two different level segments. Therefore, volumes can be calculated correctly even in the presence of overlaps, and the user is able to view a historical accumulation of the volume of water they use.

[0119] Detection may be performed at a cloud server, allowing the segmentation machine learning model to be self-training as historic and aggregate data of a variety of residences can be analyzed. For example, a cloud server may also be configured with a segmentation machine learning model that can associate a particular pattern of segments with a particular water use fixture. Additionally, the use of a cloud server may advantageously allow for a large volume of data to be stored. Accordingly, water flow data collected by the ultrasonic sensor assembly can be stored in a cloud server and can be used to train and refine the machine learning model. Self-training may occur over hours or days as more fixture detections are performed. Non-repeating features may be concluded to not be toilets, e.g. asfalse positives. Repeated matches may strengthen a machine learning model’s ability to perform fixture detection based on the pattern of segments through repetitive training.

[0120] In some embodiments, the controller and / or a cloud server may perform fixture detection with an instance rule-based machine learning model to perform segmentation and fixture detection, while a state machine can be used to perform rule adjustment based on the identified fixture. An exemplary fixture detection processing pipeline 210 is shown in FIG. 2B. Data from the ultrasonic sensor assembly (e.g. time between sending and receiving ultrasonic signals) may be fed into processing pipeline 210. The initial steps of the processing pipeline, 212-214, are described here and in further detail in the next section. Step 212 may involve determining a contribution of temperature and background noise (e.g. background water flow) to the ToF data collected by the sensing assembly. These contributions can be determined in order to improve the accuracy of what will be considered a “no flow” state, or zero flow segment, by the machine learning model.

[0121] At step 214, the controller / cloud server may determine a flow rate, as well as a moving average of a flow rate, from data collected by the ultrasonic sensor over a fixed or variable period of time. Step 216 of the processing pipeline may involve calculating a time and volume corresponding to each determined flow rate. At step 218, segmentation of the water flow data by the machine learning model may divide up the water flow dataset into level (or nearly level) segments corresponding to a constant flow rate, and transition (sloped) segments corresponding to changes of water flow rate up or down. Both raw flow rates and moving averages of flow rates may be employed by the machine model for segmentation.

[0122] Using segmentation as applied by a machine learning model may be advantageous in detection of individual water use fixtures, such as toilets, because controlled valve mechanisms are not continuously varying, and in general, a large portion of the dataset is zero flow rate. This allows for a machine learning model to reliably detect the onset of a water use event (e.g. a toilet flush) as a rapid change in flow rate. For example, in the toilet flush example of FIG. 2A, the observed flow rate would be zero up until the brief period of time following the initiation of the flush, and it would then return to zero up until the toilet is flushed again.

[0123] At step 218, a segment may be defined in accordance with a start time of the segment, a stop time of the segment, a type, e.g. direction (up, down, or level) of the segment, a flow rate at the start of the segment, a flow rate at the end of the segment, and a volume of the segment. At step 220, rules may be applied to the segments using a state machine. A state machine essentially selects the rules that govern the determination of a fixture from the segmented flow rate data when the sensing assembly is operating in different states. In some embodiments, the “state” of the sensing assembly may be “tracking / not tracking / end tracking,” “transition,” “toilet / not toilet,” etc. For example, as shown in FIG. 2C, a determination that the sensing assembly is in a state of not detecting a toilet may cause various rules to be applied by the state machine. The rules can help to determine whether the segments could correspond to other fixtures besides toilets or whether the sensing assembly is transitioning from a tracking to a non-tracking state. The rules applied by the state machine generally pertain to time, volume, or change in flow rate. Although the rules shown in FIG. 2C pertain to toilets, similar rules can also be used for other fixtures. In some embodiments, one or more machine learning models may be used to refine the rules applied by the state machine for improved accuracy and reduction of false positive matches, as shown in FIG. 2B.

[0124] An exemplary state machine is represented in FIG. 2C. While FIG. 2C is an example specific to toilet detection, other state machines may be used that are tailored to other fixtures, such as showers, washing machines, etc. Multiple instances of state machines may be running in parallel. For example, a toilet state machine may run at the same time as a shower state machine. Additionally, a general state machine (e.g. a state machine with generalized rules that are not specific to a particular building, plumbing system or fixture) may run at the same time as a state machine with rules that are tailored to a particular plumbing system (e.g. a specific residence or building). In some embodiments, every segment that the machine learning model determines to be a toilet start segment may prompt the generate a new tracking instance of a state machine.

[0125] Some rules may work for overlapping water use fixtures, for example, when a shower is running at the same time that a toilet is flushed. However, volume accumulation may not be accounted for correctly by simpler sets of rules in cases of overlapping water flow from multiple fixtures. In general, rules for segments and time may match even in overlapping situations, but volume accumulation may need to be adjusted. Thus, in someembodiments, the rules applied by the state machine may be adjusted to account for overlapping water flow from multiple fixtures. For example, the time bound rules of the state machine for measuring an accumulated volume may increase a duration of the accumulated volume calculation upon detecting an overlapping water flow situation to account for a period of lower water pressure due to the use of multiple fixtures at the same time. To account for decreased water pressure, the time bounds of each segment will be longer, but accumulated volumes will be the same.

[0126] In some embodiments, validation may be performed on the segmented data. An instance rule-based machine learning model as described herein was found to properly detect a toilet flush with 97% accuracy when validated with laboratory data. In some embodiments, the instance-rule based machine learning model may be validated through manual tagging of real user data based on visual inspection. In some embodiments, the model may yield 85% accuracy when validated in this manner using a manually tuned rule base. Manual tagging may be useful when toilets are not detected in segmented data, or when they are overdetected relative to other fixtures.No-flow Rate Detection and Leak Detection

[0127] In some embodiments, the controller / cloud server may also be configured to determine a contribution of a flow rate due to temperature and may apply a continuous temperature compensation data processing operation to flow rate samples so as to obtain highly accurate measurements without knowing an actual material temperature at any point in time.

[0128] Ultrasonic water flow rate measuring devices may face inaccuracies because, as water temperature changes, this may result in an observed a change in flow rate when it is merely a change in temperature that has instead been interpreted as a change in water flow. Ultrasonic flow rate sensors can measure water flow rate using sound and measuring time for distance travelled. The speed of sound can vary with temperature. Thus, a calibrated ultrasonic sensor can determine the contribution of temperature to a flow rate from measuring a speed of sound through the conduit if a distance between an emitter and a receiver is known.

[0129] Water in a standing conduit or pipe can adjust to the ambient surrounding temperature. When water flows, a water temperature may rapidly change to that of the source water, which is typically ground temperature. Therefore, ambient temperature may typically be higher than ground temperature or source water temperature.

[0130] A controller of the sensing assemblies described herein and / or a cloud server may be configured such that temperature is reported at various flow rates so that a relative change can be accurately determined via a continuous temperature compensation data processing operation, even if the actual material temperature is not known. Additionally, a controller and / or a cloud server may include an algorithm to determine when there is no flow, e.g. zero flow. In some embodiments, the sensing assembly may be calibrated based on a zero-flow rate determination. For example, because flow rate is relative, even if there is no water flow, an ultrasonic sensor may report a non-zero flow rate due to calibration inaccuracies or due to the effect of temperature on the flow rate. The controller and / or cloud server may be configured with one or more algorithms to account for these inaccuracies during calibration and provide a “zero-flow rate” e.g. a “no-flow rate.” Subtle changes in flow rate caused by smaller leaks or problems with water-use fixtures can be accurately and precisely detected by comparing a flow rate sample to the zero-flow rate.

[0131] In some embodiments, an algorithm may be configured to continuously search for a zero-flow rate, e.g. a no-flow rate, for a period of time by analyzing, over a two-minute time period, for example, flow rate samples obtained from the ultrasonic sensor assembly, as shown in FIG. 9. Flow can be determined by multiplying the time difference between the upstream and downstream ultrasonic signal (“tdif ’) by a constant associated with the corresponding conduit type and / or size. The time length of sample collection and analysis (e.g. the time length over which the algorithm searches for a zero) may be fixed or variable. For example, the time period may be long enough to perform a frequency analysis (e.g. a histogram analysis) on the flow rate samples, typically on several hundred samples. The frequency of sample collection may also be fixed or can be varied. While a non-zero flow rate detected using time-of-flight from an ultrasonic sensor may not be precisely constant due to noise, periods of zero flow can be identified because they produce a highly consistent flow rate. Ultrasonic flow meters can provide a normal distribution, such that they are probabilistic about an actual flow rate.

[0132] Exemplary distributions of flow rates are shown in FIGS. 3A and 3B. The time segment of FIG. 3A may be rejected as a no-flow rate because is not a normal distribution, it has multiple peaks, and it is not evenly distributed about a single value. The time segment flow rate distribution of FIG. 3B may be accepted as a no-flow rate because it is a normal distribution, which is indicative of a consistent flow rate.

[0133] A time segment flow rate distribution can be analyzed by the controller / cloud server using statistical analysis. If a distribution is highly normal, it can be considered zeroflow, e.g. no-flow. A second algorithm may be employed to determine if the center of a distribution is near a lowest sample found in a given period of time, thus excluding constant actual flow rates. For example, if the center of a normal distribution is located near a higher flow rate sample, this may be indicative that it is an actual flow rate as opposed to a zeroflow rate. When a zero-flow rate is determined, the current temperature can be recorded along with the flow rate. This process can be repeated until there are many (e.g. 20 or more) temperature / flow rate readings. For example, zero detection may be performed by using a histogram or frequency plot of all sampled data and repeated at regular intervals. For example, samples may be taken every 20 ms, or 3000 samples per minute.

[0134] Regression analysis of data collected provides a curve as shown, for example, in FIG. 3D. Such a curve may be applied continuously to every sample. As flow rate is measured and temperature is recorded, the temperature may be mapped to a curve to determine a contribution of temperature to the flow rate. This contribution can then be removed from the measured flow rate through a continuous temperature compensation data processing operation to compensate for any artificial change in measured flow rate caused by temperature.

[0135] In some embodiments, temperature can be calculated using an algorithm measuring the speed of sound through a conduit or material for any specific known conduit size, and the sensing assembly may be calibrated at no-flow (zero flow). In this way, a compensation curve as in FIG. 3C may be generated and applied to flow rate determinations through the continuous temperature compensation data processing operation. A temperature sensor is not required for measuring actual temperature since the temperature analysis focuses on a relative change in temperature.

[0136] In some embodiments, a sensing assembly as described herein may be employed to identify a leak or failure, for example a water leak. Methods may include statistical analysis and real-time flow rate analysis. A source of a leak or failure may be determined by employing fixture detection, as described in the above section, as well as end-point detection and / or advanced modeling. A water leak may comprise a continuous slow, medium, high, or very high non-stop water flow. Typically, water leaks may comprise a constant flow rate, which may increase over time. It is possible, but not typical, for a flow rate from a leak to decrease over time. Water leaks may also comprise a transient leak, e.g. a faulty toilet valve. Such a leak may occur immediately after a toilet flush.

[0137] In some embodiments, leak detection may involve the communication of data from the sensing assembly to a backend cloud server configured to perform an analysis. This can allow for real-time monitoring by a user and can provide the ability to dynamically change leak detection parameters and to employ one or more machine learning models. For example, a similar machine learning model as is used to perform segmentation and fixture detection may be used to detect flow rate signatures associated with different leaks or defects in parts. In the case of a leaky toilet flapper, for example, a machine learning model may be configured to detect the leaky toilet flapper based on the fact that a start feature of the leaky toilet flapper flow signature is the same start feature as a normal toilet but differs in volume, duration, flow rate characteristics. A publishing protocol (e.g. a compression protocol) may be utilized, which reduces the quantity of data sent to a server while preserving high detail during water changing events.

[0138] Leak detection may be performed with a sensing assembly as described herein by measuring flow rates and using these to create aggregated volume information. High- frequency sampling of flow rates can allow for an aggregate volume to be calculated from the flow rate data continuously and in parallel with volumes of segments calculated during fixture detection. In some embodiments, a leak may be detected if flow rates are in a nontypical range for a length of time longer than a period of water usage. A leak may also be detected if flow rates are above a no-flow (e.g. the calibrated “zero”) or above a predefined threshold for certain durations of time. In some embodiments, detection of a leak may be performed by observing continuous moving averages of flow rates. Moving averages, as opposed to a start of a segment of change in flow rate, may be useful for detecting smaller,more continuous leaks. A moving average may be observed over a duration of greater than or equal to 30 seconds, and the moving average may be used to determine whether a flow rate is above or below a no-flow rate for a predetermined period of time.

[0139] In some embodiments, the controller of the sensing assembly may perform leak detection, but the necessity to access high frequency sampled data would result in large amounts of data to be transmitted to a backend server over networks such as WiFi, cellular, Lora, etc. The memory of the controller may be limited and unknown to a central system, so historical analysis and / or real-time tracking may not always be possible. The use of machine learning techniques is also more limited when using the controller. Accordingly, transmission of water flow data from the sensing assembly to a backend server, such as a cloud server, may allow for a user to more easily track their historical water use and / or their water use in real-time using the sensing assembly.

[0140] To reduce a quantity of data to be transmitted and associated costs of transmitting and storing data on a cloud server, a sensing assembly may employ a “publishing” (e.g. compression) protocol. Slow-changing water flow samples may be filtered out, since they are more likely to correspond to a no-flow rate. In some embodiments, a publishing protocol may be configured to send aggregate volume readings with every sample at about a 30 second moving average. With a long moving average, together with volume, a slow continuous leak may be observed even if individual flow rate data points are not used for a determination. Leak analysis may be performed in a cloud server, allowing for tracking of very long trends, such as over days, weeks, months, years, etc. Efficient leak detection may be performed while maintaining efficient data publishing during transmission to a backend server.

[0141] In some embodiments, large leaks may be observed as a water usage above a predefined and / or configurable threshold. Normal flow rate data, as opposed to moving averages, may be used to detect larger leaks or fixture failures, since a start of a large leak or failure can be large enough to trigger the publishing protocol to send flow rate data to a cloud server.

[0142] Detection of a leak or failure associated with a water-use device or appliance may be performed by first identifying a device or appliance water usage as described with respectto fixture detection and performing a small continuous leak recognition as outlined above. By forming an association with a start of a leak and device or appliance water usage, a source of a leak can be determined. This may be performed on a cloud server and may employ ML techniques, as described above with respect to fixture detection and an instance rule-based model.

[0143] In some embodiments, present systems and methods employ data compression, or a publishing protocol when transmitting data from a sensing assembly controller to a cloud server. To observe flow rate change events, flow rate data may be measured at a high frequency millisecond time scale, for example on a scale of from about 100 ms to about 200 ms to observe flow rate change events. Because ultrasonic sensors are probabilistic, a bellshaped distribution is observed, centered about an actual flow rate, as shown for instance in FIG. 4A and as similarly described above with respect to FIG. 3B. From this distribution, assumptions about a signal -to-noise ratio (SNR) of the detected flow rate can be made. For example, the histogram of FIG. 5 A comprises 3000 data points. Assuming there is at least a period of time with zero flow, a lowest local maximum (e.g. an initial estimate of the signal to noise ratio) will be zero. Thus, the graph in FIG. 5A is not symmetric and is skewed right. If the noise is symmetric, the range to the left of the local maximum can be considered noise, and a standard deviation will provide a noise estimate in order to determine the SNR.

[0144] With knowledge of the SNR, one can determine if an observed flow rate change is due to noise or an actual flow rate change caused by a leak or a water usage event. In some embodiments, a SNR may be used when employing the data publishing protocol. A publishing protocol may employ a variable time between “published” samples, for instance from about 20 ms to about 30 seconds. In some embodiments, “published” samples may refer to samples that are chosen to be transmitted to a server. In some embodiments, an upper limit of 30 seconds, or a longer or shorter period, may be used so that a cloud server may detect an absence of data and determine that there may be a problem with the sensing assembly.

[0145] In some embodiments, a publication protocol may comprise interpolation between data points. For example, water flow data may be sampled over 200 ms time periods. A publishing protocol may measure moving averages of water flow data at, for example, 2 seconds, 10 seconds, and 30 seconds. Each sample may comprise a time stamp, a flow rate observation, and a volume observation. A time stamp may be utilized since a timecorresponding to a published sample can vary. An illustration of captured versus published samples in accordance with the publishing protocol is shown in FIG. 4B, where circles represent fixed time interval samples, squares represent 2-second moving averages, and solid circles represent a published sample.

[0146] In some embodiments, a change in flow rate may be defined if a current sample is significantly different from a 2-second moving average of all previous samples. Significance may be defined in terms of a ratio to the SNR. For example, a significant difference may be that a current sample differs from a 2-second moving average of a previous sample by about 2 times the SNR. In FIG. 4B, moving average 450 represents a first sample that is “significantly different”. When a first sample is found to be significantly different, the previous sample 451 may also be selected to be published, thereby “closing” the line segment before the change has started and thus also capturing the flat part of the segment. If the time between the previous value is more than the fixed interval (e.g. 2 seconds), the actual value may be substituted by a moving average, as the actual sample may not represent a longer interval of time accurately, and a moving average may provide a closer approximation.

[0147] In accordance with the publishing protocol, when a sample has been found to be significantly different from a 2-second moving average, subsequent samples may continue to be published until the 2 second moving average “catches up” to a previous moving average. Although 2-second moving averages are used in this example, the time period over which the moving average is determined may be varied. In some embodiments, a moving average may be determined over a 2 to 30-second time period. In some embodiments, a moving average may be determined over a time period greater than or equal to 2, 5, 10, 15, 20, or 25 seconds. In some embodiments, a moving average may be determined over a time period of less than or equal to 5, 10, 15, 20, 25, or 30 seconds. It may be desirable to use a moving average corresponding to a “flat” part of a graph as a moving average for the publishing protocol calculations, where there is no significant change in flow rate and / or volume.

[0148] For example, a sensing assembly may sample data at fixed time periods of about 200 ms, but may calculate a moving average of flow rates over 30-second time periods. Thus, data may only be transmitted to a cloud server every 30 seconds, providing for highly “compressed” data transmission in some embodiments. For instance, if there are 5 flow rate values per sample, with samples taken every 200 ms, there are 1500 sampled values perminute. If there are not significant changes in water flow rate over 30 seconds, data published every 30 seconds as a moving average would yield 10 values per minute, or a data compression of about 99%.

[0149] One caveat with publishing data in accordance with a 30-second moving average, for example, is that slowly changing values will not trigger the publication protocol because the change in flow rate will not occur fast enough to reflect a significant difference over the moving average. This may be addressed using “creep detection,” which is a comparison between the most recent sample and the last published moving average. If the difference between the most recent sample and the previously published moving average is outside of a predetermined ratio of the SNR (e.g. 2x the SNR or higher), this can trigger a single point to be published. In this manner, the gradual “creeping” of the moving average up or down, which may be indicative of a small, continuous leak, can be published to the server and made visible to the user. In some embodiments, increases or decreases in accumulated volumes over a given time period (e.g. 30 seconds) may also be used during a creep detection process. Volume may not suffer from problems with the probabilistic values as long as enough samples in time are used. This is because the over-reporting and under-reporting of volumes can additively cancel each other out over time.

[0150] A scenario where a rapidly changing value happens to cross the moving average can be addressed by artificially extending any detected change conditions. For example, when the measurement diverges from a moving average, the current and previous samples may be published. When the current sample agrees again with a previous moving average, the two subsequent measurements may also be published without regard to their difference from the moving average such that a rapid change in flow rate may also be captured and published.

[0151] If there is not a significant difference between a measured flow rate and a moving average, the sensing assembly may be configured such that a moving average will automatically be published once every 30 seconds. If the protocol simply took the current value, this would result in a semi-random value (probabilistic around actual value). For this reason, the 30 second moving average may be used. If the time since the previous point that has been published is greater than 20 seconds but less than 30 seconds, then the 10 seconds moving average can be used. If it is greater than 4 seconds and less than 10 seconds, the 2- second moving average can be used.

[0152] Accordingly, a sensing assembly as described herein may utilize fixture detection and leak detection to collect data regarding a user’s water usage. By employing zero detection and automatic temperature calibration as described herein, water-use events from fixtures as well as leaks associated with a change in water flow rate can be determined with more precision and accuracy. A publishing protocol may also enable a compact and accurate representation of water usage by ensuring that flow rate changes associated with actual flow, as opposed to noise, are reported to a server for analysis and displayed via a user interface.

[0153] In some embodiments, a cloud server and / or a controller may be configured to report a plumbing system state as normal or abnormal (e.g. leak or failure) to the user. For example, a computing device, such as a smartphone or a laptop computer, may be linked to a controller. A controller may be configured to transmit data to a computing device. A computing device may have a graphical user interface configured to display data, including water-use device and appliance water use data, water efficiency data, efficiency data over time, etc. A controller may be configured to have 2-way communication with a cloud server, and a cloud server may be configured to have 2-way communication with a device such as a smartphone or a laptop computer. In this way, a user may be able to monitor their water usage from a single point, e.g. a smartphone. A graphical user interface may be configured to provide water use in real time as well as historical water use and trends, for specific wateruse devices and appliances and for a residence overall. By streamlining the ease of displaying water use information to a user, a graphical user interface as described herein may help a user identify areas in which water use could be reduced or efficiency could be improved, thus helping to drive behavioral change by the user to incorporate more water-saving habits.

[0154] A computing device may be configured to alert a user to unusual water use or suspected leaks or failures. In some embodiments, alerts may be adjustable by a user. For example, known issues may be muted, or sensitivity levels may be adjusted so that the user receives only the most relevant alerts. A computing device interface may be configured so that a user may be able to set and customize water-saving goals and targets.

[0155] FIG. 6A shows a user interface dashboard, according to some embodiments. Item 601 can represent global timescale navigation, initially set up as “monthly”. Item 602 is timeline navigation. Item 603 can indicate flow rate and may display on / off / high usage. Item 604 can show monthly water use (e.g. gallons, liters, etc). Item 605 can show monthlyaverage water use. Item 606 can show “water conversion”, a contextual reference utilizing daily objects to explain water volume, e.g. a daily view might show a water bottle, a weekly view may show a bathtub, etc. Item 607 can provide a usage chart, in this case monthly usage in weekly navigable intervals, color coded to show above average use. Item 608 is timeline and navigation interactions, which may be used to swipe to a previous or next month. Item 609 can show usage trends. Item 610 can provide a usage breakdown by water use fixtures. Items 611 and 612 can be video elements. For example, item 611 may be a looped header video with an angle of bank configured to react to a gesture, and item 612 may be a looped footer video. FIG. 6B provides an exemplary real-time view of items 607 / 608, showing a current flow rate 613, a flow rate graph 614, and a scale 615 in minutes.

[0156] FIG. 7A and FIG. 7B provide views of a user interface notification main-page and notification sub-page, respectively. Item 701 can provide active alerts having high priority, such as leaks. Item 702 can provide a list of active alerts. Item 703 can show notifications and alerts having a medium / low priority. Item 704 may be a filter, allowing filtering by notification type. Item 705 can show alert quick details, such as water volume, flow duration, and flow rate. Item 706 can list a “severity” of an event. Item 707 can show a date / time when the assembly started to detect an event. Item 708 can be an “awareness inquiry”. Item 709 can allow a user to turn updates for an event on or off. Item 710 can allow a user to select a frequency of alerts. Item 711, “next steps”, can provide tips on how to solve an issue. Item 712 can provide a history of an event and details. As shown in FIG. 7B, a user may optionally choose to “tag” an event, where they can manually monitor and report / validate a detected leak and its severity in order to improve the accuracy of the sensing assembly.

[0157] In addition to displaying customizable reports of water usage to a user, the user interface as described herein may also allow a user to provide feedback on the quality of leak detection / fixture detection of the sensing assembly. For example, after a toilet flush is detected and displayed on the user interface, the user interface may prompt the user to confirm whether the toilet was actually flushed during that time period or whether it was a false positive detection. This may occur for events that the user has “tagged.” Feedback provided by the user can be transmitted from a mobile device back to the sensing assembly and / or a cloud server, and the user feedback can be used for validation or refinement of one or more of the machine learning models used in fixture detection or leak detection.

[0158] Fig. 8A, Fig. 8B, Fig. 8C, and Fig. 8D, show views of a user interface interactive water level indicator for consumption comparison, according to some embodiments. A present indicator may be configured to visually represent a user’s water consumption and provide a comparison to an average consumption in a selected timeframe. An interface may have a video water level background having dynamically changing numeric values. Values shown are current monthly water use and monthly average water use. The numbers may be configured to “rise” or “fall” based on a comparison between the current water use and monthly average water use.

[0159] Fig. 8A shows an example where current water use may be lower than monthly average. Fig. 8B shows an example where current water use may be higher than the monthly average. Fig. 8C shows an example where current water use may be the same as the monthly average. As shown, relative positions of the numbers and the water level background may vary and provide a visual representation of consumption trends. A monthly consumption value, determined from a user’s water use, may drive the displayed water level, and cause it to rise or fall accordingly. A water level background may include slow waves to simulate a realistic water surface.

[0160] Further, a user interface may incorporate device motion sensing capabilities. As one turns their smartphone, a water surface may follow and change its angle based on a user’s gestures, thereby creating an interactive and immersive display. Fig. 8D shows an example where a water surface may follow and change its angle based on a smartphone position. The rising and falling water levels displayed on the user interface may be configured to correspond directly to a user’s current consumption in a selected timeframe, such as a current month. This dynamic representation may enable a user to visually understand their water usage patterns and easily compare it to their average consumption within a chosen timeframe. The use of a video background, dynamic numeric values, and device motion sensing are configured to enhance user understanding and engagement in monitoring their water consumption.

[0161] The terms “coupled” or “connected” may mean that an element is “attached to” or “associated with” another element. Coupled or connected may mean directly coupled or coupled through one or more other elements. An element may be coupled to an element through two or more other elements in a sequential manner or a non-sequential manner. Theterm “via” in reference to “via an element” may mean “through” or “by” an element.Coupled or connected or “associated with” may also mean elements not directly or indirectly attached, but that they “go together” in that one may function together with the other.

[0162] The terms “upstream” and “downstream” indicate a direction of gas or fluid flow, that is, gas or fluid will flow from upstream to downstream.

[0163] The term “towards” in reference to a of point of attachment, may mean at exactly that location or point or, alternatively, may mean closer to that point than to another distinct point, for example “towards a center” means closer to a center than to an edge.

[0164] The term “like” means similar and not necessarily exactly like. For instance, “ring-like” means generally shaped like a ring, but not necessarily perfectly circular.

[0165] The articles "a" and "an" herein refer to one or to more than one (e.g. at least one) of the grammatical object. Any ranges cited herein are inclusive. The term "about" used throughout is used to describe and account for small fluctuations. For instance, "about" may mean the numeric value may be modified by ±0.05%, ±0.1%, ±0.2%, ±0.3%, ±0.4%, ±0.5%, ±1%, ±2%, ±3%, ±4%, about ±5%, or ±10%. All numeric values are modified by the term "about" whether or not explicitly indicated. Numeric values modified by the term "about" include the specific identified value. For example, "about 5.0" includes 5.0.

[0166] The term “substantially” is similar to “about” in that the defined term may vary from for example by ±0.05%, ±0.1%, ±0.2%, ±0.3%, ±0.4%, ±0.5%, ±1%, ±2%, ±3%, ±4%, ±5%, or ±10% of the definition; for example the term “substantially perpendicular” may mean the 90° perpendicular angle may mean “about 90°”. The term “generally” may be equivalent to “substantially”.

[0167] Features described in connection with one embodiment of the disclosure may be used in conjunction with other embodiments, even if not explicitly stated.

[0168] Embodiments of the disclosure include any and all parts and / or portions of the embodiments, claims, description and figures. Embodiments of the disclosure also include any and all combinations and / or sub-combinations of embodiments.

Claims

CLAIMS1. A sensing assembly configured to monitor flow in a conduit, the sensing assembly comprising: an ultrasonic sensor assembly comprising an ultrasonic sensor, wherein the ultrasonic sensor assembly is configured to physically couple to the conduit, and wherein the ultrasonic sensor is configured to send and receive an ultrasonic signal through the conduit; and a controller electrically coupled to the ultrasonic sensor assembly, wherein the controller is configured to: determine a time of flight (ToF) of the ultrasonic signal, determine a flow rate in the conduit based on the ToF, store flow data comprising the determined flow rate, and identify an individual fixture associated with the flow data using a segmentation machine learning model.

2. The sensing assembly of claim 1, wherein identifying an individual fixture associated with the flow data comprises: sampling the flow data, and separating the sampled data into a plurality of time duration segments, wherein a first time duration segment of the plurality of time duration segments is characterized by not having changes in flow rate, and wherein a second time duration segment of the plurality of time duration segments is characterized by having changes in flow rate.

3. The sensing assembly of claim 2, wherein identifying an individual fixture associated with the flow data comprises analyzing one or more of a duration, volume accumulation, direction, starting flow rate, and ending flow rate in one or more of the plurality of time duration segments.

4. The sensing assembly of claim 2, wherein identifying an individual fixture associated with the flow data comprises applying one or more rules to the time duration segments using a state machine.

5. The sensing assembly of claim 1, wherein determining the flow rate comprises applying a temperature compensation data processing operation to the ToF of the ultrasonic signal.

6. The sensing assembly of claim 5, wherein the temperature compensation data processing operation is based on a set of no-flow rate determinations at various temperatures.

7. The sensing assembly of claim 1, wherein the flow data comprises one or more of volume data, flow rate moving average data, and a time stamp of the ultrasonic signal.

8. The sensing assembly of claim 1, wherein identifying an individual fixture associated with the flow data comprises determining a no-flow rate based on the flow data.

9. The sensing assembly of claim 8, wherein the controller is configured to calibrate the ultrasonic sensor assembly using the no-flow rate.

10. The sensing assembly of claim 8, wherein the controller is configured to identify a leak or failure of the individual fixture based on the flow data and the no-flow rate.

11. The sensing assembly of claim 10, wherein identifying a leak or failure of the individual fixture by the controller comprises identifying a volume above a defined threshold from the volume data.

12. The sensing assembly of claim 8, wherein the controller is configured to communicate with a cloud server, and wherein the cloud server is configured to identify a leak or failure of the individual fixture based on the flow data and the no-flow rate.

13. The sensing assembly of claim 12, wherein identifying the leak or failure associated with an individual fixture comprises transmitting the volume data and flow rate moving average data to the cloud server and receiving a water use trend analysis from the cloud server.

14. The sensing assembly of claim 1, wherein the controller is configured to determine a speed of sound in the conduit based on the ultrasonic signal.

15. The sensing assembly of claim 1, wherein the controller is configured to transmit the flow data to the cloud server in accordance with a publishing protocol.

16. The sensing assembly of claim 15, wherein the publishing protocol comprises transmitting the flow data upon identifying the volume above the defined threshold.

17. The sensing assembly of claim 15, wherein the publishing protocol comprises publishing a flow data sample having an immediately preceding time stamp upon identifying a change in flow rate above a defined threshold.

18. The sensing assembly of claim 15, wherein the publishing protocol comprises transmitting flow data from the controller to the cloud server at variable time periods.

19. The sensing assembly of claim 15, wherein the publishing protocol accounts for a signal-to-noise ratio (SNR).

20. The sensing assembly of claim 1, wherein the sensing assembly is configured to send and receive ultrasonic signals on a millisecond time scale.

21. The sensing assembly of claim 1, wherein the sensing assembly is configured to send and receive ultrasonic signals until a moving average catches up to a flow rate in the flow data.

22. The sensing assembly of claim 1, further comprising an LED light indicator configured to indicate a strength of the ultrasonic signal.

23. The sensing assembly of claim 1, wherein the ultrasonic sensor assembly is further configured to send a test signal through the conduit and receive a response to the test signal comprising one or more test outputs.

24. The sensing assembly of claim 23, wherein the controller is configured to determine a material of the conduit based on the one or more test outputs.

25. The sensing assembly of claim 23, wherein the controller is configured to determine a diameter of the conduit based on the one or more test outputs.

26. A sensing assembly configured to monitor flow in a conduit, the sensing assembly comprising: an ultrasonic sensor assembly comprising an ultrasonic sensor, wherein the ultrasonic sensor assembly is configured to physically couple to the conduit, and wherein the ultrasonic sensor is configured to send and receive an ultrasonic signal through the conduit; and a controller electrically coupled to the ultrasonic sensor assembly, wherein the controller is configured to: determine a time-of-flight (ToF) of the ultrasonic signal, determine a flow rate and a no-flow rate in the conduit based on the ToF, store flow data comprising the determined flow rate and a no-flow rate, calibrate the ultrasonic sensor using the no-flow rate, and identify a leak or failure of an individual fixture based on the flow data.

27. The sensing assembly of claim 26, wherein determining the flow rate comprises applying a temperature compensation data processing operation to the ToF of the ultrasonic signal.

28. The sensing assembly of claim 27, wherein the temperature compensation data processing operation is based on a set of no-flow rate determinations at various temperatures.

29. The sensing assembly of claim 26, wherein the controller is configured to determine a speed of sound in the conduit based on the ultrasonic signal.

30. The sensing assembly of claim 26, wherein the controller is configured to communicate with a cloud server, and wherein the cloud server is configured to identify the leak or failure of the individual fixture based on the flow data and the no-flow rate.

31. The sensing assembly of claim 30, wherein identifying a leak or failure of the individual fixture by the cloud server comprises transmitting the flow data from the controller to the cloud server.

32. The sensing assembly of claim 26, wherein the flow data further comprises volume data, flow rate moving average data, and time stamps of ultrasonic signals.

33. The sensing assembly of claim 32, wherein identifying a leak or failure of the individual fixture by the controller comprises identifying a volume above a defined threshold from the volume data.

34. The sensing assembly of claim 26, wherein the controller is configured to transmit the flow data to the cloud server in accordance with a publishing protocol.

35. The sensing assembly of claim 34, wherein the controller is configured to transmit the flow data in accordance with the publishing protocol upon identifying the volume above the defined threshold.

36. The sensing assembly of claim 34, wherein the publishing protocol comprises transmitting flow data from the controller to the cloud server at variable time periods.

37. The sensing assembly of claim 34, wherein the controller is configured to transmit the flow data in accordance with the publishing protocol upon identifying a change in flow rate above a defined threshold.

38. The sensing assembly of claim 37, wherein the publishing protocol comprises publishing a flow data sample having an immediately preceding time stamp upon identifying a change in flow rate above a defined threshold.

39. The sensing assembly of claim 34, wherein the publishing protocol accounts for a signal-to-noise ratio (SNR).

40. The sensing assembly of claim 26, wherein identification of the leak or failure associated with an individual fixture comprises transmitting volume data and flow rate moving average data to the cloud server and performing water use trend analysis by the cloud server.

41. The sensing assembly of claim 26, wherein the ultrasonic sensor is configured to send and receive ultrasonic signals on a millisecond time scale.

42. The sensing assembly of claim 26, wherein the ultrasonic sensor is configured to send and receive ultrasonic signals over a fixed time period from about 100 milliseconds (ms) to about 200 ms.

43. The sensing assembly of claim 26, wherein the ultrasonic sensor is configured to send and receive ultrasonic signals until a moving average catches up to the flow rate.

44. The sensing assembly of claim 26, wherein the controller is further configured to identify an individual fixture associated with the flow data using a segmentation machine learning model.

45. The sensing assembly of claim 44, wherein identifying an individual fixture associated with the flow data comprises sampling the flow data; separating the sampled data into a plurality of time duration segments, wherein a first time duration segment of the plurality of time duration segments is characterized by not having changes in flow rate, and wherein a second time duration segment of the plurality of time duration segments is characterized by having changes in flow rate.

46. The sensing assembly of claim 44, wherein identifying an individual fixture associated with the flow data comprises analyzing one or more of a duration, volume accumulation, direction, starting flow rate, and ending flow rate in one or more of the plurality of time duration segments.

47. The sensing assembly of claim 44, wherein identifying an individual fixture associated with the flow data comprises applying one or more rules to the time duration segments using a state machine.

48. The sensing assembly of claim 26, further comprising an LED light indicator configured to indicate a strength of the ultrasonic signal.

49. The sensing assembly of claim 26, wherein the ultrasonic sensor assembly is further configured to send a test signal through the conduit and receive a response to the test signal comprising one or more test outputs.

50. The sensing assembly of claim 49, wherein the controller is configured to determine a material of the conduit based on the one or more test outputs.

51. The sensing assembly of claim 49, wherein the controller is configured to determine a diameter of the conduit based on the one or more test outputs.

Images