TECHNICAL FIELD

[0001 ] The present invention relates to remote monitoring, and in particular, although not exclusively, to remote monitoring of buildings, plant equipment and the like.

BACKGROUND ART

[0002] Commercial buildings generally include plant rooms, which house mechanical equipment used in the day-to-day operation of the building. This equipment generally includes water pumps, and heating, ventilation and air-conditioning (HVAC) equipment.

[0003] It is generally essential to these buildings that the plant equipment in the plant rooms is functioning correctly. As such, the functioning of the plant equipment is generally monitored carefully by building management.

[0004] A problem with such systems of the prior art is that such monitoring generally identifies failures or serious flaws in the equipment when it has already started causing problems in the building. As such, any repairs are generally urgent, and thus expensive.

[0005] Certain systems exist with automatic monitoring functions, that alert a user to the failure. In particular, a pump or fan may be configured to detect that it is not pumping or blowing, and automatically issue an alert to the user. A problem with such systems is that they also generally detect failure, or conditions very close to failure. As such, the urgency and cost of repairs mentioned above is still present.

[0006] In short, these systems are generally reactive to faults, and thus generally identify faults once repairs have already become urgent and thus expensive.

[0007] Certain attempts have been made to avoid unpredictable repairs, including by replacing or overhauling plant equipment periodically. While this may reduce the amount of unpredictable repairs, it generally means that plant equipment is replaced or overhauled unnecessarily, causing unnecessary cost.

[0008] Similarly, periodic inspection of equipment may be performed to detect failures. However, it is generally hard to predict failure without disassembly of equipment, which is costly and also disruptive.

[0009] As such, there is clearly a need for improved remote monitoring systems and methods. [0010] It will be clearly understood that, if a prior art publication is referred to herein, this reference does not constitute an admission that the publication forms part of the common general knowledge in the art in Australia or in any other country.

SUMMARY OF INVENTION

[001 1 ] The present invention is directed to remote monitoring systems and methods, which may at least partially overcome at least one of the abovementioned disadvantages or provide the consumer with a useful or commercial choice.

[0012] With the foregoing in view, the present invention in one form, resides broadly in a remote monitoring method for remotely monitoring at least one item, the method including:

receiving, on a data interface, operating data relating to the item;

determining, using at least one processor, one or more operating parameters at least in part according to the received operating data; and

receiving subsequent operating data; and

determining that the subsequent operating data is outside of the one or more parameters;

issuing a notification to a user remote to the item in response to determining that the subsequent operating data is outside of the one or more parameters; and

updating the one or more operating parameters based upon feedback from the user in response to issuing the notification.

[0013] Advantageously, the method may identify potential issues of the item before a failure is imminent, which allows for maintenance, repair or replacement of the item in a manner that minimises disruption and cost. As the operating parameters are updated based upon feedback from the user in response to the notification, the system is able to learn based upon such input.

[0014] The method may further comprise predicting failure of the item based upon subsequent operating data and the one or more operating parameters.

[0015] The feedback from the user may include information indicative of an accuracy of the predicted failure.

[0016] The one or more operating parameters may include normal operating parameters, wherein the method further includes determining that the subsequent operating data correlates with other data in response to determining that the subsequent operating data is outside of the normal operating parameters.

[0017] The notification may include a request for confirmation from the user whether the correlation with the otherdata justifies the subsequent operating data being outside ofthe normal operating parameters. This is particularly useful in situations where abnormal operational data is justified by abnormal surrounding conditions.

[0018] Preferably, the item comprises plant equipment of a commercial building, or a commercial building itself.

[0019] Preferably, the data is transmitted to a remote server, wherein determining that the subsequent operating data is outside of the one or more parameters is at the remote server.

[0020] Preferably, the data is transmitted to the remote server at least in part wirelessly. Suitably, the data is transmitted from a location proximate to the item, to a location remote to the item, wirelessly. Such configuration may alleviate the requirement for communications equipment in association with the item.

[0021 ] Preferably, the data is captured at least in part by a remote monitoring device, in proximity to the item, and transmitted to the remote server at least in part, using a cellular data interface.

[0022] The remote monitoring device may analyse and/or pre-process the captured data prior to transmission to the remote server.

[0023] The data may be transmitted from the remote monitoring device at or near real time.

[0024] Preferably, the operating data includes data of one or more sensors coupled to, or provided in association with, the monitored item, and for capturing data relating thereto. The sensors may include accelerometers, transducers (for measuring sound or for measuring vibration), flow rate sensors, temperature sensors, pressure sensors, and/or electrical monitoring equipment (e.g. to measure current or voltages).

[0025] Preferably, the sensors are configured to capture data external to the item (e.g. vibration caused by the item).

[0026] Preferably, the operating data includes data from the item. The data interface may be coupled directly to a data interface of the item to receive said data.

[0027] Preferably, the operating parameters include one or more operating boundaries.

[0028] The operating parameters may be implicitly defined with reference to the operating data. The operating parameters may be defined with reference to an average of the operating data. [0029] The operating data and the operating parameters may be defined according to a time of day. In other words, different operating data may be defined for different times of the day.

[0030] The notification may include an alert. The alert may include a short message service (SMS) message.

[0031] The method may include identifying actual failure of the equipment, and issuing a notification to a user remote to the item based upon the actual failure.

[0032] Preferably, the operating data includes data of a similar or equivalent item. This may enable large amounts of data to be considered when predicting a failure.

[0033] Preferably, the method includes updating the operating parameters according to the failure. This may enable the method to constantly improve in predicting failures.

[0034] The method may include verifying failure of the item.

[0035] In another form, the invention resides broadly in a remote monitoring system for remotely monitoring at least one item, the system including:

a remote monitoring device, configured to capture operating data relating to the item;

at least one server, remote to the remote monitoring device, and configured to: receive, on a data interface, operating data relating to the item; determine, using at least one processor, one or more operating parameters at least in part according to the received operating data; and

receive subsequent operating data; and

determine that the subsequent operating data is outside of the one or more parameters;

issue a notification to a user remote to the item in response to determining that the subsequent operating data is outside of the one or more parameters; and

update the one or more operating parameters based upon feedback from the user in response to issuing the notification.

[0036] Preferably, the remote monitoring device includes one or more sensors configured to capture the operating data. Preferably, the sensors are interchangeable.

[0037] Any of the features described herein can be combined in any combination with any one or more of the other features described herein within the scope of the invention. [0038] The reference to any prior art in this specification is not, and should not be taken as an acknowledgement or any form of suggestion that the prior art forms part of the common general knowledge.

BRIEF DESCRIPTION OF DRAWINGS

[0039] Various embodiments of the invention will be described with reference to the following drawings, in which:

[0040] Figure 1 illustrates a remote monitoring system, for monitoring plant equipment, according to an embodiment of the present invention.

[0041] Figure 2 illustrates a schematic of a remote monitoring device, according to an embodiment of the present invention.

[0042] Figure 3 illustrates a schematic of a server, configured to receive sensor data, environmental data and plant equipment data, and generate notifications based thereon, according to an embodiment of the present invention.

[0043] Figure 4a illustrates a screenshot illustrating sensor data over time, in one embodiment of the invention;

[0044] Figure 4b illustrates a screenshot illustrating subsequent sensor data corresponding to the sensor data of Figure 4a but at a different point in time.

[0045] Figure 5a illustrates a screenshot illustrating historical sensor data over time, in one embodiment of the invention;

[0046] Figure 5b illustrates a screenshot illustrating actual sensor data corresponding to the historical sensor data of Figure 5a.

[0047] Figure 6a illustrates a screenshot illustrating historical sensor data over time, in one embodiment of the invention;

[0048] Figure 6b illustrates a screenshot illustrating actual sensor data corresponding to the historical sensor data of Figure 6a.

[0049] Figure 7 illustrates a method for monitoring plant equipment, according to an embodiment of the present invention.

[0050] Figure 8 illustrates a method for monitoring a building or plant equipment, according to an embodiment of the present invention. [0051 ] Preferred features, embodiments and variations of the invention may be discerned from the following Detailed Description which provides sufficient information for those skilled in the art to perform the invention. The Detailed Description is not to be regarded as limiting the scope of the preceding Summary of the Invention in any way.

DESCRIPTION OF EMBODIMENTS

[0052] Figure 1 illustrates a remote monitoring system 100, for monitoring plant equipment, according to an embodiment of the present invention. As outlined in further detail below, the system 100 may predict failure of the plant equipment before the failure is imminent, which allows for maintenance, repair or replacement of the plant in a manner that minimises disruption and cost.

[0053] The system 100 includes a remote monitoring device 105, configured to monitor plant equipment 1 10, such as a pump, or heating, ventilation and air-conditioning (HVAC) equipment. In particular, the remote monitoring device 105 includes one or more sensors 1 15 (not illustrated), coupled to, or provided in association with, the monitored plant equipment 1 10, and for capturing data relating thereto. The sensors 1 15 may include accelerometers, transducers (for measuring sound or for measuring vibration), flow rate sensors, temperature sensors, pressure sensors, and/or electrical monitoring equipment (e.g. to measure current or voltages).

[0054] Furthermore, the remote monitoring device 105 is coupled directly to a data interface of the plant equipment 1 10, and captures data therefrom using one or more pre-defined protocols. The data of the plant equipment 1 10 may include internal operating parameters of the plant equipment 1 10, or internal sensor data of the plant equipment 1 10, for example.

[0055] The remote monitoring device 105 then sends the captured data to a remote server 120. This is performed wirelessly from the remote monitoring device 105 to a wireless base station 125, such as a Third Generation Partnership Project (3GPP) Long Term Evolution (LTE) base station, or fourth generation broadband cellular network (4G) base station. This enables the remote monitoring device 105 to be installed in a plant room without requiring any pre-wired or existing telecommunications systems in the plant room.

[0056] The remote monitoring device 105 may analyse and/or pre-process the captured data prior to transmission to the remote server 120. As an illustrative example, in the case of a pump the remote monitoring device 105 may determine a volume of flow over a particular time period and for the pump using flow rates and time. The remote monitoring device 105 may then transmit the processed data (e.g. the volume of flow) rather than the raw data (e.g. flow rates). [0057] The remote monitoring device 105 may include an internal clock, to enable timestamps to be added to data. This is particularly useful if time critical measurements are made, or if data is collated and uploaded to the server 120 periodically. Alternatively, the internal clock may be used to create accurate measurements over time, such as measurements according to time of day, or for other periods of time.

[0058] The data may be transmitted at or near real time, and the reception time may be sufficiently indicative of a time at which the data was captured. This is particularly useful for non-time critical data, such as hourly flow reports or the like.

[0059] The remote server 120 logs and analyses the received data. From the received data, one or more normal operating boundaries are defined. These boundaries may be defined explicitly (e.g. between 10 and 12 Litres per Hour) or implicitly (e.g. ±20% of a historical average). Furthermore, the boundaries may be defined with reference to typical operating parameters that vary over time, and thus may also vary over time.

[0060] The normal operating boundaries may be defined such that further analysis may be performed when operation is outside of the one or more boundaries. Such further analysis may be to determine whether there is a reason for operation outside of the boundaries, and may include determining a correlation with one or more other datasets, such as weather. As an illustrative example, water usage may go outside of its normal operating boundaries on very hot days, as a result of the hot weather. As such, increased water usage may be justified by the increase in temperature.

[0061 ] Operating boundaries may also be defined such that operation outside of the one or more boundaries is indicative of an abnormal event (such as a pump failure), or indicative of a future failure. In particular, one or more of the operating boundaries may be configured to detect a future failure before the failure is imminent, to allow for maintenance, repair or replacement of the plant equipment 1 10 in a manner that minimises disruption and cost.

[0062] When the server 120 has analysed the received data, it determines whether the data is outside of the one or more boundaries, or whether there is any potential justification for being outside of the boundaries (e.g. through correlation with another parameter or dataset). The server then issues either an alert or other communication to one or more nominated persons 125, using one or more computing devices 130. The computing device 130 is illustrated in Figure 1 as a laptop computer, but the skilled addressee will readily appreciate that the computing device 130 may comprise a smartphone, a tablet computer, a phone, a pager or the like.

[0063] In case a correlation with another parameter or dataset is found that may justify being outside of the one or more boundaries, confirmation is requested of the one or more nominated persons 125 in a notification. The persons 125 may then respond either confirming that the operation is normal (and that the deviation from normal is justified), or that the operation is not normal, despite the correlation found. This feedback enables better decisions to be made in the future.

[0064] In case no such correlation is found, or that the parameters indicate that a fault is present, notifications in the form of alerts are sent to one or more persons 125. The alerts may be particularly useful to notify users of an issue when they are not necessarily monitoring the computing device 130, e.g. after hours.

[0065] Alternatively, a notification may be provided when non-urgent information (e.g. operating parameters indicating a non-imminent issue) is to be provided to the user, and an alert may be provided when urgent information (e.g. operating parameters indicating an imminent failure) are provided to the user. Reinforcement learning through user feedback allows the system to determine the urgency, and improve its learning over time.

[0066] As an illustrative example, a notification relating to a pump may including one or more of the following details: Pump Installation location; Customer details; an indication of the type of fault (e.g. pump unit / motor wear requiring maintenance); Vibration graphs showing the irregularity compared to the normalised graph; an indication of the severity of the fault (e.g. moderate non-scheduled maintenance required), and time information, such as a time date and frequency of the irregularity.

[0067] In one embodiment, the notifications are provided in an application of the user. The application may allow the user, or an administrator, to modify how the notifications are received, e.g. sounds, vibrations, turned off at night, etc, and what the notification contain. For example, different notifications may be provided to different users based upon their position, or any other suitable factor.

[0068] The notifications may also include contact details of one or more persons who may need to be contacted in relation to the issue, which simplifies the process by the user for taking the next steps (e.g. investigation or repair).

[0069] The use of notifications in an application has an advantage over SMS and email, as SMS and email notifications are generally hard to manage, are not dynamic (e.g. don’t provide current status), and tend to lead to notification fatigue as all SMS and email messages go through the same applications. [0070] In another embodiment, the alerts or notifications may be supplemented by, or comprise a short message service (SMS) or email message.

[0071 ] The one or more persons 125 may then investigate, and organise the maintenance, repair or replacement of the plant equipment 1 10. Feedback is then provided back to the system, which enables the system to improve its learning over time. The feedback may confirm a fault, wear or status of one or more components.

[0072] In some embodiments, the one or more persons 125 may monitor a wide range of plant equipment, and a dashboard may be provided to show the range of plant equipment being monitored and any alerts or notifications in association therewith.

[0073] The remote monitoring device 105 may include a main body, to which the sensors 1 15 are releasably coupled. This enables the remote monitoring device 105 to be easily reconfigured from configuration to another by changing the sensor 1 15.

[0074] Figure 2 illustrates a schematic of a remote monitoring device 200, according to an embodiment of the present invention. The remote monitoring device 200 may be similar or identical to the remote monitoring device 105.

[0075] The remote monitoring device 200 includes a processor 205, a memory 210 coupled to the processor 205, the memory 210 including instruction code executable by the processor 205 for capturing data and sending that data to a remote server.

[0076] A local interface 215 is also coupled to the processor 205, which enables local communication with one or more sensors 220a-220n, and an interface of the plant equipment 225. A remote interface 230, which is coupled to the processor 205, enables communication with the remote server.

[0077] Data is thus captured using the local interface 215 and the sensors 220a-220n, and transmitted to the server using the remote interface 230. The local interface may be in any suitable form but can include a contact closure interface (e.g. to plant equipment), an RS422 MODBUS interface (e.g. to other modules or plant equipment), or an i2C/SPI or A/D interface (e.g. to vibration, temp, current or pressure sensor). Similarly, the remote interface may take any suitable form, but is advantageously a cellular data interface, such as 3GPP LTE data interface.

[0078] While not illustrated, the skilled addressee will readily appreciate that the remote monitoring device 200 is powered. The remoter monitoring device may be powered using a mains power adapter, a battery, or any suitable means (e.g. a generator coupled to a flow sensor). If battery powered, the remote monitoring device 200 may be installed without having any power requirements in the area in which it is installed, which is clearly advantageous.

[0079] Figure 3 illustrates a schematic of a server 300, configured to receive sensor data 305, environmental data 310 and plant equipment data 315, and generate notifications based thereon. The server 300 may be similar or identical to the remote server 120 of Figure 1 .

[0080] The server 300 includes an analytics engine 320 which receives the sensor data 305, the environmental data 310 and the plant equipment data 315, and analyses same with reference to historical data 325 and reference data 330.

[0081 ] In particular, the analytics engine compares the sensor data 305 and plant equipment data 315 with the historical data 325 and the reference data 330, taking into account the environmental data 310, if appropriate.

[0082] The historical data 325 may comprise the sensor data 305 and the plant equipment data 315 for past periods, and may include averaging, filtering or the like. The reference data 330 on the other hand, may comprise reference data relating to a failure of the plant equipment, or similar or equivalent plant equipment, and thus comprise data indicative of a future failure.

[0083] The analytics engine 320 may provide a failure prediction 335 or an abnormal event 340. The failure prediction 335 may include details of a non-imminent failure that is predicted, such as when the failure is predicted to occur, together with suggested actions. The abnormal event 340 may comprise a failure or issue relating to an abnormal and unexpected event.

[0084] A notification engine 345 is configured to receive the failure prediction 335 or an abnormal event prediction 340, and generate a notification (and/or alert) based thereon. The notification or alert may be provided to an external system 350 (e.g. by SCADA) or to a user 355 (e.g. by messaging or notifications on a computing device).

[0085] As such, a user, such as the user 355, may organise maintenance, repair or replacement of the plant.

[0086] The historical data is constantly updated based upon the received plant equipment data 315 and the sensor data 305. The reference data is not automatically updated, but instead the notification engine provides data relating to the failure prediction 335 or the abnormal event 340 to a verification engine 360. The verification engine 360 then enables a user to verify the data, and update the reference data if appropriate. As such, the system constantly learns based upon failure predictions and abnormal events, and thus improves its prediction of failures. [0087] The analytics engine 320 may incorporate various means for identify the failure predictions 335 and the abnormal events. In one embodiment, artificial intelligence methods are employed to enable the analytics engine 320 to not only make accurate predictions, but to enable the analytics engine 320 to improve over time and more data is made available.

[0088] As outlined above, supervised and reinforcement learning is used to analyse multiple data inputs to allow the system to learn and better understand what are normal patterns and abnormal patterns. This generally includes prompting the user for confirmation whether a particular outcome is correct or not, the answer to which is used to assist in further learning.

[0089] Comparing multiple datasets allows the algorithms to find correlation and to ignore known matching data events. For example, an extreme heat event (i.e. hot weather) may be uncommon, and outside of what is normal, but when it does happen, it may cause water usage, for example, to also go outside of what is normal. By determining such correlation, the system is not only able to determine whether the system is operating normally with respect to general historical data, but also whether the system is operating normally in the particular circumstances (which themselves may be abnormal).

[0090] In addition to determining whether the plant equipment is operating normally or not, the system may be used to determine an urgency associated with abnormal operation. For example, a piece of equipment may be operating abnormally, but is unlikely to fail in the near future, and thus can be kept for some time. In another example, the data may indicate an imminent failure, which is thus very urgent.

[0091 ] Figure 4a illustrates a screenshot 400a illustrating sensor data 405 over time, and a frequency composition 410 of said sensor data 405. The frequency composition 410 is determined using a Fourier-style transformation, such as the Fast Fourier Transformation, which is well known in the field of data analytics.

[0092] Amplitude operating boundaries 415 are generated based upon the sensor data 405, and in particular such that the sensor data 405 is always within the operating boundaries during normal operating conditions. The amplitude operating boundaries may be generated based upon an initial operating period (e.g. sensor data from the first month of operation), with an additional buffer (e.g. 5%). Alternatively, the operating boundaries 415 may be initially manually set by a user, and updated over time.

[0093] Figure 4b illustrates a screenshot 400b illustrating sensor data 405’ and associated frequency composition data 410’, corresponding to the sensor data 405 and frequency composition 410, but at a different point in time. [0094] At time 420, the sensor data 405’ is outside of the operating boundaries, and thus a notification issues. This may indicate a potential future failure in the monitored equipment, which in turn allows for maintenance, repair or replacement of the plant in a manner that minimises disruption and cost.

[0095] While the operating boundaries are absolute and amplitude based, other operating boundaries may exist in the frequency domain, for example, and including implicit or relative boundaries. As an illustrative example, operating boundaries may exist in the frequency domain where each frequency component is compared to other frequency components to identify resonant frequencies, or frequency compositions out of the ordinary. Such frequency-base operating boundaries may be particular useful in identifying bearing wear in a pump, for example.

[0096] The sensor data 405 of Figure 4a and 4b is vibration data, but the principals may be applied to other types of sensor data such as sound data, temperature data and the like.

[0097] Figure 5a illustrates a screenshot 500a illustrating averaged sensor data 505 over time for a pump, and for different times of the day. The sensor data indicates a number of runs of the pump per hour, and is provided for each hour in the day.

[0098] Figure 5b illustrates a screenshot 500b illustrating sensor data 505’ and corresponding to the averaged sensor data 505, but for a particular day (rather than averaged).

[0099] One or more implicit operating boundaries may be defined according to the averaged sensor data 505, including sensor data being within ± 1 run per hour.

[00100] As indicated by disparity 510, the sensor data 505’ is two pump runs greater than the historical average of zero runs per hour for that time period, and thus a notification issues.

[00101 ] Figure 6a illustrates a screenshot 600a illustrating average sensor data 605 over time, illustrating run minutes per hour of a pump.

[00102] Figure 6b illustrates a screenshot 600b illustrating sensor data 605’ corresponding to the sensor data 605 but at a specific point in time.

[00103] One or more implicit operating boundaries may be defined according to the averaged sensor data 605, similar to as outlined above.

[00104] Initially the sensor data 605’ maps the historical sensor data 605 well, but then starts to diverge, and at time 610, the sensor data 605’ is showing zero run time, clearly outside of the operating boundaries, and thus a notification issues. This may indicate an actual future failure in the monitored equipment.

[00105] In addition to identifying that sensor data may be outside one or more operating boundaries, the system may also provide indications of the type of faults that may be present. In the case of the pump of Figures 5a and 5b, the constant running of the pump may be an indication of a leak. In the case of the system of Figures 4a and 4b, the additional vibration may be indicative of wear. In the case of the system of Figures 6a and 6b, the lack of pumping may be an indication of failure of the pump.

[00106] The above examples are indicative only, and illustrate single sensor data configurations. The system may, however, predict future plant equipment failures through the use of multiple environmental and machine/electrical incorporated sensors including but not limited to: accelerometers, sound measurements, temperature, pressure, currents, voltages, time, existing machine communication protocols and usage. As an illustrative example, the system may combine temperature with operating times, to take into account particularly hot or cold conditions.

[00107] The system as such may offer real time proactive analysis using deep leaning and machine learning models of any existing and newly installed plant machinery, such as motors, generators, pumps, variable speed drives, power supplies, using on or multiple environmental and machine/electrical incorporated/imbedded sensors. By using real time data and stored historical data the predictive software can determine future plant equipment failures. The relevant analysed data is sent to customers and repair companies via an easy to use application, in an email or through notifications.

[00108] When faults are investigated or repaired, a checklist may be provided to the person(s) responsible for the investigation or repair. The information from this checklist may be provided back to the system, to improve future accuracy of the algorithm through reinforced learning. The system may also be used to monitor human performance (human performance on repeat calls, time on-site etc.).

[00109] Figure 7 illustrates a method 700 for monitoring plant equipment, according to an embodiment of the present invention. The method 700 may be similar or identical to the method implemented on the system 100.

[001 10] At step 705, operating data relating to an item being monitored is received. The operating data may be data captured by one or more sensors of a remote monitoring device, for example. [001 1 1 ] At step 710, one or more operating parameters are determined based upon the received operating data. The operating parameters may define baseline parameters, or boundaries to which the operating parameters should be bound in normal operating conditions. Alternatively, the operating parameters may comprise average operating parameters.

[001 12] At step 715, subsequent operating data is received. The subsequent operating data may be in the same or a different format to the operating data received in step 705.

[001 13] At step 720, a risk of failure is predicted based upon received sensor data and the operating parameters. The risk may be determined as a percentage risk, and that risk may be compared to a threshold to determine if failure is predicted.

[001 14] If failure is predicted, a notification is sent to the user in step 725. Alternatively, the operating parameters are updated in step 710, and the steps are repeated.

[001 15] Figure 8 illustrates a method 800 for monitoring a building or plant equipment, according to an embodiment of the present invention. The method 800 may be similar or identical to the method implemented on the system 100.

[001 16] At step 805, operating data relating to an item or building being monitored is received. The operating data may be data captured by one or more sensors of a remote monitoring device, for example, and in the case of building monitoring, may monitor the building at its periphery (e.g. water or power going in or out).

[001 17] At step 810, one or more“normal” operating parameters are determined based upon the received operating data. The normal operating parameters may define baseline parameters, or boundaries to which the operating parameters should be bound in normal operating conditions.

[001 18] At step 815, the method determines that operating parameters of the item or building being monitored are outside of the determined normal operating parameters.

[001 19] At step 820, the method determines whether the operating parameters of the item or building that are outside of the determined normal operating parameters correlate with one or more other parameters or datasets. This enables the method to identify whether other parameters or datasets (e.g. weather data) may justify the deviation from“normal” operating parameters, in response to them being outside of the parameters, thereby alleviating the need to consistently check for such correlation.

[00120] If the operating parameters correlate with one or more other parameters or datasets, at step 825 a user is prompted to confirm whether the correlation justifies the deviation from normal. As an illustrative example, abnormally high temperatures may justify abnormally high water usage. Depending on the answer from the user, the method may learn and improve its future decisions.

[00121 ] If the operating parameters do not correlate with any other parameters or datasets (and therefor the method is unable to identify any potential justification from the deviation from normal), a notification is sent to the user in step 830.

[00122] The user may investigate and/or perform repair on the building or plant, and may provide feedback to the method based upon results of the investigation or the repair to enable the system to make better decisions in the future.

[00123] The systems and methods described above are particularly suited to retrofitting to any existing buildings or plant equipment without the need for pre-existing communications or sensors. In some embodiments, the systems and methods may integrate with third party or existing building management systems, e.g. SCADA systems.

[00124] In some embodiments, different data is provided to different users, depending on their need and role. In particular, a user interface may be provided that is customizable allowing different levels of information to be presented to different users. As an illustrative example, building managers may only require water usage and high level alerts whereas maintenance companies may require more in depth information. The user interface may be configured such that data can be accessed at any time. Such configuration enables full data access to those needing it, while minimizing notification fatigue.

[00125] The systems and methods described above may be used to detect potential problems or maintenance requirements before human interaction. This in turn reduces time to site, maintenance costs and improves human performance. Furthermore, the systems and methods may be used to save water and electricity by giving building management (or other users) the ability to change the operation of the building. This in turn may enables buildings to gain better energy star ratings.

[00126] As the methods and systems are always looking over the building or plant performance, this frees up building management for other tasks.

[00127] As outlined above, the systems and methods utilise machine learning techniques, which include supervised, unsupervised and reinforcement learning, which enables the system to quickly and efficiently learn. [00128] By comparing abnormal data to other data sources, e.g. outside weather, justification of the abnormal data can be found, and confirmed by a user. This is particularly useful when abnormal circumstances (e.g. very hot weather) causes abnormal data (e.g. high water usage). Feedback in relation to such correlation enables the systems and methods to learn such patterns over time and without any pre-existing context.

[00129] The use of time of day and day of week enables the system to take into account information such as business hours compared to non-business hours usage or data, and weekend compared to weekday usage or data. The use of a calendar even allows for public holidays and one-off or irregular events to be considered.

[00130] Advantageously, the systems and methods may reduce the need for unnecessary maintenance and unnecessary replacement of components reducing costs to repair companies and plant owners, and disruption. Furthermore, by proactively analysing plant equipment, maintenance can be conducted more precisely rather than broad time-based maintenance, which in turn allows repair/maintenance companies to become more efficient increasing profits and customer satisfaction.

[00131 ] When equipment failures can be predicted before they occur (or a imminent), scheduled shut down and repairs may be performed rather than emergency repairs which often cause more disruption to building occupants and cost more to repair due to the urgency,

[00132] In addition to the notifications mentioned above, real time data may be provided, e.g. to repair companies and plant owner, saving time and allowing repair companies to react quickly by arriving at site with better knowledge of the problem.

[00133] While the above embodiments have been described with reference to plant equipment, e.g. in a commercial building, the systems and methods may be applied to the automotive industry (e.g. monitoring vehicles), or to heavy industry (e.g. monitoring generators, excavators, cranes etc.)

[00134] In the present specification and claims (if any), the word ‘comprising’ and its derivatives including‘comprises’ and‘comprise’ include each of the stated integers but does not exclude the inclusion of one or more further integers.

[00135] Reference throughout this specification to ‘one embodiment’ or‘an embodiment’ means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, the appearance of the phrases ‘in one embodiment’ or ‘in an embodiment’ in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more combinations.

[00136] In compliance with the statute, the invention has been described in language more or less specific to structural or methodical features. It is to be understood that the invention is not limited to specific features shown or described since the means herein described comprises preferred forms of putting the invention into effect. The invention is, therefore, claimed in any of its forms or modifications within the proper scope of the appended claims (if any) appropriately interpreted by those skilled in the art.