TITLE: MULTI-SENSOR SYSTEM AND METHOD

TECHNICAL FIELD

The invention relates to methods, systems and devices for use in distance and proximity sensing applications.

In particular, the invention relates to a multi-sensor system and associated method utilising multiple sensor types.

Reference throughout the specification is made to the invention as relating to or including a level sensor, although this should not be seen as limiting.

BACKGROUND ART

Sensing systems are employed in a wide proliferation of applications and circumstances.

The sensors used in such systems monitor a commensurately wide range of metrics or parameters, such as temperature, distances, numeric event registrations, forces, acoustics, and pressure, to name but a few.

Although many highly sophisticated and sensitive sensors have been devised, they may be ill-suited to many applications with constraints on factors such as unit costs, form factor, robustness, flexibility, and so forth. For example, applications involving long-term monitoring of a plurality of distributed sites, such as the levels of municipal refuse bins place an inexorable premium on all the aforesaid constraints.

Sensor accuracy is naturally also important, with different sensor types having differing performance strengths and characteristics. Prior monitoring solutions have often adopted sensor systems utilizing a single type of sensor.

Such single sensor solutions however inevitably lead to compromises in the accuracy and/or reliability of the readings. As an example, ultrasonic sensors accurately measuring fluid levels in a storage tank may perform poorly in measuring noisy and/or moving water in e.g. sewage/stormwater systems. In contrast, a laser may give highly precise distance measurements of planar, reflective non-absorbing surface but may be impaired by environments with the converse characteristics, e.g. diffuse, absorptive, dark and/or surfaces otherwise generating low-reflectivity.

Prior municipal refuse bin level monitoring solutions have only adopted sensor systems utilizing a single type of sensor, often an ultrasonic distance sensor mounted on the underside of the lid of the refuse bin and measuring the relative distance to the refuse material located in the bin. However, litter bin ultrasonic sensors are prone to unreliable distance reading, e.g. giving misleadingly short distance measurements when the litter bin is empty or near empty and the bin liner billows into the ultrasonic emitter beam path.

It would clearly be desirable to have sensor systems that addressed at least some of these shortcomings.

Considering again, the above-discussed application of monitoring refuse bins in order to apply an effective servicing strategy, it would also be desirable for a sensing system to provide one or more of the following, including:

- ability to be retrofitted to existing bins;

- data transmission and two-way communication capacity over a low-power wireless network;

- level detection, of solids and/or fluids;

- environmental sensing;

- vibration sensing;

- orientation sensing;

- extended operational longevity;

- data accuracy confidence enhancement from multiple sensor sources;

- remotely configurable.

It is an object of the present invention to address the foregoing problems or at least to provide the public with a useful choice.

All references, including any patents or patent applications cited in this specification, are hereby incorporated by reference. No admission is made that any reference constitutes prior art. The discussion of the references states what their authors assert, and the applicants reserve the right to challenge the accuracy and pertinency of the cited documents. It will be clearly understood that, although a number of prior art publications are referred to herein, this reference does not constitute an admission that any of these documents form part of the common general knowledge in the art, in New Zealand or in any other country.

It is acknowledged that the term‘comprise’ may, under varying jurisdictions, be attributed with either an exclusive or an inclusive meaning. For the purpose of this specification, and unless otherwise noted, the term‘comprise’ shall have an inclusive meaning - i.e. that it will be taken to mean an inclusion of not only the listed components it directly references but also other non-specified components or elements. This rationale will also be used when the term‘comprised’ or 'comprising' is used in relation to one or more steps in a method or process. Further aspects and advantages of the present invention will become apparent from the ensuing description which is given by way of example only.

DISCLOSURE OF INVENTION

As expounded above, each sensor type exhibits certain characteristic performance attributes, benefits and shortcomings. If the pertinent application involved is unaffected by these shortcomings, the use of a single sensor type may be entirely sufficient and appropriate. Often, however, sensing applications are adversely impacted by a sensor’s characteristic for at least part of its required operational performance envelope.

Considering specific examples of sensor types, ultrasonic sensors are widely employed for various level-sensing, positional and proximity sensing tasks, including solids and liquid level-sensing in bins, tanks and silos. This wide-scale usage is due to several beneficial characteristics including low cost, relatively compact form-factor, reliability, non-destructive, unaffected by optical characteristics such as colour, reflectivity, transparency or

opaqueness, unaffected by illumination levels and not highly affected by dirt, dust, high- moisture environments or soiling of the sensor surface.

Conversely, limitations of ultrasonic sensors include having readings affected or inhibited by temperature and pressure, noisy environments, soft materials/fabrics, cross-talk between any other close ultrasonic sensors, vacuums and an inability to locate objects moving within a 3-dimensional space.

In contrast, laser range/distance sensors, exhibit the advantageous characteristics of, high directionality (> 5 degrees), high distance resolution (+- 2mm), temperature and pressure independent, unaffected by acoustic interference, whilst being adversely affected by highly reflective surfaces (glass, chrome plating, mirrors etc), highly absorbent and/or diffuse surfaces. Thus, it will be appreciated that although both laser and ultrasonic distance sensors operate as distance sensors, there are clearly applications where either sensor may yield superior results with respect to the other, dependant on the environmental conditions and the nature of the measured surface.

According to a first aspect of the present invention there is provided a multi-sensor system, configured to provide, from multiple sensors, a determination of a distance measurement indicative of a separation between at least one distance sensor and a target surface within a predetermined target range, said sensor system including at least two types of distance sensors from the group including:

- laser,

- ultrasonic,

- infrared,

- ultraviolet,

- radar, or

- any other form of contactless, distance-sensing sensor,

each distance sensor generating corresponding sensor data and said distance measurement determination including one or more sensor metrics including said sensor data, and

wherein, determination of said distance measurement utilises sensor data from a prioritised distance sensor selected, at least in part, according to a calculation of whether one or more sensor metrics meets one or more prioritisation criteria.

As used herein, the terms“distance-sensing sensor” and“distance sensor” also include proximity sensors, and any other form of contactless, proximity-sensing sensor, including magnetic flux sensors, reed switches, ambient light sensors, audible sonic sensors, in addition to laser, ultrasonic, ultraviolet, radar, or any other form of contactless, distance sensing sensor.

As used herein, the terms“distance data” and“distance sensor data” may include data of an emitted and received signal such as Time of Flight (TOF), phase shift or a distance calculated from TOF or phase shift, e.g. a distance proportional to the product of said TOF and the emission velocity/speed of the signal.

As used herein, the term“sensing signal” refers to a signal emitted and/or received by a sensor and may include, for example, laser, infrared, ultraviolet, other light or other electromagnetic waves or sound waves e.g. ultrasound.

A“low-power wireless network” as described herein should be understood to include a wireless network capable of providing the necessary data transfer rates for Machine to Machine (M2M) and Internet of Things (loT) devices, including networks with minimalist protocols, not requiring devices to continually maintain an open receive-window.

It should be appreciated that reference herein to a“computer processor”,“processor” or the like in the singular should be interpreted to include the plural e.g. multiple computer processors of which any individual processor may be located locally, remotely or virtually on e.g. cloud systems.

Reference herein to a“computer processor”,“processor” or“processing means” should be interpreted to include not only conventional computer processors such as found in desktop, mobile, wearable devices, loT devices, sensors, cameras and the like but also

programmable Integrated Chips or any device capable of performing calculations using electronic data.

The term“resolution” as used herein indicates the smallest reading or change in reading that can be reported by a sensor, system or device.

The term“precision” as used herein indicates the smallest reading or change in reading that can be taken satisfactorily and reliably by a sensor, system or device.

The term“accuracy” when referring to a distance indicates how close the reading is to the actual or true distance.

The term“mathematical manipulation”, includes any function, definition, relationship, interaction, interdependence, correlation, weighting, bias, preference and/or any other statistical, logical, or inferential influence, or relationship.

As used herein, the term“distance” and“distance measurement” includes distance, proximity, and absolute or relative distances and proximity.

The term“target surface” includes any form of object, material, solid, liquid surface, vapour, droplets, and/or particles and is not restricted to refuse, trash, litter, or rubbish.

As used herein, the term“bin” is not limited to refuse bins, but also includes any receptacle, container, housing, tank, silo, enclosure, walled-void, or any other object or location capable of retaining some form of measurable contents.

As used herein, the terms“empty bin”,“empty state”,“bin empty” should not be interpreted strictly as an absolute value and use of the terms herein should be interpreted as a label, status, or indicator that the bin is empty, near-empty or effectively empty, and/or may be treated as empty or effectively empty by the multi-sensor system.

As used herein, the term“orientation sensor” may include devices or systems capable of indicating an orientation or change in orientation of an object and may include for example, a gyroscope (e.g. rotary, vibratory, optical gyroscopes, gyrotheodolites), accelerometers, compasses, or any other device or system capable of indicating an orientation, alignment or change therein.

The terms“level”,“fill-level” or the like used herein refer to a‘level’ to which a container, bin, or other volume is filled, e.g. this may be a percentage or absolute value. The term“level sensor” as used herein should thus be interpreted to refer to a sensor, device or system that is used to determine or provide information for determining the level.

Reference herein to“refuse”,“trash”,“litter”,“rubbish” should be understood to include any type of discarded waste, e.g. agricultural, residential, industrial, commercial or public waste.

Preferably, said sensor data includes distance data from at least one of said distance sensors.

Preferably, said determination of said distance measurement includes;

- using distance data from a single prioritised distance sensor;

- disregarding or excluding distance data from a non-prioritised distance sensor which fails to meet one or more prioritisation criteria; or

- utilising distance data from two or more prioritised distance sensors.

Preferably, utilising distance data from two or more prioritised distance sensors includes calculating a distance measurement from:

- an average value of distance data from said two or more prioritised distance sensors;

- a maximum value of distance data said two or more prioritised distance sensors;

- a minimum value of distance data said two or more prioritised distance sensors, or

- any other mathematical manipulation of distance data from said two or more prioritised distance sensors.

Preferably, said prioritisation criteria further includes data from:

- environmental metrics, and/or

- target surface metrics.

The distance data need not necessarily include an absolute distance measurement and instead may include data from which distance may be derived, e.g. time of flight of a known speed signal, phase shift of a waveform or a differential between an emitted and received amplitude, intensity or another variable of an emitted signal.

Preferably, the sensor data includes at least one of:

- reflectance;

- wavelength;

- amplitude;

- intensity;

- signal path divergence;

- accuracy;

- resolution;

- frequency;

- power spectrum;

- precision, and/or

- proximity.

Preferably, said sensor metrics include sensor data, including at least one of pressure, temperature, orientation, vibration, volume and proximity data respectively provided by at least one corresponding:

- pressure sensor;

- temperature sensor;

- orientation sensor;

- vibration/movement sensor;

- volume/resonant frequency sensor;

- proximity sensor.

Preferably, said environmental metrics data is derived from ambient and/or localised environmental conditions within said predetermined distance range, said environmental metrics data including at least one of:

- lighting;

- noise;

- temperature;

- humidity;

- proximity; - pressure;

- atmospheric composition.

Preferably, said target surface metrics includes at least one of:

- reflectivity;

- absorptivity;

- specularity;

- diffusivity;

- opacity;

- orientation;

- rigidity;

- texture;

- regularity.

Preferably, said determination of the prioritised distance sensor is, at least in part, calculated from a predetermined calibration of a sensor metric according to one or more target surface metrics.

Preferably said determination of the prioritised distance sensor includes a pre-set calibration calculated using at least one target surface metric.

Preferably, said target surface metrics are externally derived, externally defined, empirically determined, anticipated or projected.

To illustrate the above with non-limiting examples, ultrasonic sensors employed in level sensing applications such as monitoring litter bins, wheeled bins, dumpsters, portable toilets and tanks are known to require monitoring of, and compensation from, an environmental metric such as temperature, whilst being unaffected by environmental metrics such as lighting levels, or target surface metrics such as colour. Thus, when monitoring levels in larger refuse or recycling containers such as dumpsters or clothing bins with relatively stable environmental conditions and/or predictable target surfaces (i.e. clothing), an ultrasonic sensor provides multiple beneficial attributes. If a multiple-input level sensor of the present invention was configured with an ultrasonic distance sensor, paired with an infrared distance sensor, the corresponding sensor metrics of the two distance sensors would differ significantly.

As is well understood, infrared sensors work on the principle of emitting and detecting reflected light waves in the InfraRed (IR) spectrum, while ultrasonic sensors use the same principle with ultrasonic sound waves. Thus, in the above example, the determination of the distance measurement may include prioritisation criteria for an environmental metric such as a threshold ambient light level and/or a target surface metric (e.g. minimum reflectivity) meeting predetermined prioritisation criteria.

The interior of an enclosed clothing bin (typically sited in unsecured public areas and thus configured to prevent unauthorised removal of the deposited clothing) would typically be unlit, in contrast, an open clothing bin located at a recycling centre may receive strong direct sunlight. The clothing would also have varying reflectivity levels depending on the material, composition, coatings or presence of additives.

Infrared distance sensors have difficulty in detecting the reflected sensing signal against high ambient light levels. If sensor metrics such as intensity or reflectance of the infrared sensing signals were used for the prioritisation criteria in a high ambient light level environment, the different ambient bin light levels and variable reflectivity of a target surface of clothing would generate significant irregularity in the measured intensity/reflectance levels making accurate measurement difficult.

Thus, in this example, for given prioritisation criteria corresponding to a signal level required for reliable readings, this would only be met by the ultrasonic distance sensor, this would result in the system prioritising of the ultrasonic distance sensor data over the infrared distance sensor data.

In alternative applications where, for example, the differences in the two distance sensor attributes are less marked, it may be preferable to apply different prioritisation criteria to the different distance sensors due to the effect of particular environmental or target surface metrics.

Axiomatically, there are a plethora of possible combinations of sensor types and associated sensor metrics that may be utilised, and the above example is merely by way of illustration.

Preferably, the distance sensors include at least one ultrasonic distance sensor and at least one laser distance sensor.

It should be appreciated that the laser distance sensor need not be limited to a particular wavelength of light. Preferred embodiments may use a red laser e.g. using a wavelength between 635-670nm, but it will be appreciated that it may be feasible to use other wavelengths depending on the application. Moreover, it should be noted that depending on the application, in some embodiments the laser distance sensor may be substituted with another form of optical distance sensor.

Preferably, the multi-sensor system further includes at least one;

- housing;

- computer processor electrically coupled/connected to at least one of the distance sensors;

- communication system capable of transmitting data over a wireless communication medium;

- power supply,

said processor configured to perform said determination of said distance measurement.

The computer processor may be located in the housing or provided remotely.

Preferably, the multi-sensor system further includes at least one remote computer processor, remote from said housing.

In an alternative embodiment, the multi-sensor system may include;

- at least one housing, incorporating at least one:

o computer processor electrically coupled to at least one of the distance

sensors;

o communication system capable of transmitting and/or receiving data over a wireless communication medium;

o power supply,

- at least one remote computer processor, remote from said distance sensors, said remote computer processor configured to perform said determination of said distance measurement.

The communication system is preferably also capable of receiving data over the wireless communication medium.

The communication system may include any device, computer chip or system capable of transmitting data over a wired or wireless network. Preferably, the network is a wireless network such as a cellular network or a low-power wide-area network such as NB-loT, LoRa, LoRaWAN or Sigfox, Zigbee, Bluetooth, radio, UHF or other wireless network.

According to one aspect of the present invention, there is provided a method for providing a determination of a distance measurement indicative of a separation between at least one distance sensor and a target surface, the separation being within a predetermined target range, said method performed by a multi-sensor system including a computer processor and at least two types of distance sensors from the group including:

- laser;

- ultrasonic;

- infrared;

- ultraviolet;

- radar;

- any other form of contactless, distance sensor,

each distance sensor generating corresponding sensor data and said distance measurement determination including one or more sensor metrics including said sensor data, and

wherein, determination of said distance measurement utilises sensor data from a prioritised distance sensor selected, at least in part, according to a calculation of whether one or more sensor metrics meets one or more prioritisation criteria, said method including the computer processor:

a) receiving sensor data from each distance sensor;

b) calculating whether said sensor data meets prioritisation criteria;

c) determining a distance measurement using sensor data in the form of distance data from at least one of the distance sensors meeting prioritisation criteria.

The prioritisation criteria at step c) may be the same, different or include the criteria at step b) or vice versa.

The sensor data utilised at step b) may be a subset of the sensor data received at step a). Preferably, said determination of said distance measurement includes;

- utilising distance data from a single prioritised distance sensor;

- disregarding distance data from a non-prioritised distance sensor which fails to meet one or more prioritisation criteria; or

- utilising distance data from two or more prioritised distance sensors.

Preferably, said utilising distance data from two or more prioritised distance sensors includes calculating a distance measurement from: - an average value of distance data from said two or more prioritised distance sensors;

- a maximum value of distance data from said two or more prioritised distance sensors;

- a minimum value of distance data from said two or more prioritised distance sensors, or

- any other mathematical manipulation of distance data from said two or more prioritised distance sensors.

It will be apparent to one skilled in the art, that the present invention may be applied to the beneficial combination of other sensor types and is not necessarily limited to ultrasonic and laser sensors as now described in more detail below.

It can be readily understood that the basic principles involved in distance measurements are also a key underpinning of many other forms of sensing, such as proximity sensing, position, orientation, object detection, movement sensing and more. Contactless distance sensors are overwhelmingly utilised throughout a diverse range of level sensing applications.

Ultrasonic sensors are the predominant sensor type employed in level sensing applications, for both liquids and solids. However, ultrasonic sensors are not without their drawbacks. In noisy and/or atmospherically unstable (e.g. high or fluctuating temperatures) environments, their readings may be compromised.

In the field of level sensing of refuse, it is clearly economically desirable to be able to use the same level sensor in different refuse receptacles such as litter bins, larger wheeled bins, dumpsters as well as other tanks and silos. However, the peculiarities of refuse collection impose numerous challenges of a single device to perform effectively in all such situations.

Clearly, the nature of rubbish itself is inherently varied and unpredictable, both in its intrinsic nature and in the timetabling of its deposits. This generates the twin challenges of accurately interpreting any level-sensing data inside a bin and co-ordinating such information with an effective strategy for timely emptying of the bin.

The scope and nature of the former issue differ with the differently-sized collection receptacles. In larger containers such as a dumpster, the relatively wide footprint of the container walls enables an overhead (lid-mounted) ultrasonic sensor’s relatively broad emitter signal path (i.e. a cone) to emit over a largely unimpeded path across the surface of typical rubbish accumulation.

In contrast, smaller municipal litter bins (such as those dispersed about shopping precincts, main streets, schools, and other urban areas of high pedestrian footfall) are relatively narrow in proportion to their base, small (typically less than 1 m deep) and typically fitted with a soft flexible plastic bin liner.

Emptying of such litter bins by contractors or council workers usually involves opening a lid, extracting the plastic liner and refuse therein and installing a fresh empty liner and re-closing the lid. If the bin liner is installed rapidly or without care the liner may not fully open, the operator relying on refuse deposition to force open the bin liner. Furthermore, the empty bin liner, being very light, flexible and malleable, is easily displaced inside the bin without the weighting effect of any overlaying rubbish and readily billows inwardly towards the centre.

The ultrasonic sensor distance-measuring acoustic emission extends from the sensor in a cone, before being reflected from any surfaces encountered. In the aforementioned examples the empty bin liner/bag itself thus impinges on the path of the ultrasonic signal, thereby generating potentially misleading short distance readings, implying the litter bin is full and requires emptying.

A level sensor employing a single ultrasonic distance sensor may thus be compromised when used in litter bin applications. Time of flight recorded by the ultrasonic distance sensor during a measurement event is affected by the path length which may result in the pressure wavefront reflecting off more than one surface creating a longer path and time of flight.

Other effects may be the absorption of sound by the materials in the bin or indeed the liner bag itself where insufficient energy in the pressure wavefront fails to trigger the ultrasound receiver and one or more of the multiple“chirps” from the ultrasound is lost resulting in a reduction in accuracy.

There is thus a challenge in ascertaining whether the distance measurements received from a level sensor at any particular instance are an accurate correlation to the actual fill-level of a bin. Potentially misleading factors range from:

- false readings from bin liner reflection of a billowed, displaced or incorrectly installed bin liner, e.g. one that is not pressed flush against the bin wall and base;

- light trash being temporarily supported above the bottom of the bin by friction with the bin liner that is billowed, displaced or incorrectly installed;

- elongate or unusually shaped trash, triggering close distance sensor readings, whilst still only occupying a small portion of the volume of the bin;

- misaligned sensor if lid is not closed or replaced flush after opening;

- highly reflective or absorptive bin contents, generating unexpected reflected signals;

- environmental changes, e.g. temperature, wind, humidity.

The degree of confidence in a distance reading may thus be increased by one on more corroborating indicia. Naturally, on-site visual inspection should occur each time the bin is serviced for emptying and refurbished with a fresh liner. However, this is not an appropriate or efficient means of corroborating partial bin-fill readings, as servicing should only be conducted when there is already high confidence the bin is full.

Enhancing the confidence in the interpretation of bin-fill distance measurements is therefore preferably provided by incorporating a further sensor using a different sensing signal to avoid duplication of potential errors.

Thus, according to a further aspect of the present invention, the multi-sensor system includes a level sensor, the level sensor including said at least two types of distance sensors and wherein said two types of distance sensors include a laser distance sensor and an ultrasonic distance sensor, respectively utilising a laser sensing signal and an ultrasonic sensing signal.

According to another aspect of the present invention, there is provided a level sensor including at least two types of distance sensors in the form of a laser distance sensor and an ultrasonic distance sensor, respectively utilising a laser sensing signal and an ultrasonic sensing signal.

Preferably, the at least one laser distance sensor provides laser distance data and the at least one ultrasonic distance sensor provides ultrasonic distance data.

Preferably, said sensor metrics include reflectance data derived from at least one distance sensor. Preferably, said distance sensor is said laser sensor.

The combination of both an ultrasonic and laser distance sensor, preferably within a common housing, provides a level sensor with notable advantages. The above-described ability to identify and then utilise the distance data from the most appropriate distance sensor has particular applicability in the field of bin level sensing.

Moreover, several prevailing characteristics of typical litter bin monitoring place significant constraints to achieve a practical, accurate and cost-effective system, namely, the need for:

- a low unit cost, due to a large number of sites to be monitored;

- an ability to be retrofitted to existing bins, thus reducing capital costs (and thus

viability) of implementation;

- remote sensor monitoring (and configuration) ability via data transmission over a two- way communication network:

- to have the communication network optimised for a low-energy consumption, e.g. for Machine to Machine (M2M) and Internet Of Things (loT) usage;

- level detection of irregular and unpredictable target materials;

- temperature sensing, giving insight to environmental changes (possibility impacting sensor performance), fire warnings and organic decomposition/fermentation activity;

- vibration/orientation/tilt sensing e.g. to indicate if the bin had been opened, damaged or knocked over;

- extended operational sensor longevity, providing a multi-year lifespan without battery replacement;

- ability to confidently and accurately interpret sensor readings without requiring on-site intervention/checking, - enhanced by data from multiple sensor sources.

Many of the above attributes are often encompassed by terms such as‘smart monitoring’ and have become increasingly viable and effective since developments such as M2M communications and loT networks. Regarding litter bin level sensing, it will thus be apparent that several design choices are directly influenced by the above attributes, particularly the choice of distance sensor and means of manipulating of the data obtained from the sensors.

Although more elaborate and powerful ultrasonic and laser sensors (and most other sensor types) are available, they may be unfeasible for economic or practical usage in such applications.

Instead, it will be understood that minimising power consumption is a key constraint in the choice of any component to enhance the operational longevity of the installed level sensor. The ability of low-power RF networks such as Sigfox ®, LoRa/LoRaWAN ®, NB-loT

(Narrowband loT), CAT M1 , LTE and the like, to provide large network coverage with a limited number of base-stations is dependent on the use of low data rates thus constraining the potential sensors that may be used.

The range of the radio frequency (RF) link, for a given power output, is determined by the data rate, i.e. a lower rate provides a longer range. Thus, there is an incentive to minimise the data quantity (and frequency) required to be transmitted by the monitoring level sensor. Providing a high network capacity is also attributable to the use of ultra-narrow band modulation (UNB), concentrated over small bandwidths, using a small message protocol limited, in the case of Sigfox, to 12 bytes frames.

Therefore, to minimise power usage, it is prudent to also minimise the quantity of transmitted data and perform any data processing remotely. Although such data processing could clearly be performed on-device, and as such falls within the scope of the present invention, the subsequent description focuses on the use of remote processing of at least part of the sensor metric data acquired from the distance sensors.

It should be appreciated that in embodiments involving remote-data processing via a network connection, the multi-sensor system may include both the distance sensors and a remote processor that processes the data. In an idealised scenario, remote, distributed monitoring of bins by smart sensors would yield a simple fill-level reading of the various bin levels, enabling optimisation of a corresponding collection/emptying schedule. Underfilled bins would thus not be emptied prematurely (and wastefully) while full bins would not become overfilled while awaiting servicing.

In practice, existing systems generate bin fill-level readings that can be highly variable and require appreciable interpretation/processing to output meaningful conclusions. Data provided by an accurate bin-level determination provides evidence accountability for commercial waste management activities.

As discussed previously, these difficulties are due to the unpredictability and variability of the deposited rubbish itself, as well as the timing of the deposits. Compounding these uncertainties are the additional possibilities of bins being emptied (or not emptied) unscheduled, unauthorised emptying/opening, and vandalism or exceptional environmental issues (fire, high winds, flooding, etc) displacing or impacting the bin and or the level sensor.

Utilising a single source of distance measurement sensor data to determine bin-fill level, inherently reduces the ability to cross-check readings to establish confidence levels in the fill-level accuracy. The present invention mitigates this risk by incorporating two or more different types of distance sensors. Furthermore, the present invention incorporates additional sensor metrics from at least one of the distance sensors to augment the decision making logic involved in prioritising usage of one distance sensor over the other(s).

It will be readily understood that although the following embodiment utilises reflectance from the laser distance sensor, this is for illustrative purposes only, and does not preclude reflectance from the ultrasonic distance sensor, or any other sensor metric from any sensor being utilised in the determination of said distance measurement.

Reflectance is essentially a measure of the detected intensity of a sensing signal (emitted by one of the distance sensors) after reflection from the target surface. In the case of a laser sensing signal, the reflectance, as a function of distance, for a known target surface with predetermined target surface metrics, typically exhibits a predictable characteristic response. This enables reflectance to be utilised as a corroborating input to cross-check distance measurements from the laser distance sensor.

Conversely, if there is high confidence or certainty, in the accuracy of both the distance measurement and the associated reflectance values, this may be used to derive information about an unknown target surface and/or target metric.

Although trash placed in litter bins is inherently variable, as discussed previously, identifying the trash composition is not in-itself necessarily a priority, unless it interferes with accurate level-sensing. However, the present invention advantageously utilises a means of identifying particular contents of a litter bin to simultaneously overcome a significant impediment to level-sensing.

As previously identified, a new bin liner fitted to a litter bin can generate false fill-depth readings when the empty bin liner billows inwardly or otherwise intercept an emitted signal from a distance sensor. The ultrasonic and laser distance sensors both obtain distance readings from the most proximal upper surface of the bin or its liner and then determine the short depth reading as a full bin. However, even if the bin liner is pushed flush with the walls of the bin, both the laser and ultrasonic distance sensors can give inaccurate distance readings or, in some instances, no readings at all.

The multi-sensor system may thus utilise reflectance data from at least one distance sensor, preferably said laser sensor, to compare the distance data with reflectance data for an empty liner. Given an empty bin liner (and, consequently, that the target surface i.e. the plastic liner, is known, the measured distance and/or reflectance may be compared to

corresponding expected reference distance and reflectance values to form the basis of the prioritisation criteria.

The multi-sensor system may thus be able to identify that the reflectance sensor data of a bin with a fitted empty plastic liner exhibits a detectable divergence comparative to corresponding reference reflectance sensor data, e.g. the reflectance sensor data obtained from the same liner material in a substantially planar orientation, substantially orthogonal to the laser sensing signal.

Alternatively, the multi-sensor system may be able to identify that the reflectance sensor data of the fitted empty plastic liner exhibits a detectable divergence between:

- the measured distance corresponding to a given reflectance level, and

- a known, reference separation between the distance sensor and the target surface for said given reflectance level.

Thus, even when the bin liner is known to be empty, (e.g. immediately after being emptied) and the actual distance to the target surface is known (or confidently inferred) to exceed a predetermined minimum distance, the determination of a distance measurement using the laser distance sensor may produce an invalid result, more specifically, either a short or nil reading.

However, the level sensor may advantageously utilise detection of this invalid distance measurement result as a fulfilment of prioritisation criteria that the bin liner is actually empty.

Preferably, said sensor metrics include laser reflectance data and laser distance data.

Preferably, the distance sensors include at least one ultrasonic distance sensor and at least one laser distance sensor and said determination of said distance measurement includes utilising laser distance data and ultrasonic distance data and calculating a distance measurement from:

- - an average value of the laser distance data and ultrasonic distance data;

- - a maximum value of the laser distance data and ultrasonic distance data;

- - a minimum value of the laser distance data and ultrasonic distance data, or

- - any other mathematical manipulation of the laser distance data and ultrasonic distance data.

In one embodiment, prioritisation criteria for using sensor data from the at least one ultrasonic distance sensor and/or at least one laser distance sensor includes a

predetermined or calculated correlation of the laser distance data and ultrasonic distance data.

Preferably, the correlation is a predetermined distance range, e.g. there is said correlation if both the laser distance data and ultrasonic distance data are within the predetermined range.

Preferably, the distance sensors include at least one ultrasonic distance sensor and at least one laser distance sensor and said determination of said distance measurement includes using distance data from the at least one ultrasonic distance sensor or the at least one laser distance sensor. Preferably, the distance sensors include at least one ultrasonic distance sensor and at least one laser distance sensor and said determination of said distance measurement excludes distance data from the at least one ultrasonic distance sensor or the at least one laser distance sensor if the:

- ultrasonic sensor data fails to meet one or more prioritisation criteria, or

- laser sensor data fails to meet one or more prioritisation criteria.

Preferably, the laser sensor data includes reflectance sensor data.

Preferably, the target surface is a surface of one or more objects inside at least one bin.

Preferably, said objects includes materials and substances.

In one embodiment, at least one of the objects is a flexible bin liner.

Preferably, a determination of said distance measurement indicates an empty state if reflectance sensor data from said laser sensor is less than a first predetermined reflectance threshold defined by a first function.

Preferably, a determination of said distance measurement includes a calculation defined by a third function if:

- reflectance sensor data from said laser sensor is greater than a first predetermined reflectance threshold defined by a first function, and if

- reflectance sensor data from said laser sensor is less than a second predetermined reflectance threshold defined by a second function, and if

- there is a predetermined or calculated correlation of the laser distance data and

ultrasonic distance data.

Preferably, a determination of said distance measurement utilises the laser sensor data if the reflectance sensor data from said laser sensor is greater than said second predetermined reflectance threshold, the distance measurement determined as being equivalent to, or derived from, said laser distance data.

In one embodiment, determination of said distance measurement utilises sensor data from a prioritised distance sensor selected, at least in part, according to a calculation of whether one or more sensor metrics meet the following prioritisation criteria:

- if reflectance sensor data from said laser sensor is less than a first predetermined reflectance threshold defined by a first function, then determination of said distance measurement indicates an empty state, else

- if reflectance sensor data from said laser sensor is greater than said first

predetermined reflectance threshold and less than a second predetermined reflectance threshold defined by a second function, and if

- there is a predetermined or calculated correlation of the laser distance data and

ultrasonic distance data, the distance measurement is determined as a calculation defined by a third function, else

- if reflectance sensor data from said laser sensor is greater than said second

predetermined reflectance threshold, the distance measurement is determined as equivalent to, or derived from, said laser distance data.

In one embodiment, determination of said distance measurement utilises sensor data from a prioritised distance sensor selected, at least in part, according to a calculation of whether one or more sensor metrics meet at least one of the following prioritisation criteria: - if reflectance sensor data from said laser sensor is less than a first predetermined reflectance threshold defined by a first function, then determination of said distance measurement indicates an empty state;

- if reflectance sensor data from said laser sensor is greater than said first

predetermined reflectance threshold and less than a second predetermined reflectance threshold defined by a second function;

- if there is a predetermined or calculated correlation of the laser distance data and ultrasonic distance data, the distance measurement is determined as a calculation defined by a third function, and/or

- if reflectance sensor data from said laser sensor is greater than said second

predetermined reflectance threshold, the distance measurement is determined as equivalent to, or derived from, said laser distance data.

Preferably, if none of said prioritisation criteria is met, said determination of said distance measurement determines the distance measurement as undefined.

Preferably, said first function is given by Ai x e ^{~UB 1} where Ai = a constant, L = laser distance data and Bi = a constant.

Preferably, said second function is given by: A ₂ x e ^~UBl where A ₂ = a constant, L = laser distance data and B ₂ = a constant.

Preferably, said third function is given by one of:

- an average value of the laser distance data and ultrasonic distance data;

- a maximum value of the laser distance data and ultrasonic distance data;

- a minimum value of the laser distance data and ultrasonic distance data, or

- a mathematical interrelationship of the laser distance data and ultrasonic distance data.

The constants A and B may be extrapolated from experimental data, empirically derived or otherwise ascertained based on the application and user requirements.

Preferably, said distance measurement determination is defined as an indication the bin is in an empty state if said laser reflectance data and laser distance data meet prioritisation criteria including at least one of:

- measured reflectance sensor data being below a predetermined reflectance

threshold, and/or

- measured reflectance sensor data corresponding to measured distance data below a predetermined distance threshold.

The above-described ability to resolve the potentially ambiguous and misleading fill-level readings that may occur with an empty plastic bin liner is a powerful and important insight for remote monitoring. It allows the potential elimination of wasteful false-positive readings that the litter bin requires emptying when the bin is in fact empty. It may also be incorporated as part of a multi-stage interpretation of the sensor data to establish accurate fill-level measurement.

In one embodiment there is provided a method of minimising the occurrence of bin-emptying prompts due to false indications in a bin monitoring system, the bin monitoring system including the multi-sensor system as aforementioned and configured to prevent a said bin emptying prompt issuing to a user or device indicating the bin is to be emptied, said method including providing an indication the bin is empty if said laser reflectance data and laser distance data meet prioritisation criteria including at least one of:

- measured reflectance sensor data being below a predetermined reflectance

threshold;

- measured reflectance sensor data corresponding to measured distance data below a predetermined distance threshold.

Preferably, a said threshold may be:

- a function;

- an absolute value;

- a displacement/separation value from a reference value or function.

According to one aspect, the multi-sensor system includes at least one computer processor configured with computer-readable instructions, the computer-readable instructions including instructions to conduct said determination of said distance measurement.

The computer processor may be electrically coupled/connected to the distance sensors, or alternatively, may be located remotely to said distance sensors and configured to receive the sensor data via at least one wireless communication medium.

Preferably the computer processor is part of a computer of a remote monitoring system for monitoring the at least one bin. The computer is preferably a server computer connected via a wireless network to a further processor electrically connected to the distance sensors.

Preferably, said computer processor is configured to generate data indicating the bin is empty if said laser reflectance data and laser distance data meet prioritisation criteria including at least one of:

- measured reflectance sensor data being below a predetermined reflectance

threshold;

- measured reflectance sensor data corresponding to measured distance data below a predetermined distance threshold.

Preferably the computer-readable instructions include instructions to:

a) process sensor data from each distance sensor;

b) calculate whether said sensor data meets prioritisation criteria;

c) determine a distance measurement using sensor data in the form of distance data from at least one of the distance sensors meeting prioritisation criteria.

Preferably, said determination of said distance measurement includes utilising distance data from two or more prioritised distance sensors.

In a further embodiment, the computer-readable instructions include instructions to calculate a distance measurement from:

- an average value of distance data from said two or more prioritised distance sensors;

- a maximum value of distance data from said two or more prioritised distance

sensors;

- a minimum value of distance data from said two or more prioritised distance sensors, and/or

- any other mathematical manipulation of distance data from said two or more prioritised distance sensors.

As used herein, the above reference to threshold criteria (including said first and second threshold criteria) and to functions for said thresholds (including said first, second and third functions) is provided for exemplary purposes and are not limiting. Any number of threshold criteria and/or any associated functions defining the threshold criteria may be utilised, in any sequence, or permutation.

Reference herein is made to various aspects and embodiments of the present invention. For clarity and to aid prolixity every possible combination, iteration or permutation of features, aspects and embodiments are not described explicitly. Thus, it should be appreciated that the disclosure herein includes any combination, iteration or permutation unless explicitly and specifically excluded.

BRIEF DESCRIPTION OF DRAWINGS

Further aspects and advantages of the present invention will become apparent from the following description which is given by way of example only and with reference to the accompanying drawings in which:

Figure 1 shows an upper perspective view of a level sensor for use in a multi-sensor system according to a first preferred embodiment of the present invention;

Figure 2 shows the level sensor of figure 1 , shown from a lower perspective;

Figure 3 shows an upper perspective view of the internal components of the level sensor of figure 1 ;

Figure 4 shows a lower perspective view of the internal components of the level sensor of figure 1 ;

Figure 5 shows two exemplary wheeled bins with a level sensor of figures 1 -4,

mounted to the underside of the bin lids;

Figure 6 shows a gross pollutant trap or stormwater solid waste collection cage fitted with a level sensor of figures 1 -4;

Figure 7 shows the level sensor of figures 1 -4, in a vertically sectioned view of an

empty exemplary bin with a bin fill-level of approximately 0%;

Figure 8 shows the level sensor of figures 1 -4, in a vertically sectioned view of the bin of figure 7 with a bin fill-level of approximately 50%;

Figure 9 shows the level sensor of figures 1 -4, in a vertically sectioned view of an

empty exemplary bin containing assorted solid objects with a bin fill-level of approximately 50%;

Figure 10 shows the level sensor of figures 1 -4, in a vertically sectioned view of an

empty exemplary bin with an empty, displaced bin liner;

Figure 1 1 shows the level sensor of figures 1 -4, in a vertically sectioned view of an

empty exemplary bin of figure 7, showing an open bin lid;

Figure 12 shows a graphical representation of laser sensor reflectance measured

distance for a variety of target surfaces and bin liner configurations; Figure 13 shows a schematic diagram of some of the functions of the level sensor of figures 1 -4;

Figure 14 shows a high-level schematic system diagram of communication structures of a remote monitoring system using a multi-sensor system according to one embodiment of the present invention.

BEST MODES FOR CARRYING OUT THE INVENTION

Drawing reference table

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings.

While the present invention will be discussed in conjunction with the following embodiments, it will be understood that they are not intended to limit the present invention to these embodiments alone. On the contrary, the present invention covers alternatives,

modifications, and equivalents which may be included within the spirit and scope of the present invention as described herein and as defined by the appended claims.

Furthermore, in the following detailed description of the present invention, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, embodiments of the present invention may be practised without these specific details. In other instances, well- known methods, procedures, components, and circuits have not been described in detail so as not to unnecessarily obscure aspects of the present invention.

Although specific advantages have been enumerated above, various embodiments may include some, none, or all of the enumerated advantages.

Other technical advantages may become readily apparent to one of ordinary skill in the art after review of the following figures and description.

It should be understood at the outset that, although exemplary embodiments are illustrated in the figures and described below, the principles of the present disclosure may be implemented using any number of techniques, whether currently known or not. The present disclosure should in no way be limited to the exemplary implementations and techniques illustrated in the drawings and described below.

Unless otherwise specifically noted, articles depicted in the drawings are not necessarily drawn to scale.

Some portions of the descriptions in this document are presented in terms of procedures, logic blocks, processing, protocols and other symbolic representations of operations on data bits within a computer memory. These descriptions and representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work.

In the present application, a procedure, logic block, process, function, or the like, is a self- consistent sequence of steps or instructions leading to a desired result. Reference herein will also be made to various“algorithms” which should be understood to refer to one or more computer-implemented processes, procedures, functions and/or calculations that are capable of accessing, reading, processing, modifying, creating or otherwise manipulating data.

The“steps” of each method are those requiring physical manipulations of physical quantities. Usually, although not necessarily, these quantities take the form of electrical or magnetic signals capable of being stored, transferred, combined, compared, and otherwise

manipulated in a computer system.

It should be borne in mind, however, that all of these and similar terms are to be associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless specifically stated otherwise as apparent from the following discussions, it is appreciated that throughout the present invention, discussions utilizing the terms such as "aborting," "accepting," "accessing," "adding," "adjusting," "analyzing," "applying,"

"assembling," "assigning," "balancing," "blocking," "calculating," "capturing," "combining," "comparing," "collecting," "creating," "debugging," "defining," "delivering," "depicting," "detecting," "determining," "displaying," "establishing," "executing," "filtering," "flipping," "generating," "grouping," "hiding," "identifying," "initiating,"“investigating,” "interacting," "modifying," "monitoring," "moving," "outputting," "performing," "placing," "positioning," "presenting," "processing," "programming," "querying,"“receiving” "removing," "repeating," "resuming," "sampling,"“scanning,” "selecting,"“sending,” "simulating," "sorting," "storing," "subtracting," "suspending," "tracking," "transcoding," "transforming," "transferring,"

"transforming," "unblocking," "using," or the like, refer to the action and processes of a computer system, or similar electronic computing device, that manipulates and transforms data represented as physical (electronic) quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system memories or registers or other such information storage, transmission or display devices.

To aid brevity and clarity, reference herein will also be made to hardware devices in the singular, however, such reference should be interpreted to also include multiple components forming the device and/or multiple devices sharing the function, e.g. reference herein to a “server” should be interpreted to include multiple servers, distributed servers, cloud-based servers and the like.

To aid brevity and clarity, reference herein will also be made to software in the singular, however, such reference should be interpreted to also include multiple components forming the device and/or multiple devices sharing the function, e.g. reference herein to a“server” should be interpreted to include multiple servers, distributed servers, cloud-based servers and the like.

Figures 1 -1 1 show a level sensor (1 ) of a multi-sensor system according to one embodiment of the present invention.

The following embodiment is described with respect to use of the level sensor (1 ) in measuring the fill-level of material in bins. However, this should not be seen to be limiting as the level sensor (1 ) may be used for numerous applications where relative distance measurements are required and, as a subset, applications where fill-level is to be determined, e.g. gross pollutant traps, skips, sewerage tanks and other tanks, silos, containers or applications where it’s desirable to measure the fill-level.

The level sensor (1 ) is configured to provide distance data including distance measurements between the level sensor (1 ) and a target surface (30). The level sensor (1 ) uses multiple inputs to determine the distance measurements. The level sensor (1 ) returns distance measurements within a predetermined target range, in preferred embodiments being the range or space between the level sensor (1 ) and the bottom area of the interior of the bin (19).

The level sensor (1 ) includes at least two types of distance sensors and in the embodiment shown includes an ultrasonic distance sensor (2), comprising an emitter (2a) and receiver (2b) pair. The ultrasonic distance sensor (2) is of known type and is not be explained in detail further, notwithstanding that the ultrasonic sensor (2) emits ultrasonic frequency sound (32) from emitter (2a) and measures the Time-Of-Flight (TOF) of the reflected sound waves (34) as detected by receiver (2b).

The level sensor (1 ) also includes a laser sensor (3) located in a well or recess (4) in the housing (5) of the level sensor (1 ). The recess (4) helps protect the laser sensor (3) from damage or substances coating the protective glass cover of the laser sensor (3). Laser distance sensors such as laser sensor (3) are also well-known in the art and will not be explained in detail further, notwithstanding that the sensor is also an optical distance sensor utilising an emitter and receiver pair and measuring distance based on Time-Of-Flight (TOF) of a laser sensing signal (31 , 33).

In the level sensor (1 ), each distance sensor (2, 3) generates corresponding sensor data derived from sensor metrics. In the embodiment shown in figures 1 -10, the sensor metrics include the distance data which includes at least the TOF data of the laser sensor (3) and the ultrasonic sensor (2).

The level sensor (1 ) may be programmed to perform a calculated distance measurement (6) and generate distance data with the distance measured. Flowever, there are advantages in instead sending the raw data, (e.g. TOF) to a remote computer for performing the calculation to determine the distance measured and so in preferred embodiments, the calculations are performed remotely.

The multi-sensor system makes a determination of the distance measurement by utilising the distance data from a prioritised distance sensor, i.e. either the ultrasonic sensor (2) or laser sensor (3). The prioritised distance sensor (2 or 3) is selected, at least in part, according to a determination of whether data from one or more sensor metrics exceeds at least one prioritisation criteria.

The prioritisation criteria include data from environmental metrics and/or target surface metrics.

The environmental metrics data from the environmental metrics are derived from ambient and/or localised environmental conditions within the predetermined target range and may include, for example, lighting; noise; temperature; pressure; and/or atmospheric composition.

The target surface metrics include properties or characteristics of the target surface (30) and may include, for example, reflectance; absorption; secularity; diffusivity; opaqueness;

orientation; rigidity; texture; and/or regularity.

Figures 3 and 4 show the internal components of the level sensor (1 ) within a waterproof housing (5).

Figure 3 shows an upper side of a PCB (20) which is provided to mount the circuitry and sensors (2, 3) of the level sensor (1 ). A power supply is provided in the form of a pair of batteries (21 ) mounted in a holder (not shown) and electrically connected to the circuitry on the PCB (20). The ultrasonic sensor emitter (2a) receiver (2b) are mounted to the PCB along with laser sensor (3). An antenna (22) for wireless communication connectivity is also mounted to the PCB (20). A socket (29) is provided for wired programming, diagnostics, reporting and debugging.

Figure 4 shows the underside of the PCB (20) with associated circuitry. The circuitry includes multiple components and for clarity, only pertinent components will be identified herein, including:

- a communication system provided in the form of a wireless transceiver (23) that is able to transmit and receive data over a wireless network, e.g. a cellular network, Internet Of Things (loT) network, Zigbee, Bluetooth, radio, UHF or other wireless network.

- Global Navigation Satellite System (GNSS) receiver (24) for providing location data. In the embodiment described herein, the wireless network includes the SigFox™ network which utilises an Ultra-Narrow Band (UNB) radio spectrum of 192kHz width at about 868MHz in Europe and 902MHz or 928MHz for other locales. The SigFox™ network is designed for very low-power low-data applications with uplink data transmission limits of about 12 bytes per data message sent and 6 data messages per hour. The SigFox™ network is bidirectional and can accommodate downlink data transmission limits of about 8 bytes per data message sent.

- computer processor provided in the form of an onboard computer

processor/microcontroller (25).

- temperature sensor provided in the form of thermistor (26).

- orientation sensor provided in the form of accelerometer/tilt sensor (27).

- magnetic switch (28) is provided and can be activated by an external magnet carried by an operator. The magnetic switch (28) is used as a wireless switch for activating and initialising the device.

The ultrasonic sensor (2) has an approximately 20° emission arc, 9-10° on either side of the centre. The laser sensor (3) is an ST® VL53L0X model, with an emitter and detector pair comprising a laser diode and optical sensor. The laser sensor (3) is designed for use as a proximity sensor but returns laser distance data and laser reflectance data which is used as explained in more detail below.

Figure 5 shows an exemplary wheeled bin (7) with a level sensor (1 ) mounted to the underside of the bin lid (8) via a double-sided adhesive pad. The ultrasonic sensor (2a, 2b) and laser sensor (3) face downwards into the bin (7). A dumpster (9) is also shown in Figure 5 with an inclined lid (10). The level sensor (1 ) on the dumpster (9) is attached to an angled mounting bracket (1 1 ) that in turn is attached to the underside of the dumpster lid (10).

Naturally, in alternative embodiments the sensors (2, 3) could be mounted in a portion of the housing (5) that is inclinable with respect to the rest of the housing (5) such that the orientation of the sensors (2, 3) can be adjusted without requiring use of an angled mounting bracket.

Figure 6 shows a stormwater solid waste collection cage (12) or‘gross pollutant trap’. The cage (12) filters and collects solid waste egressing from stormwater pipe (13). The level sensor (1 ) is mounted to a mounting bracket (14) attached to the side of the pipe (13). The level sensor (1 ) may thus be used to determine the fill level of the cage (12) from distance data reported by the sensors (2, 3).

Application of the level sensor (1 ) will now be described with respect to an exemplary application involving fill-level determination in municipal bin (19) such as shown in Figures 7- 1 1 .

Figures 7-1 1 show a vertically sectioned view of an exemplary bin (19). The litter bin (19) has a plastic bin liner (15) located on an inner can (16) within the bin (19) and material receiving openings (17) underneath the hinged or removable lid (8). The level sensor (1 ) is attached to the underside of the lid (8).

It will be appreciated that there are numerous types of bins and the examples shown in Figures 7-1 1 are simplified drawings, provided for illustrative purposes only.

Figure 7 shows the litter bin (19) in an empty state with no material present in the bin liner (15).

In Figure 7, the bin liner (19) has been pushed flush against the walls of the inner can (16) allowing a substantially unobstructed void between the distance sensors (ultrasonic sensor (2) and laser sensor (3)) in the level sensor (1 ) and their target surface (30), i.e. the surface of the bin liner (15) the bottom of the bin (19). The empty bin liner (15) in figure 7 represents a bin fill-level of 0%, i.e. completely empty, and corresponds to a‘maximum’ distance measurement (6) value for level sensor (1 ) in the embodiment.

The laser sensor signal (31 ) emitted from the laser sensor (2) is directed downwards towards bin liner (15) at the approximate centre of the bottom of the bin (19) and thereafter reflects outwards therefrom, according to the orientational and physical nature of the bin liner (15) at that point. A dark bin liner (15) surface clearly absorbs more incident light than a relatively light bin liner and so results in a lower reflectance measurement of the reflected laser sensor signal (33).

Moreover, an undulating, crumpled, bin liner scatters light more effectively and also results in a lower reflectance measurement of the reflected laser sensor signal (33) than a planar, shiny, surface, substantially orthogonal to the incident laser sensor signal (31 ).

As shown in Figure 7, the ultrasonic sensor signal (32) is emitted in the same general direction as the laser sensor signal (31 ) towards the bottom centre of the bin liner (15) and, similarly, the reflected ultrasonic sensor signal (34) reflects from any surface of the bin liner (15) encountered. It has been found that although it may be anticipated that a detectable reflected laser sensor signal (33) would be received readily by the level sensor (1 ), a reliably discernible signal is not detected. Furthermore, it has also been found that a reliably discernible reflected ultrasonic signal (34) is not received by the level sensor (1 ) either.

Figure 8 shows the litter bin (19) in an approximately 50% fill-level state, with the refuse (18) at a level approximately the half the height of the interior receptacle height. In figure 8, the ultrasonic sensor distance measurement (6a) and the laser sensor distance measurement (6b) are shown within a 10cm correlation.

Figure 9 shows the litter bin (19) with a number of objects inside and in particular two boxes (18a, 18b) inside, with the smaller box (18a) resting against the larger box (18b) inclined against the side of an inner can (16).

In figure 9 however, the ultrasonic sensor distance measurement (6a) and the laser sensor distance measurement (6b) have a magnitude discrepancy of greater than 10cm.

Figure 10 shows the litter bin (19) with a bin liner (15) that has been fitted without being fully opened and pushed to the bottom of the litter bin (19). In this state the ultrasonic sensor signal (32) from the ultrasonic sensor (2) encounters the surface of the displaced bin liner (15) at the upper region of the bin (19). Thus, the corresponding reflected ultrasonic sensor signal (34) generates a short ultrasonic sensor distance measurement indicating the litter bin (19) % fill-level is very high/full even though there is actually no material or objects in the bin liner (15).

The laser sensor signal (31 ) is effectively a point source with little signal divergence for a predetermined target range of less than 1 m - typical of most litter bins. Thus, the laser sensor signal (31 ) is able to pass into the remaining central void or opening in the bin liner (15) before its corresponding reflected laser sensor signal (33) reflects from its oblique incidence with the highly irregular, undulating and/or crumpled surface of the bin liner (15). Counterintuitively, although a detectable portion of the reflected laser sensor signal (33) does typically reach the laser sensor (3), and the reflectivity magnitude of the received reflected laser sensor signal (33) is low, the calculated laser sensor distance measurement (6b) returns incorrectly low/short readings, i.e. high (near full) fill-level % values.

Figure 1 1 shows the litter bin (19) with the bin lid (8) tilted open, after removal of a full bin liner (15) of material (not shown) and replacement with an empty bin liner (15). Tilt detection by the accelerometer/tilt sensor (27) gives the level sensor (1 ) the ability to notify the system of service events, vandalism and other events that result in the orientation of level sensor (1 ) changing. Tilt detection also allows the level sensor (1 ) to avoid taking measurements when the orientation of the device is not as expected.

The tilt sensor (27) is also suited for use in wheeled rubbish bins (19) with a hinged lid and portable toilets.

The level sensor (1 ) is fitted with an ST Microelectronics LIS2DH accelerometer which is permanently powered and has an interrupt line that can wake up the device from a low- power standby state.

There are two parts to tilt detection; firstly, waking the level sensor (1 ) from standby when a significant change in orientation occurs; secondly, once awake determining the actual orientation of the level sensor (1 ).

After the level sensor (1 ) is woken by an accelerometer interrupt mechanism (because orientation has changed) or when woken by normal mechanisms such as magnetic switch (28) or periodic timer, device orientation must be determined to decide if a notification must be sent or at least if it is acceptable to take a measurement.

Only sampling acceleration immediately after wakeup could result in measurements being taken while the device is still moving. While this can be used to determine current orientation, ideally the resting state after movement is complete is desired. Waiting a long period of time then re-sampling acceleration could result in missing the tilt event altogether, for example, a bin lid opened quickly, then shut.

Thus, a preferable solution is to sample acceleration periodically until it is deemed to be tilted or stable. Peak acceleration detection is also possible. The actual algorithm used to analyse the samples and thresholds will vary on a per-application basis.

Raw values of X, Y, and Z acceleration can be used to determine if the level sensor (1 ) is outside a particular range of safe angles, however, these are only completely effective when rotation is exactly about the X, Y or Z axis.

Thus, a more effective method for determining level sensor (1 ) orientation is to calculate the angular difference between the level sensor (1 ) reference vector (gravity for a level sensor (1 ) installed on a horizontal plane) and the measured acceleration vector. The reference vector can always be gravity, or where sensors are installed on a significant angle may be “zeroed” after installation.

The result is the actual angle the level sensor (1 ) differs from its reference angle which may be reported back to a server (29) if required.

A determination of whether the level sensor (1 ) is in a tilted or normal state, given the current (and possibly previous) orientation samples, may be performed according to the following algorithm:

- After a wake-up event, sample for up to one second at 100ms intervals giving up to ten samples, storing each sample as an orientation angle.

- When more than four consecutive orientation angles occur that are similar (e.g.

within 10 degrees), take an average of four as final orientation. Stop sampling.

- If no similar results occur after 10 samples, device is not stationary, average final four samples as final orientation.

Compare final orientation to threshold to determine if a notification message must be sent. Or send orientation anyway and allow server (29) to determine if orientation is acceptable.

Figure 12 shows the reflectance magnitude of multiple reflected laser sensor signals (33) plotted against the corresponding laser sensor distance measurement (6b). The individual plots (35 - 43) cover the following exemplary target surfaces (30) typically found in litter bins (19) including;

plot (35): Cardboard (35)

plot (36): ESD bag (36)

plot (37): Polystyrene (37)

plot (38): White shopping bag (38)

plot (39): Tin can (39)

plot (40): Crumpled paper (40)

plot (41 ): Transparent plastic bag (41 )

plot (42): Black plastic container (42)

plot (43): A sock (43)

Figure 12 also shows comparative plots (44 - 47) of the bin liner (15) in the following conditions:

- plot (44): fitted in the bin (19)) with the bin liner (15) allowed to settle hang

freely/loosely;

- plot (45): flat on a horizontal surface at a distance measurement (6) equal to the

maximum bin depth;

- plot (46): fitted in the bin (19) with the bin liner (15) pressed flat on the bottom of the inner bin can (16);

- plot (47): loose in the bin, where the sensor was moved to the edge of the bin and back to the centre.

Plots (35-43) were obtained by multiple readings taken with the different types of target surfaces (30) located at varying distances below the laser sensor (3). Figure 12 also shows a first threshold (48), a second threshold (49) and an encirclement, demarcating invalid results (50).

In the embodiment shown in figure 12, the first threshold (48) is defined by a first function Ai x e ^-L/B1 where Ai = a constant, L = laser distance data and Bi = a constant.

The second threshold (49) is defined by a second function A ₂ x e ^{~LIB 2} where A ₂ = a constant, L = laser distance data and B ₂ = a constant.

The plots in figure 12 show several significant details. It will be noted that the measured laser sensor signal reflectance data plotted on the Y-axis (51 ) is a logarithmic scale, while the measured laser sensor distance data (6b) plotted on the x-axis is linear. The

exponential-dependant first and second thresholds (48, 49) thus plot as straight lines.

Considering the approximate trend of all the various target surface (30) plots (35 - 43), it can be seen they also approximately correlate to a broadly similar exponential relationship, albeit with different proportionality constants, as is well understood. It will be further appreciated that the first and second thresholds (48, 49) may be defined, in alternative embodiments (not shown) by alternative distance and/or reflectance dependant functions, such as inverse, or inverse square-distance functions.

Considering the details of the plots in figure 12 more closely it can be seen that, as expected, the detected reflectivity diminishes with increasing distance in all plots (35-43). Even a separate reference plot (45) of the bin liner (15) measured without the constraints of being fitted inside the inner can (16) of the bin (19) shows a correlation (albeit at lower reflectance values) with the other target surface (30) plots (35-43), despite the absorbent, non-reflective nature of (predominately black plastic) bin liners (15). However, once the bin liner (15) is fitted inside the bin (19), the plots (44, 46-47) for the detected reflectance levels for the empty bin liner (15) show a pronounced change, which is utilised to derive several important outcomes of the level sensing system.

Firstly, the reflectivity plot (46) when the bin liner (15) is pressed flat against the bottom of the bin (19) is not shown in figure 12 as no detectable reflectivity readings were obtained.

Secondly, when the bin liner (15) is fitted to the inner can (16) and allowed to hang loosely (i.e. not pressed flush to the inner sides of the bin (16)), all the corresponding distance readings are significantly short (between approximately 100-350mm), even though the true distance readings should span a 100-800mm range. Significantly, although the readings of plot (44) are thus clearly incorrect, they are distinguishable from all the other plots (35-43). This enables a threshold function (e.g. such as a first function corresponding to threshold (48)) to be used to apply corresponding prioritisation criteria.

Thus, when measured reflectance levels such as shown in plot (44) are recorded, and the reflectance data is compared to the first threshold (48) value defined by the above-defined first function, the outcome is all the measured reflectance values of plot (44) failing to exceed the first threshold (48). Consequently, it may be then determined that the bin liner (15) is likely to be empty, and the bin (19) fill-level status is set to 0%. This understanding is highly significant, as an orthodox interpretation of such short distance measurement (6b) results would be an inference the bin (19) was in fact nearly or totally full, thus potentially triggering a futile bin (19) emptying and refurbishment visit.

It can also be seen from plot (47) in figure 12 that the reflectance values measured as the position of the laser sensor (3) (and, therefore, also the orientation) and the corresponding laser sensor signal (31 ) is varied from its typical central lid (8) position (as shown in figures 7-10) towards the outer rim of the bin (19). The resulting distance measurements would be expected to show a corresponding change from a short distance measurement (6) at the rim of the bin (19), dropping away sharply to a long measurement (6) corresponding to the full depth of the bin (19).

Instead, only the measurements at the rim produced accurate distance measurements (6), while the remainder again produces invalid short distance readings, shown located within the demarcation boundary line (50). Application of the plot (44) reflectance data with the first function establishes that only the rim-mounted laser sensor distance measurements (6) exceed the first threshold (48).

It follows from the results of plots (35-47) that reflectance measurements of the centre of the bin (19) (not the rim) that fall below the distance-dependant exponential first function defining a first threshold (48) are only exhibited by the empty bin liner (15).

In such cases, the distance measurement (6a) of the ultrasonic sensor (2) (not shown in figure 12) may be disregarded, given the confidence in the above conclusion of the status (i.e. empty) of bin liner (15). Or to state in another context, for measured laser reflectance values falling below the first threshold (48) (i.e. a form of prioritisation criteria), distance measurements (6a) from the ultrasonic sensor (2) reading are disregarded.

It may be established through testing, or otherwise through calculation or derivation that one distance level sensor may outperform another. In the embodiments described with reference to figures 1 -12, the distance measuring the performance of the laser sensor (3) has been determined to outperform the ultrasonic sensor (2) for all reflectance values exceeding the second threshold (49) defined by the second function.

If the reflectance sensor data from the laser sensor (3) for a given distance measurement (6b) places the reading in the intermediate region between the first and second thresholds (48, 49), a third function may be used to determine the distance measurement. The employment of the third function may, itself be dependent on one or more further

prioritisation criteria, such as:

- the laser sensor (3) distance measurement (6b) and the ultrasonic sensor (2)

distance measurement (6a) values are within a predetermined correlation or range, e.g. 10 cm.

Self-evidently, numerous alternative correlations may be employed.

The constants used in the threshold functions mentioned above are determined through sensor calibration or from reference data for the application.

The third function itself may equally be defined in any chosen manner. In the embodiment described with respect to figures 1 -12, the third function is:

- an average value of the laser distance data and ultrasonic distance data;

- a maximum value of the laser distance data and ultrasonic distance data;

- a minimum value of the laser distance data and ultrasonic distance data, or

- a mathematical interrelationship of the laser distance data and ultrasonic distance data.

We may now thus consider further the embodiments shown in figures 7 - 10, in light of the above analysis of figure 12.

As earlier described, in Figure 7, the bin liner (15) has been pushed flush against the walls of the inner can (16) and thus replicates the circumstances of the reflectance/distance plot (46) from the laser sensor (3), i.e. no discernible reading reliably detected.

As previously stated, the determination of distance measurement indicates an empty state if reflectance sensor data from the laser sensor (3) is less than a first predetermined reflectance threshold (48) defined by a first function.

Consequently, this results in the reflectance sensor data clearly failing to exceed the first threshold (48), therefore leading to the determination of an empty bin-liner (15) present in the bin (19).

Figures 8 and 9, both show litter bins (19) with % bin-fill situations of approximately 50%. If the reflectance sensor data obtained in the example of figure 8, gave values falling between the first and second thresholds (48, 49), then the following prioritisation criteria may be applied to the sensor data from both the ultrasonic sensor (2) and the laser sensor (3) to determine the distance measurement, i.e. a determination of said distance measurement (6) includes a calculation defined by a third function if: reflectance sensor data from the laser sensor (3) is greater than the reflectance threshold (48) defined by the first function, and if

reflectance sensor data from the laser sensor (3) is less than the second reflectance threshold (49) defined by the second function, and if

- there is a predetermined or calculated correlation of the laser distance data and ultrasonic distance data.

The third function may be defined in numerous forms including, but not limited to;

- an average value of the laser distance data and ultrasonic distance data;

- a maximum value of the laser distance data and ultrasonic distance data;

- a minimum value of the laser distance data and ultrasonic distance data, or

- a mathematical interrelationship of the laser distance data and ultrasonic distance data.

As stated earlier, in figure 8, the ultrasonic sensor distance measurement (6a) and the laser sensor distance measurement (6b) are shown within a 10cm correlation. Thus, assuming the level sensor is configured to implement the first alternative of the third function, the distance measurement (6) would be given by an average value of the laser sensor distance data (6b) and ultrasonic sensor distance data (6a). The application of the third function in this example was itself dependant on the prioritisation criteria of the laser sensor distance data (6b) and the ultrasonic sensor distance data (6a) meeting a predetermined or calculated correlation, e.g. 10 cm in this example.

In contrast, in figure 9, as earlier stated, the ultrasonic sensor distance measurement (6a) and the laser sensor distance measurement (6b) have a magnitude discrepancy of greater than 10cm.

Thus, if the above example correlation prioritisation criteria were not applied, the ultrasonic sensor distance (6a) and laser sensor distance (6b) may still be used to calculate the distance measurement according to the definition of the third function.

Alternatively, for example, if the ultrasonic sensor distance measurement (6a) and the laser sensor distance measurement (6b) have a magnitude discrepancy of greater than 10cm, the prioritisation criteria may be defined to set the distance measurement (6) as undefined.

Considering figure 10, it can now be seen that the status of the bin liner (15) is comparable to the circumstances of plot (44) in figure 12. As also stated earlier, in figure 10 the ultrasonic sensor distance measurement (6a) and laser sensor distance data (6b) both produce short measurements.

When compared to the first threshold (48) via the first function, the corresponding reflectance sensor data from said laser sensor (3) in figure 10, all fall below the threshold (48). As per previously given examples, this outcome that the reflectance sensor data from the laser sensor (3) in figure 10 falls below the threshold (48) may be then used as a fulfilment of prioritisation criteria that the bin liner (19) is actually empty.

It will be appreciated that although in this example configuration, this fulfilment of prioritisation criteria that the bin liner (15) is empty, prioritises sensor data from the laser sensor (3), over the ultrasonic sensor (2), the converse may be applied if desired.

It is reiterated again that, although examples given herein may give prioritisation criteria prioritising, for example, a laser sensor (3) over an ultrasonic sensor (2), these are illustrative only. The invention may be configured with prioritization criteria including any chosen sensors, and sensor metrics, prioritized and configured in any desired manner. Figure 13 shows a schematic diagram of some of the functions of the level sensor (1 ).

The left-most column shows the raw data provided by the level sensor (1 ). This raw data includes:

- Tilt Flag or orientation data - if the accelerometer/tilt sensor (27) detects that the level sensor (1 ) has been moved, tilted or otherwise, re-orientated past a set threshold, the processor (25) may generate a data packet including information that a reorientation has occurred. This data packet may contain simple binary data (e.g. 1 - tilt / 0 - no tilt) or more complex data including accelerometer measurement data. The ultrasonic and laser distance data will also be affected by the sensor orientation, hence the need to measure orientation to compensate for any reorientation from a‘calibrated’ or‘initial’ state.

- Over Temperature Flag or temperature data - Measurements provided by

ultrasonic sensor (2) are affected by temperature and will report different measures at different temperatures, typically at 0.17% error from calibrated temperature per degree Celsius.

Therefore, providing temperature data enables the ultrasonic distance data to be more accurately determined by compensating using the temperature data.

Additionally, if the ultrasonic sensor, for example, becomes too hot then the data provided may be too inaccurate. The processor (25) may generate a data packet including temperature information of the device based on the resistance of thermistor (26).

Alternatively, or in addition, the processor (25) may generate a data packet including information that thermistor (26) has passed a threshold temperature. Such an Over Temperature Flag may also be useful in indicating a potential fire in the bin (19).

- Ultrasonic Distance - The ultrasonic sensor (2) provides distance data in the form of TOF.

- Laser Distance - The laser sensor (3) provides distance data in the form of TOF.

- Laser Reflectance - The laser sensor (3) also provides reflectance data related to the target surface (30). This data may include intensity received by the laser receiver and intensity emitted.

The raw data is then used in a measurement phase as shown in the second column. The orientation, temperature and Ultrasonic distance data is utilised to calculate an Ultrasonic distance measurement. Similarly, the laser distance data, temperature and orientation data are used to calculate a laser distance measurement.

Laser reflectance is calculated as a percentage from received intensity vs emitted intensity.

In the third column, the bin fill level percentage is calculated using the ultrasonic distance data, laser distance data, laser reflectance data and bin dimension data in the form of the calculated or measured Bin Full Distance and Bin Empty Distance. This bin dimension data is shown in the 4 ^th column of Figure 13.

Figure 14 shows a high-level schematic system diagram of communication structures of a remote monitoring system (100) using a multi-sensor system according to one embodiment of the present invention. The remote monitoring system utilises a distributed network of level sensors (1 ) installed in bins (19). The level sensors (1 ) form the base layer in the drawing.

The level sensors (1 ) shown in the figure are physical loT devices that may use Low Power Radio communications technologies such as Sigfox (101 ) or LoRaWAN (102) or use cellular technologies (103) such as NB-loT or 4G-LTE.

Each level sensor (1 ) periodically transmits a data packet of sensor data back through the appropriate network service provider (101 -103) to the server (29).

The server (29) processes the raw data to make a determination of the distance

measurement of the separation between the level sensor (1 ) and the target surface, e.g. rubbish (18) or bin liner (15). This determination is used as an indicator of the fill-level of the bin (19).

The server (29) logs and transmits data relating to the distance measurements or fill-level of the bin (19) to servers (105) from which customers or other users may access the information at web interfaces (104) or systems management interface (106).

The server (29) also provides a bin monitoring user interface (107) and a device

management dashboard (108) to authorised users.

The determination of the distance measurement utilises the sensor data from a prioritised distance sensor (2 and/or 3) selected according to a calculation of whether one or more sensor metrics meets one or more prioritisation criteria.

The server (29) periodically receives sensor data from the level sensors (1 ). This sensor data includes laser reflectance data and laser distance data from the laser distance sensor (3) and ultrasonic distance data from the ultrasonic distance sensor (2).

On receiving the data the server (29) is configured to run software that calculates whether the laser reflectance data meets prioritisation criteria and generates corresponding data providing information about the fill-level of the bins to which level sensors (1 ) are installed.

The calculations include a structured function determining:

a) if the reflectance sensor data from the laser sensor (3) is less than a first reflectance threshold (defined by first function Ai x e ^_L/Bl where Ai = a constant, L = laser distance as reported by laser distance data and Bi = a constant) then determination of the distance measurement indicates an empty bin, else

b) if the reflectance sensor data is greater than the first reflectance threshold but less than a second reflectance threshold (defined by second function A ₂ x e ^~UBl where A ₂ and B ₂ are constants) then the distance measurement is defined as the minimum value of the laser distance data and ultrasonic distance data, else

c) if the reflectance sensor data from the laser sensor is greater than the second

threshold then the distance measurement is determined to be equivalent to the laser distance data, else

d) set the distance measurement as undefined.

The constants used in the threshold functions mentioned above are determined through sensor calibration or from reference data for the application. The Network Service Provider (101 -103) receives the data packets from sensors (1 ), often from multiple radio towers. After confirming the packet is from a sensor (1 ) on a valid service plan and combining all received packets, the Network Service Provider (101 -103) generates a data packet to send to the server (29). In generating the data packet additional information is added to the sensor data such as received signal strength, the devices network identifier and geolocation information.

Network service providers offer many options for connection to the server (29) and payload format (eg HTTP Post, REST, Microsoft loT Hub, HTML, JSON, etc...). Typically, the Network Service Provider (101 -103) will send an event within approximately 10 seconds of reception from the level sensor (1 ).

The basic up-link data flow involves server (29) receiving the data packet from the Network Service Provider (101 -103). Using a network identifier, the server (29) obtains or determines a unique identifier for each level sensor (1 ). This identifier allows the server (29) to determine the appropriate actions and data flow required for that type of device.

Typically, this determination would involve:

a) Decoding of the sensor data using a determination of distance measurement as

described above. The sensor data is decoded into metric values and flags and status information extracted.

b) Additional“calibrated” values are generated. This calibration may also include, for example, generation of a rubbish bin full level in percent from a raw distance in centimetres.

c) Analytics are performed on the received data such as filtering, sensor fusing, bounds checking, geofencing etc.

d) The raw data, calibrated data, the information appended by the network service

provider and analysis results are stored in an event record in a database on a data store connected to the server (29).

e) Data is tested against a device type and client-specific set of criteria to determine if a client event needs to be generated. If needed, an event is generated on the particular client’s API or preferred event mechanism.

f) Data is tested against a device type and specific set of criteria to determine if a

device management dashboard event needs to be generated and dispatched.

g) Client servers (105) receive the event (if one was generated) and dispatches it

appropriately. Otherwise a dashboard webpage (106) or application requests from the Server (29) the most recent event or event history required to generate its visualisations or actions.

Down Link events follow a similar but reverse process to the Up Link process:

a) The server (29) is requested to send data corresponding to an event to a level sensor (1 ). The server (29) confirms the requested event is valid for a particular client and the type of sensor (1 ).

b) The server (29) converts the event to the correct binary payload to be sent down to the sensor (1 ). c) The payload is sent to the appropriate Network Service Provider (101 -103). At this point, it is held pending by the service provider (101 -103) until the next time the level sensor (1 ) contacts its Network Service Provider with an Up Link that also checks for a pending Down Link.

d) The level sensor (1) receives the Down Link data packet and the onboard processor (25) processes the data packet. The onboard processor (25) is programmed to perform the requested action.

e) Depending on the particular level sensor (1 ) or type of event, the level sensor (1 ) may; send an Up Link message immediately in response to the event, send a response in the next required Up Link or not send a response at all and just send a normal Up Link next time one is required.

It needs to be understood that there exist implementations of other variations and

modifications of the invention and its various aspects, as may be readily apparent to those of ordinary skill in the art, and that the invention is not limited by specific embodiments described herein. Features and embodiments described throughout this specification may be combined with and without each other. It is therefore contemplated to cover any and all modifications, variations, combinations or equivalents that fall within the scope of the basic underlying principals disclosed and claimed herein.

Modifications, additions, or omissions may be made to the systems, apparatuses, and methods described herein without departing from the scope of the disclosure. For example, the components of the systems and apparatuses may be integrated or separated. Moreover, the operations of the systems and apparatuses disclosed herein may be performed by more, fewer, or other components and the methods described may include more, fewer, or other steps. Additionally, steps may be performed in any suitable order. As used in this document, “each” refers to each member of a set or each member of a subset of a set but does not imply ALL members of the set or subset.