AU2019322484A1

AU2019322484A1 - Method, apparatus and system for measuring urination events for cows

Info

Publication number: AU2019322484A1
Application number: AU2019322484A
Authority: AU
Inventors: Paul Robert Shorten; Brendon WELTON
Original assignee: AgResearch Ltd
Current assignee: AgResearch Ltd
Priority date: 2018-08-17
Filing date: 2019-08-15
Publication date: 2021-02-18
Also published as: WO2020036495A1

Abstract

The present invention relates to an apparatus, system and non-invasive methods of detecting and measuring urination events for cows. The invention involves the use of apparatus and a system in a method of detecting changes in the tilt angle of the spine of a cow and measuring the duration of those changes. If the changes and duration meet predetermined criteria, then it may be determined that a urination event has occurred for the cow. This data may be assessed and used for managing and/or minimising the potential environmental impact of nitrogen derived from urine of cows on the environment.

Description

A METHOD, APPARATUS AND SYSTEM FOR MEASURING URINATION EVENTS FOR COWS

Statement of Corresponding Applications

This application is based on the provisional specification filed in relation to New Zealand Patent Application No. 745397, the entire contents of which are incorporated herein by reference.

Field of Invention

The present invention relates to an apparatus, system and methods of measuring urination events for cows. The invention has particular application to non-invasive techniques of measuring urination events for cows by detecting changes in the tilt angle of the spine during the urination event.

Background to the Invention

The implications of excessive nitrogen being excreted in urine by grazing livestock such as dairy cows and sheep onto farmed land is well known. Dairy farm systems in particular are prone to high nitrogen losses, largely due to only about 10 to 30% of the nitrogen consumed by a dairy cow being converted to products (in the form of milk, meat, fibre). The remainder of the consumed nitrogen is primarily excreted in urine and deposited -on the soil in concentrated areas at high nitrogen application rates equivalent to about 500 to 1,200 kilograms nitrogen per hectare.

The nitrogen derived from this urine is well in excess of that required by pasture plants and after transformations in soil, it is prone to being lost via leaching or gaseous emissions pathways. Nitrogen derived from the urine of grazing animals has been found to contribute significantly to nitrous oxide emissions, ammonia emissions and nitrate leaching to groundwater. Therefore, a good opportunity exists to reduce nitrogen losses from grazed pasture land is by specifically targeting individual urine patches.

There is evidence that cows exhibit repeatable phenotypic variation in urination event volume and the frequency of urinations events per day. This difference in urination event volume and daily urination frequency between individual cows can be as much as two-fold. Furthermore, the nitrogen load of an individual urination event is also highly correlated with the time from the previous event, the time of day and the duration of the urination event.

This means that cows which urinate more frequently per day coupled with a lower volume per urination event tend to excrete lower amounts of nitrogen per urination event and thus represent a lower risk to the environment. This is because pasture plants in the urine "patch" (the area of pasture onto which urine is discharged) can utilise a greater proportion of the urine nitrogen for plant growth when urine is deposited at a lower nitrogen application rate. Conversely, for cows that have less urination events per day coupled with a greater volume per urination event, the amount of nitrogen excreted in the urine patch is higher; a greater amount of the nitrogen is unable to be utilised by pasture plants and therefore this nitrogen is vulnerable to being lost to gaseous emissions and/or being leached into waterways.

As there is little variance between individual cows in terms of the urine flow rate, knowing the duration of the urination event allows the volume of the urination event to be calculated based on existing calibration curves from urine sensor studies that indicate that urine volume is proportional to the duration of the urination event.

Even though individual cows may excrete similar total amounts of nitrogen in urine over a period of 24- hours, irrespective of whether they urinate, for example, 12 times a day versus 18 times a day, it is preferable when considering minimising urine nitrogen loading on pasture and/or losses of nitrogen from grazed pasture land, that a dairy herd contain cows that urinate more frequently per day with a lower volume per urination event resulting in lower amounts of nitrogen per urination event.

However, there can be variance in the urination nitrogen load between individual cows and individual urination events for a particular cow. This means that a single urination event should not be relied upon for determining the impact on nitrogen loading by a specific cow. It would be preferable to make this determination based on a series of urination events for greater accuracy.

Being able to measure the frequency and duration of individual urination events over a specified period of time to determine urination frequency, average volume of urine excreted per urination event and thus the amount of nitrogen per urination event, can assist in making appropriate management decisions, both at herd level and for individual cows.

The best way of achieving this is to measure the duration of individual urination events, preferably over a period of time covering at least 72 hours, to determine an average for the individual cow. However, in practice, it is difficult to monitor the time, frequency and duration of individual urination events.

One method may involve manually observing the cow in the field and timing the length of any urination events that may occur. The longer the period of time the cow is observed, the more urination events are observed. Thus, a more accurate average for the duration of the urination event and therefore the volume of urine excreted in that event, can be determined. A more accurate estimate of urination frequency (events/day) can also be obtained.

Cows typically urinate 12 times in a 24-hour period. This means it is not good practice to monitor the animal for a limited period of time, for example across three hours per day where cows would be expected to urinate only one to two times. It would be preferable that any such monitoring take place over a 24 to 72-hour period. While this may mean more accurate determination of the timing of the urination event, and estimation of any diurnal time patterns in frequency and duration of urination events, it is clearly time-consuming and not practical, particularly when dealing with more than one animal.

This approach also requires a clear field of continuous observation; this may be difficult to achieve in practice as the cow of interest is typically in close proximity to other cows of the herd. Of course, the individual cow being monitored could be isolated from the herd, but this may affect its behaviour and lead to changes in patterns, such as frequency of urination events, that could affect the accuracy and therefore potentially bias any determinations of urination parameters for that cow.

An alternative to this would be to use sensors or other equipment placed into or proximate to the urinary tract. This would do away with the need for a person to watch one or more cows for extended periods of time and allow it to retain, as much as is possible, its normal behaviour.

However, care would need to be taken when securing sensors (some of which may require adhesive patches, or the like, to hold the sensing equipment and associated paraphernalia in place). It can also be difficult to apply invasive sensors to the urinary tracts of cows as many will exhibit avoidance behaviours. Placement of invasive sensors for detecting urination events, some having a weight of 500 to 1,500 grams, can take considerable time. Care would need to be taken that the cow, or the person attempting to locate the sensor, is not injured when attaching, inserting or checking the operation of the sensor. This provides an extra cost and constraint to phenotyping larger numbers of animals.

There is also the difficulty of ensuring that the sensors remain in situ for an appropriate period of time. In a typical farm environment, the sensors may become detached, dislodged or malfunction as the animal gets up and down, generally ambulates around the farm or knocks into farm/milking shed, fences, races or posts.

Cows can also generate large amounts of mucus that can block or hinder the operation of the urine sensors. A sustained high-mucus confined environment can lead to cow discomfort and/or infection.

Furthermore, it will be appreciated that cows are exposed to considerable amounts of fluids, in the form of rainfall, condensation, mud, dung and urine, all of which have the potential to affect the properties of adhesives that may be used to secure some types of sensing equipment.

Another method may involve the use of indoor metabolism stalls or similar facilities to house cows. These are constructed with sensors and/or collection devices or the like that are arranged to detect and collect the excretions of the cow. Furthermore, this method is not practical when dealing with large dairy herds, which can comprise hundreds of individual cows. It would entail considerable expense for stalls to be built for all the individuals of the herd, so in practice, only a limited number of stalls would be available.

The farmer would need to select cows to go into these stalls for a period of time, diverting them as appropriate as they transit through the dairy farm. This would then require the farmer to avoid inadvertently selecting a previously tested cow to go into the stalls, which can be difficult when dealing with a large dairy herd. This process would require careful management of the herd and may be potentially very time consuming.

For temperate grazed pasture systems whereby cows graze outdoors, this approach may not be representative of normal grazing behaviours that the cow would experience in its daily life. For example, the cow is typically offered harvested feed in portable bins and animal movements are restricted when in indoor metabolism stalls. Care must then be taken when extrapolating observations derived from such indoor metabolism stalls to outdoor field conditions. In addition, metabolism stalls do not typically provide information on individual urination events; instead they provide aggregate information typically on a daily basis (e.g. total volume of urine excreted per day).

As already noted, dairy cows can excrete similar total amounts of nitrogen in urine on a daily basis but can exhibit large variations between cows in the amount of nitrogen excreted on a per urination event basis. This variation can be harnessed to minimise the environmental impact of grazing livestock.

Object of the Invention

It is an object of the invention to provide an apparatus, system and method for detecting a urination event, its duration and frequency in one or more cows, and thus determining an average volume of urine excreted per urination event and daily urination frequency.

Alternatively, it is an object of the invention to provide an apparatus, system and method for non- invasive measurement of the duration of urination event in one or more cows.

Alternatively, it is an object of the invention to provide an apparatus, system and method for determining the extent of urine volume on land that is derived from the urine of one or more cows.

Alternatively, it is an object of the invention to provide an apparatus, system and method for determining the extent of nitrogen loading on land that is derived from the urine of one or more cows.

Alternatively, it is an object of the invention to provide an apparatus, system and method for a user to manage a cow or a herd of cows to minimise their impact on nitrogen loading on land.

Alternatively, it is an object of the invention to provide an apparatus, system and method for classifying cows for their urine nitrogen environmental impact. Alternatively, it is an object of the invention to at least provide the public with a useful choice.

Summary of the Invention

According to a first aspect of the invention, there is provided a method of detecting a urination event for a cow, the method including the steps of:

1. identifying a first position proximate to the hipline of the spine of the cow and identifying a second position on the spine of the cow a distance between 300 to 500 millimetres forward of the first position;

2. monitoring for changes in the angle of the spine at the first and second positions;

3. measuring the length of time that any changes in the angle of the spine is maintained at the first and second positions; and

4. determining whether the changes in the angle of the spine and the length of time corresponds to criteria defining a urination event.

According to a further aspect of the invention, there is provided a method of detecting a urination event for a cow, the method substantially as described above and including an additional step of:

5. calculating an approximate volume of urine excreted per urination event for the cow.

According to a further aspect of the invention, there is provided a method of detecting a urination event for a cow, the method including the steps of:

1. placing a first orientation sensor on the cow in a first position proximate to the hipline of the spine and placing a second orientation sensor in a second position on the spine of the cow a distance between 300 to 500 millimetres forward of the first orientation sensor;

5. calculating an approximate volume of urine excreted per urination event for the cow. According to a further aspect of the invention, there is provided an apparatus, when used in a method of detecting a urination event for a cow (the method including the steps substantially as described above) wherein the apparatus includes: an orientation sensor, wherein the sensor is configured to be placed on a point of the spine of the cow, and wherein the sensor includes: a data collection device; and a processor, wherein the processor is configured to record and measure data relating to changes in the angle of the spine at the first and second positions and the length of time the angle is maintained.

According to a further aspect of the invention, there is provided a system when used in a method of detecting a urination event for a cow, the method including the steps substantially as described above, and wherein the system includes: a first and second orientation sensors configured to be placed on the cow in a first position proximate to the hipline of the spine and in a second position on the spine a distance between 300 to 500 millimetres forward of the hipline respectively and record changes in the angle of the spine and the duration of said changes, the sensors substantially as described above; and a processor configured to correlate changes in the angle of the spine and the length of time to criteria defining a urination event and determining whether a urination event has occurred for said cow.

The invention is a method, apparatus and system for measuring the duration of urination events in a cow. The invention detects the changes in the angle of the spine of the cow as it urinates, monitors the duration of these changes and from this is able to determine the approximate volume of urine excreted by that cow during the urination event. Over a period of time, an average for urine excreted per urination event can be determined for the cow. This information can be used to assist in management decisions relating to the animal, for example, whether it should be removed from the dairy herd. It can also be used to provide information on the genetic correlations and heritability of urination frequency, time of urination, urination duration, urination volume and nitrogen load per event.

A cow should be understood to be a dairy or beef cow, which are kept and raised for their value as commodities, either through their flesh (as meat) or from their by-products (milk, hides, offal). Reference will now be made throughout the remainder of the present specification to the cow being a dairy cow, i.e. cows kept for their production of milk. A healthy dairy cow urinates on average around 12 times a day, although this is dependent on the diet. Each time a cow urinates should be understood to be a urination event.

During observation trials, in which seven to eight animals were observed from 9am to 3pm over 6 days, it was noted that 122 out of 126 observed urination events had a duration greater than 5 seconds (97%). There was an average of 11.5 urination events per animal per day, with a mean duration of 10.8 seconds, a standard deviation of 4.1 seconds and a minimum/maximum event duration of 2 and 24 seconds, respectively. Urination events of duration less than 5 seconds are not expected to pose a leaching risk (these infrequent events are expected to have a nitrogen load of less than 4 grams per urination event).

Urination events in cows is associated with a characteristic change in body posture. The primary animal movement change is the "arching" or spinal flexion of the spine; there is no movement in the rear legs and the vulva remains at the same height above the ground (i.e. there is no "squatting").

During this arching movement, the cow's spine undergoes a certain degree of flexion at the beginning of the urination event followed by spinal extension at the termination of the urination event. This movement is not subtle; the change in orientation of the spine can be as much as 5° to 25° and thus is visually detectable. By monitoring the duration of this displacement, the length of time of the urination event can be determined.

For the purpose of describing the invention and its use for adult cows, the spine should be understood to have two points of interest. The first point of interest being the hip line (point A), and the second (point B) being at least 400 millimetres forward (i.e. a point closer to the head of the cow) of the hip line.

More precisely, point A is the centre of the hip line (or wing of ilium bone) at the first sacral vertebrae (SI) and point B is at the centre point between the first lumbar vertebrae (LI) and the last thoracic vertebrae (T13). These positions (A and B) undergo a clearly detectable range of movement at the start and conclusion of each urination event.

However, the recited points of interest are not meant to be limiting and other points, such as the vertebrae adjacent SI and T13, may be used as points of interest, although the range of tilt movement associated with the arching behaviour may not be as significant.

It will be appreciated that there may be some minor variations in the position of point B relative to point A due to differences in the length of the torso of individual cows. While in exemplary embodiments of the invention, the anatomical point B will be about 400 millimetres of point A, this is not intended to be limiting. Variations to the specified distance is permitted although generally the anatomical point B will usually be within 300 to 500 millimetres of point A. As noted above, these measurements are for adult, full grown cows. Persons skilled in the art will appreciate that if juvenile cows are being measured, the distance between points A and B will be proportionally less.

Reference to two points of interest is not meant to be limiting; for greater accuracy and/or redundancy, additional points of interest could be selected. However, as noted above, points A and B are preferred since these correspond to the points of the spine in the cow that undergo the greatest range of movement at the start and end of the urination event.

In particular, at point A, the hip line of the spine, when the urination event is initiated or begins, the spine tilts upwards by between 8° to 22.5°. Simultaneously, or at least substantially at the same time, at point B, a point approximately 400 millimetres forward of the hip line, the spine tilts downward by between 8° to 22.5°.

This tilting or displacement of the spine at these points is maintained throughout the duration of the urination event.

Then, at the termination of the urination event, i.e. when it ends, the spine tilts downward at point A by between 8° to 22.5°. Correspondingly, the spine tilts upward at point B by between 8° to 22.5°. Thus, the spine of the cow returns to its normal posture. It should be noted that reference to change in tilt refers to change in pitch in the direction of the spine toward the head of the cow.

A variety of models may be utilised in order to classify whether the event was an act of urination and to predict the duration of the urination event.

Thus, in one exemplary embodiment of the invention, the criteria for one model for determining whether a urination event has occurred are:

1. An increase in tilt of at least 8° at the centre of the spine at the hipline;

2. A corresponding decrease in tilt of at least 8° at the centre of the spine at a location approximately 400 millimetres in front of the hipline;

3. A specified period of relative inactivity during which the increase/decrease in tilt angle is maintained with a sustained difference in tilt angle of at least 16° and at least one time during this period, with a difference in tilt angle of at least 17°;

4. A decrease in tilt of at least 8° at the centre of the spine at the hipline; and

5. A corresponding increase in tilt of at least 8° at the centre of the spine at a location approximately 400 millimetres in front of the hipline. An additional criterion, to reduce the risk of false positives caused by other behaviours, may include a limitation that the period of "relative inactivity during which the increase/decrease in tilt angle is maintained" must be at least five seconds or more before the changes in tilt angle of the spine can be positively determined as being a urination event.

These changes in tilt angle in the spine, when held for a period of time of at least five seconds, are strongly indicative of a urination event. Cows generally do not exhibit other behaviours that correspond to this movement. Therefore, being able to detect these changes in the tilt angle of the spine and their duration, allows the frequency of urination events, duration of the urination event and the volume of urine excreted per event, to be monitored.

It should be appreciated that the above recited criteria may be modified or otherwise adapted depending on the model to be used.

Urination volume is proportional to the duration of the urination event, with an average flow rate of 4 litres per minute. The nitrogen load of an individual event is also highly correlated with the time from the previous event, the time of day and the duration of the urination event.

The change in tilt angle of the spine may be detected in a number of ways.

In a first exemplary embodiment, the present method involves the use of an apparatus in the form of an orientation sensor. This should be understood to be a sensor that measures the orientation of the spine, and in particular any changes in respect to the tilt of the spine.

In the first exemplary embodiment of the present invention, the orientation sensor is an accelerometer and shall be referred to as such throughout the remainder of the specification. However, this is not meant to be limiting. Depending on circumstances, other orientation sensors such as magnetometers and gyroscopes can be used instead of accelerometers, or they may also be used in conjunction with accelerometers to provide information about changes in the tilt angles of the spine of the animal.

Accelerometers are preferred, due to being relatively inexpensive when compared to other types of orientation sensors such as magnetometers or gyroscopes. Thus, accelerometers are more cost effective to use with a larger group of cows at any one time.

It should also be appreciated that the extent of change in tilt angle of the spine is such that a high degree of accuracy is not essential; the accelerometer simply needs to be capable of detecting a change of about 8° or more. This also contributes to keeping the expense of the present invention to a minimum as a more basic type/model of accelerometer may be used. However, this is not intended to be limiting and more advanced accelerometers may be used. For example, tri-axis accelerometers can provide more accurate information on the change in tilt of the spine during urination. In exemplary embodiments of the invention, the accelerometer includes a housing for its various components. Given its likely exposure to potentially harsh inclement weather, it is preferable that the housing be watertight.

The housing may include a base portion, which in use, is the surface of the accelerometer that is placed in contact with the cow. The base portion may include an attachment surface to which a bonding agent, such as high-strength adhesive, can be applied. Alternatively, the base portion may have a previously prepared adhesive layer which is activated prior to being attached to the cow by removing a shielding strip.

In some embodiments, the user attaching the accelerometer may prepare the area to which it is to be attached by shaving the hair at the desired location. This allows the bonding of the accelerometer to be against at least a portion of the skin of the cow rather than its hair. Alternatively, each accelerometer may be integrated into a strap which the user attaches to the cow by wrapping it around the body, taking care that the accelerometer is positioned upper most on the spine.

As noted above, changes in tilt angle of the spine may be detected in a number of ways. In a second exemplary embodiment of the invention, the present method uses images of the cow and image processing to ascertain the degree of tilt along the spine.

In this second exemplary embodiment, the present method involves the use of an imaging apparatus in the form of a camera.

A camera should be understood to encompass conventional digital cameras, 3D cameras, stereo vision, structure from motion, laser scanner, time of flight cameras and LiDAR. These may all be used to provide information on the change in tilt of the spine associated with urination and fall within the scope of the present invention. Persons skilled in the art will readily identify cameras that would be suitable for use in the present invention, bearing in mind location and physical constraints.

In exemplary embodiments of the invention, the camera is attached directly to the cow being monitored. This allows for continuous and unobstructed observation over the monitoring period.

Cameras may be used remotely from the cow; for example, a camera may be mounted to a post near the grazing area of the cow being monitored. However, unless the cow is closely confined, the camera may need to have appropriate resolution and field of vision to capture urination events if they occur relatively distant from the camera position. Alternatively, a drone equipped with imaging apparatus could be utilised. The imaging apparatus/camera includes a housing for its various components. Given the likely sensitivity of the camera to water and fluid, it is desirable that the housing be watertight. However, the housing would still need to allow for an appropriate field of vision.

In exemplary embodiments, the housing of the camera may be provided with a base portion to allow it to be attached to the cow with adhesive or the like. As with the accelerometer, this may require the user to prepare the site of attachment by shaving it to reduce or eliminate any hair that may be present.

However, in some embodiments, the housing of the imaging apparatus may be configured to receive a strap, harness or the like. The strap may engage directly with the housing or alternatively, may engage with a base plate or similar structure to which the housing is attached in a snap-lock type arrangement.

The use of a strap in this embodiment makes it ideal for placing the imaging apparatus on the neck or upper back of the cow, facing rearwards, and may be quicker to install.

The accelerometer or imaging apparatus should be understood to include a processor, such as a programmable logic controller (PLC) or processor.

In exemplary embodiments of the invention, the processor is configured to measure and record data relating to changes in the angle of the spine and the duration of said changes.

In exemplary embodiments of the invention, the processor is configured to analyse the data and determine whether changes in the angle of the spine and the duration of said changes should be correlated with the criteria set out above defining a urination event and be identified as such. However, it should be appreciated that in some embodiments of the invention, the analysis of the measured and recorded data may be performed by a central processing station or computer.

The accelerometer or imaging apparatus should be understood to include a data collection device to collect tilt angle data over a period of time and store it for later retrieval and analysis.

In exemplary embodiments of the present invention, the data collection device is a hard drive or memory device connected or otherwise linked to the processor.

In exemplary embodiments of the present invention, the accelerometer or imaging apparatus includes a power source in the form of a battery.

For ease and lack of expense, lithium button cell batteries are preferred for the accelerometer, but this is not meant to be limiting. The batteries may be larger (or smaller), depending on the size of the housing of the accelerometer and the period of time the cow is to be monitored. Smaller batteries are preferred since this allows the size of the overall housing to be kept to a minimum and be less obtrusive to the animal. For the imaging apparatus, which may have higher power consumption, the battery may be a AAA, AA or 9-volt battery. Again, smaller batteries are preferred so that the overall housing is kept as small as possible.

In alternative embodiments, as the back, and thus the spine of the cow, is exposed to sunlight for several hours a day, a solar cell may be used as a source of power for the accelerometer or imaging apparatus. This may be particularly useful if the measurements are to be performed during the summer months when there is a greater likelihood of extended periods of sunlight.

In some embodiments of the invention, the accelerometer or imaging apparatus may include a means for wirelessly transmitting tilt angle data, either in real time or at regular intervals (e.g. as the cow transits a milking shed or another point of interest on the farm) to a remote processing station for analysis. However, it will be appreciated that this may place additional demands on available power for the accelerometer or imaging apparatus.

In some embodiments of the invention, the accelerometer or imaging apparatus may include a GPS unit for tracking functionality. In alternative embodiments, the accelerometer or imaging apparatus may be communicative with a separate GPS unit which may be carried by the animal in an ear tag, collar or the like.

Having this GPS tracking functionality may allow the user to correlate the timing of a urination event with a specific location of the farm, both on an individual and herd level. This data may be useful in managing the movements of a cow, or herd, about the farm in order to minimise the environmental impact of nitrogen excreted in urine.

When the accelerometer embodiment of the present invention is in use, the user will attach at least a pair of accelerometers to the spine of the cow to be monitored; one at the hipline and one about 400 millimetres forward of the hipline. If desired, additional accelerometers can be placed at regular intervals along the spine.

Once activated, the accelerometers will measure and record changes in the tilt angle of the spine, and the length of these changes. Over time, this would build up a profile of individual urination events for the animal.

After a desired period of time, typically 72-hours, the user will retrieve the accelerometer. The collected data can be retrieved and processed through a central processing station, such as a computer. From this data, an average volume of urine excreted per urination event can be determined for the cow.

The simplicity of attachment and removal of the accelerometers allows the user to more easily assess a group of cows. Furthermore, the non-invasive manner of attachment and removal is less stressful for the animals being assessed. Being relatively inexpensive, a stockpile of accelerometers may be built up and reused on a regular basis for groups or herds of cows.

Alternatively, when the imaging embodiment of the present invention is in use, the user will attach the imaging apparatus to the cow, orientating it so that the back and rear quarters of the animal is within the field of vision. In particular, a first point at the hipline and a second point about 400 millimetres forward of the hipline should be within the field of vision, although other locations along the spine within the field of vision can also provide further information on the arching behaviour. Once activated, the imaging apparatus will monitor changes in the tilt angle of the spine, as determined by the difference in relative heights between the areas of interest at the start, middle and end of the urination event, together with the duration of these changes. Over time, this would build up a profile of individual urination events for the animal.

After a desired period of time, typically 72-hours, the user will retrieve the imaging apparatus. The collected data can be retrieved and processed through a central processing station, such as a computer. From this data an average volume of urine excreted per urination event can be determined for the cow.

The information derived from using the present invention can be used in a variety of ways. For example, it may be used for selecting breeding stock that preferentially exhibit the desirable phenotype of a low amount of nitrogen per urination event. It also has value as a tool for regulatory compliance relating to nitrogen loss from farms.

Further aspects of the invention, which should be considered in all its novel aspects, will become apparent to those skilled in the art upon reading of the following description which provides at least one example of a practical application of the invention.

Brief Description of the Drawings

One or more embodiments of the invention will be described below by way of example only, and without intending to be limiting, with reference to the following drawings, in which:

Figure la is a schematic of a cow showing the extent of the approximate curvature of the spine when standing and (in dashed lines) when urinating;

Figure lb is a schematic of a cow showing approximate placement of the accelerometer of one embodiment of the present invention;

Figure 2 is a schematic of the accelerometer housing;

Figure 3a is a graph of the relationship between the measured and predicted duration of urination events for the 21 predicted urination events of a calibration trial following application of Model A; Figure 3b is a graph of the relationship between the measured and predicted duration of urination events for the 129 predicted urination events of a validation trial following application of Model A;

Figure 4a is a graph of changes in tilt angle of the spine at the hipline of a cow over a first period of time;

Figure 4b is a graph of changes in tilt angle of the spine at a point 400 millimetres forward of the hip line of the cow of Figure 3a, over the same period of time shown in Figure 3a;

Figure 5a is a graph of changes in tilt angle of the spine at the hipline of the cow of Figures 3a and

3b over a second period of time;

Figure 5b is a graph of changes in tilt angle of the spine at a point 400 millimetres forward of the hip line of the cow of Figure 4a, over the same period of time shown in Figure 4a;

Figure 6 is a schematic of the imaging apparatus of an alternative embodiment of the present invention; and

Figure 7 is a contour map of the back region of a cow based on data collected from the imaging apparatus of Figure 5.

Detailed Description of Preferred Embodiments of the Invention

In Figure la, a cow (generally indicated by arrow 100) is illustrated; the hip line point of the spine is identified as A. Located further along the spine, closer to the head (H) of the cow is a second point of interest (B).

Point A is the centre of the hip line (or wing of ilium bone) at the first sacral vertebrae (SI) and point B is at the centre point between the first lumbar vertebrae (LI) and the last thoracic vertebrae (T13). These two points correspond to points of the spine that undergo a distinct tilting at the start and end of a urination event. This causes an overall change or displacement in the general orientation of the spine, as indicated by the dashed lines, for the duration of the urination event. The inventors have determined that monitoring for changes in the angle of the spine at these points allow a cow's urination events, and in particular their duration, to be monitored.

In one embodiment of the invention, this may be achieved through the placement of orientation sensors at these points along the spine.

Figure lb, showing the cow (100) in a top and side view, illustrates the placement of the sensors (102, 104, 106, 108). In this view, four orientation sensors, in the form of accelerometers, are located along the spine, the first (102) being positioned 200 millimetres behind the hipline and then every 200 millimetres along the spine towards the head (H).

Sensor 102 is placed at the fifth sacral vertebrae. Sensor 104 is positioned at the first sacral vertebrae and subsequent sensors are located at the fourth lumbar vertebrae (sensor 106) and the point between the first lumbar vertebrae (LI) and the last thoracic vertebrae (T13) (sensor 108) respectively. Those skilled in the art will appreciate that orientation sensors can be placed at any vertebrae on the sacral, lumbar, thoracic and cervical parts of the spine however, it is preferable to place them on the vertebrae that undergo the most change in tilt angle during the urination event, for ease of detection and greater accuracy.

Although four sensors are shown in this view, it is not essential; in some embodiments only two or three sensors may be used. It will be appreciated that this means the spacing between adjacent sensors may need to be adjusted accordingly.

More than four sensors may be employed to attain more accurate predictions of urination events. These sensors can be located on any sacral, lumbar, thoracic and cervical vertebrae. However, it should be appreciated that orientation sensors located on the tail (coccygeal vertebrae)(T) do not provide reliable information on urination events; tail raising behaviour is also associated with dung events and insect avoidance.

The sensors may be standard, off-the-shelf, accelerometers. The inventors used the Hobo Pendant^® G data logger, manufactured by Onset^® Instruments (www.onsetcomp.com). This accelerometer is able to measure three-dimensional movement, including tilt, in up to three axes and records up to seven days of data in 0.01 second intervals. It is powered by a CR2032 lithium button cell battery and is particularly useful due to its relatively low weight of about 18 grams making it fairly unobtrusive to the cow when in position.

The Hobo Pendant^® G data logger has a measurement range of ± 3 g; 29.4 m/s² and accuracy of ± 0.075 g; 0.735 m/s² at 25°C (± 0.105 g; 1.03 m/s² from -20°C to 70°C). Resolution is at 0.025 g; 0.245 m/s². Stored data is extracted with a USB interface cable which can be connected to a central processing station, such as a computer or handheld unit running appropriate software.

Illustrated in a side view in Figure 2, the accelerometer (200) includes a housing (202) of polypropylene to ensure that its components are kept watertight. Being made of a plastic material helps keep the overall weight of the accelerometer to a minimum. However, other materials such as aluminium may be used instead. The base (204) of the housing (202) is suitable for expeditious attachment to the cow. The user may apply high-strength adhesive (not shown) to the underside (206) of the base of the housing and the point of the spine (not shown) at which it is to be positioned. This may require the user to prepare the location point on the spine by shaving it to reduce or eliminate hair and ensure a good adhesive bond.

Regardless of the method of securing the accelerometer to the cow, it is desirable that it remain in place for at least 48 to 72 hours. This is a sufficient period of time to monitor the number and duration of urination events and determine a reasonably accurate average for the cow. However, this is not meant to be limiting; in some cases, it may be acceptable to determine a urination frequency/duration average based on data collected over a 24-hour period or over a period greater than 72-hours depending on the desired accuracy.

Regardless of the length of time that the accelerometer is in place, it must have a source of power for the processor measuring and recording the tilt angle data. This power source may be in the form of a battery such as a lithium cell button (not shown). The accelerometer also includes a hard drive or memory device (not shown) to store the collected data for later retrieval and analysis.

The placement of the sensors along the spine is important as they capture the extent of the tilt or angling of the spine that occurs during the urination event. As previously discussed, a urination event must meet five criteria in order to be identified as such. In one example, referred to as Model A, these criteria are:

3. A period of relative inactivity during which the increase/decrease in tilt angle is maintained for a period of 5 consecutive seconds, with a sustained difference in tilt angle of at least 16° and at least one time during this period, with a difference in tilt angle of at least 17°;

5. A corresponding increase in tilt of at least 8° at the centre of the spine at a location approximately 400 millimetres in front of the hipline.

An algorithm can be developed based on these parameters and applied by a processor to tilt angle data collected from a cow fitted with accelerometer sensors at appropriate points along its spine.

For greater accuracy, the algorithm can include an additional constraint to limit false positives that may occur. For example, cows are known to use one of their rear hooves to scratch their bodies, particularly the sides of necks and heads to alleviate an irritation (such as an insect bite or the like). Such movement may cause changes in the orientation of the spine that fall within the criteria outlined above. In Model A above, criteria 3 stipulates that the period of relative inactivity, that follows the initial tilting of the spine, must be five consecutive seconds or more. In alternative models, this period of relative inactivity may be increased or decreased if desired.

Arching of the spine also occurs prior to a large rear kicking event which may involve one or both rear legs. This arching event is very rapid and is complete in less than one second. Large back arching events of very short duration can also occur during rapid locomotion and these events are excluded based on the criteria 1 to 5 above.

Furthermore, additional accelerometers may be used to assist in improving accuracy and eliminating false positives. For example, accelerometers could be located on the rear legs to detect the characteristic movements associated with scratching or kicking events.

Persons skilled in the art will appreciate that while there may also be flexing of the spine at points A and B as the cow lowers and raises itself from the ground, these events will typically fall outside the criteria defining a urination event as there is no period of relative inactivity during which the increase/decrease in tilt angle is maintained.

It should be appreciated that further features of the changes in spine orientation during a urination event may be used to develop an algorithm for use in the present invention. For example, one model may employ Linear Discriminant Analysis (LDA) based on 18 selected features.

The 18 selected features, calculated over time intervals where the difference in tilt angle is greater than 0°, are:

1. The maximum difference in tilt angle;

2. The number of seconds the difference in tilt angle is greater than 0°;

3. The number of seconds the difference in tilt angle is greater than 8°;

4. The standard deviation in the difference in tilt angle;

5. The standard deviation in the difference in tilt angle restricted to periods when the difference in tilt angle is greater than 15°;

6. The standard deviation in the difference in tilt angle restricted to periods when the difference in tilt angle is greater than 20°;

7. The standard deviation in the difference in tilt angle restricted to periods when the difference in tilt angle is greater than 25°;

8. The maximum number of consecutive seconds with a difference in tilt angle of at least 16°;

9. The maximum number of consecutive seconds with a difference in tilt angle of at least 10°;

10. The maximum number of consecutive seconds with a difference in tilt angle of at least 8°; 11. The number of seconds to reach the maximum difference in tilt angle;

12. The number of seconds the rate of change in tilt is greater than 2.75 degrees per second;

13. The number of seconds the rate of change in tilt is greater than 10% of Feature 1 (above) degrees per second;

14. Parameter bi from the fit of Equation 1 to the difference in tilt angle over the time interval where the difference in tilt angle is greater than 0°;

15. Parameter b₂ from the fit of Equation 1 to the difference in tilt angle over the time interval where the difference in tilt angle is greater than 0°;

16. Parameter b from the fit of Equation 1 to the difference in tilt angle over the time interval where the difference in tilt angle is greater than 0°;

17. Parameter b -bi-b from the fit of Equation 1 to the difference in tilt angle over the time interval where the difference in tilt angle is greater than 0°;

18. The number of seconds the difference in tilt angle is greater than 80% of Feature 1 (above); and where the difference in tilt angle is characterised by Equation 1 as follows:

Other classifications may employ these 18 features, or at least the majority of these 18 features; for example, Naive Bayes (NB), Support Vector Machine (SVM), Classification Tree (CT) and k-Nearest Neighbour (kNN) classifications may be used for all 18 features while a Stepwise Generalised Linear Model (SGLM) could use 16 features (the two excluded features are the maximum number of consecutive seconds with a difference in tilt angle of at least 10° and 16° respectively; features 8 and 9 in the list above).

A calibration trial was conducted with four non-lactating dairy cows over a two-day measurement period. The cows grazed predominantly perennial ryegrass/white clover pastures and had free access to drinking water throughout the study period. The cows were visually observed for individual urination events by trained technical research staff in conjunction with video recording devices positioned around the grazed area to record the time and duration of urination and dung events from individual cows over a 6-hour measurement period (09:00 to 15:00). A total of 24 urination events and 24 dung events were recorded from the 4 cows over the two-day trial period. The duration of the sufficient back arching event in the calibration trials was used as a predictor of the duration of the urination event using Model A as defined above. A comparison of observed and predicted classification of urination events in the calibration trial is tabulated below in Table 1 (TP - true positive, FN - false negative, FP - false positive, TN - true negative).

Table 1

¹ Of the 24 predicted but unobserved urination events 18 were observed to be dung events and the remaining 6 were scratching events.

² This is the number of 15 seconds intervals where no event was observed and no event was predicted to occur (there are 1440 15 second intervals per 6 hour period per cow). Urination events are approximately 15 seconds in duration.

³ The FI score is the harmonic mean of Sensitivity and Precision and provides a balanced measure of binary classification.

The relationship between the measured and predicted duration of urine events for the 21 predicted urination events in the calibration trial 1 with Model A (R² = 0.29, RMSE = 4.7 seconds, N = 21, P = 0.01, slope = 0.77 ± 0.28, intercept = 0.26 ± 4.91, bias = 3.65 seconds) is graphically represented in Figure 3a.

Other models may be used to predict the classification of urination events. These may amend the existing criteria; for example, Model B may be identical to Model A other than criteria 3 being amended such that to qualify as a urination event, the back arching must be sustained for at least 10 seconds. In addition, further criteria could be added. For example, Model B may include an additional criteria where the number of seconds that the rate of change in tilt is greater than 2.75 degrees per second has to be less than 5 seconds to qualify as a urination event. A comparison of observed and predicted, using Model B, classification of urination events from the calibration trial is tabulated below in Table 2.

Table 2

¹ The 2 predicted but unobserved urination events were observed to be dung events.

A validation trial was conducted with 15 non-lactating dairy cows over a 4-day period using the same methodology as described for the calibration trial. Experimental cows were visually assessed for time and duration of urination events over a 6-hour measurement period (09:00 to 15:00). A total of 135 urination events were observed from the 15 cows over the four-day trial period.

Using Models A and B, comparison of observed and predicted classification of urination events from the validation trial is tabulated in Tables 3 and 4 respectively.

Table 3

¹ This is the approximate number of 15 seconds intervals where no event was observed and no event was predicted to occur (there are 1440 15 second intervals per 6 hour period per cow). Urination events are approximately 15 seconds in duration.

Figure 3b graphically represents the relationship between the measured and predicted duration of urine events (model A) for the 129 predicted urination events with measured event duration of the validation trial for Days 1 to 4 (R² = 0.16, RMSE = 3.0 seconds, N = 129, P < 0.001, slope = 0.29 ± 0.06, intercept = 5.11 ± 0.88, bias = 4.62 seconds).

Table 4

An example in practice of the implementation of the method, apparatus and system can be seen in Figures 4a and 4b; these respectively show in one cow of the calibration trial, the measured tilt angle of the spine at the hip line (point A in Figure la) in degrees and at a point 400 millimetres from the hip line (point B in Figure la) over a period of time from approximately 10:02 to 10:04 (in these examples, time will be expressed as hounminutes. seconds).

From this it will be seen that at approximately 10:02.35 the accelerometer at point A underwent a significant increase in tilt from an initial 72° to around 86° to 88°.

This tilt was maintained for 12 seconds, in accordance with criteria 1 to 5 listed above, during which the cow urinated. This was correlated with visual observation of the urination event estimated to have a 14 second duration.

Following initiation of the urination event, the relative tilt of the spine is maintained, until at about 10:02.47, when urination ceased and the spine started to return to its normal orientation. Thus, the change in tilt of the spine of the cow as it begins and ends the urination event determines the length of the urination. From this, the approximate volume of urine excreted during the urination event can be calculated.

Over the period from 10:03.20 to 10:03.40 small changes in tilt angle are visible at both locations on the spine. However, these are not synchronized, large in magnitude or sustained in duration and are associated with normal animal movement. These changes are not classified as a urination event, not meeting the criteria 1 to 5 of Model A set out above.

A further example is shown in Figures 5a and 5b. These show a time period in the afternoon for the cow of Figures 3a and 3b. It will be seen that a peak was observed from about 13:05.10 to 13:05.30. As this fulfils the criteria for a urination event, this can be positively correlated to a urination event lasting about 19 seconds. This was confirmed by visual observation of the urination event, estimated as being of 20 seconds duration.

The technology predicts that this cow (in Figures 4a, 4b, 5a, 5b) urinates twice over a six-hour period from 09:00 to 15:00, which is consistent with a visual observation of two urination events by this cow during the same time period.

Determination of the average volume of urine excreted per urination event for a particular cow and daily urination frequency can then be used to help make management decisions in relation to that cow. For example, cows that urinate more frequently coupled with a low volume per urination event (or the sires of these cows) may be preferentially selected for breeding as such animals may have a lower impact on nitrogen loading on pasture. Alternatively, they may be retained in the herd while cows that urinate less frequently are culled or separated from the herd into groups that would graze pasture that are remote from water ways.

An alternative embodiment of the invention uses imaging as a means to determine changes in the tilt angle of the spine of the cow during a urination event. This calls for the use of imaging apparatus such as a camera which is preferably attached to the cow. An example of a camera (500) used in the present method is illustrated in Figure 6.

The camera (500) is an Occipital StructurelO™ 3D camera which utilises structured light to obtain a depth image. Its housing is relatively unobtrusive, being 119 millimetres in length with a height of 29 millimetres. The camera has an overall weight of 95 grams.

The camera lens (502) has a horizontal field of view of 58 degrees, a vertical field of view of 45 degrees and a framerate of 30 to 60 frames per second. To save power usage and maximise data storage, the user may opt to select a frame rate of 30 frames per second. A USB port (504) allows for the retrieval of the image data at the end of the monitoring period.

The data extracted from the images can be used to produce a digital elevation map (DEM) for the surface of the back of the cow. In particular, this DEM can be used to generate a contour map of the spine and surrounding regions of the cow. An example of such a map is illustrated in Figure 7.

The contour map defines the peaks, ridges, saddles and valleys of the upper surface of the back of a cow as per a traditional contour map. The spine of the cow is represented by the dotted line running from left to right across the graph. The small circle denotes the position of the spine at the edge of the field of view. The camera is at a fixed location 600 millimetres above the animal. The local peak of the contour surface at the centre of the animal defines the location where the sacral spine meets the tail (represented by the asterisk at the extreme left end of the dotted line).

The steepest ascent/descent method and the discrete Laplace transform can be applied to the image data to locate the region of the spine and then specify its location in three-dimensional space. The processor, whether in the camera or in a central processing station such as a computer, will perform the necessary analysis and transform work.

In particular, regions of the spine relative to the hip bone or sacral/tail junction can be determined and the changes in local tilt or slope can be determined in these regions as they change during a urination event. The duration of these changes provides an approximate correlation with the length of the urination event, subject to the criteria previously set out above with respect to the accelerometer embodiment, and from this the volume of urine excreted can be calculated. Over time, this will lead to development of a profile for average volume of urine excreted per urination event.

Unless the context clearly requires otherwise, throughout the description and the claims, the words "comprise", "comprising", and the like, are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense, that is to say, in the sense of "including, but not limited to".

The entire disclosures of all applications, patents and publications cited above, if any, are herein incorporated by reference.

Reference to any prior art in this specification is not, and should not be taken as, an acknowledgement or any form of suggestion that that prior art forms part of the common general knowledge in the field of endeavour in any country in the world.

The invention may also be said broadly to consist in the parts, elements and features referred to or indicated in the specification of the application, individually or collectively, in any or all combinations of two or more of said parts, elements or features. Where in the foregoing description reference has been made to integers or components having known equivalents thereof, those integers are herein incorporated as if individually set forth.

It should be noted that various changes and modifications to the presently preferred embodiments described herein will be apparent to those skilled in the art. Such changes and modifications may be made without departing from the spirit and scope of the invention as claimed in the appended claims. It is therefore intended that such changes and modifications be included within the present invention.

Claims

1. A method of detecting a urination event for a cow, the method including the steps of: a. identifying a first position proximate to the hipline of the spine of the cow and identifying a second position on the spine of the cow a distance between 300 to 500 millimetres forward of the first position; b. monitoring for changes in the angle of the spine at the first and second positions; c. measuring the length of time that any changes in the angle of the spine is maintained at the first and second positions; and d. determining whether the changes in the angle of the spine and the length of time corresponds to criteria defining a urination event.

2. The method as claimed in claim 1, wherein the method includes an additional step of: e. calculating an approximate volume of urine excreted per urination event for the cow.

3. A method of detecting a urination event for a cow, the method including the steps of: a. placing a first orientation sensor on the cow in a first position proximate to the hipline of the spine and placing a second orientation sensor in a second position on the spine of the cow a distance between 300 to 500 millimetres forward of the first orientation sensor; b. monitoring for changes in the angle of the spine at the first and second positions; c. measuring the length of time that any changes in the angle of the spine is maintained at the first and second positions; and d. determining whether the changes in the angle of the spine and the length of time corresponds to criteria defining a urination event.

4. The method as claimed in claim 3, wherein the method includes an additional step of: e. calculating an approximate volume of urine excreted per urination event for the cow.

5. The method as claimed as in any one of claims 1 to 4, wherein the second position on the spine of the cow is 400 millimetres forward of the second position.

6. The method as claimed in any one of claim 1 to 5, wherein the criteria defining a urination event are: a. An increase in tilt of at least 8° at the centre of the spine at the hipline; b. A corresponding decrease in tilt of at least 8° at the centre of the spine at a location approximately 400 millimetres in front of the hipline; c. A specified period of relative inactivity during which the increase/decrease in tilt angle is maintained with a sustained difference in tilt angle of at least 16° and at least one time during this period, with a difference in tilt angle of at least 17°; d. A decrease in tilt of at least 8° at the centre of the spine at the hipline; and e. A corresponding increase in tilt of at least 8° at the centre of the spine at a location approximately 400 millimetres in front of the hipline.

7. The method as claimed in claim 6, wherein in criteria c., the period of "relative inactivity during which the increase/decrease in tilt angle is maintained" is at least five seconds or more.

8. An apparatus, when used in a method of detecting a urination event for a cow as claimed in claim 3, wherein the apparatus includes: an orientation sensor, wherein the sensor is configured to be placed on a point of the spine of the cow, and wherein the sensor includes: a data collection device; and a processor, wherein the processor is configured to record and measure data relating to changes in the angle of the spine at the first and second positions and the length of time the angle is maintained.

9. The apparatus as claimed in claim 8, wherein the orientation sensor is an accelerometer.

10. The apparatus as claimed in claim 8, wherein the orientation sensor is an imaging apparatus.

11. The apparatus as claimed in any one of claims 8 to 10, wherein the data collection device is a hard drive or memory device.

12. The apparatus as claimed in any one of claims 8 to 11, wherein the orientation sensor includes a power source, wherein the power source is a battery.

13. The apparatus as claimed in any one of claims 8 to 12, wherein the orientation sensor includes a wireless transmitter.

14. The apparatus as claimed in any one of claims 8 to 13, wherein the orientation sensor includes a GPS unit.

15. The apparatus as claimed in any one of claims 8 to 14, wherein the orientation sensor includes a housing.

16. The apparatus as claimed in claim 15, wherein the housing includes a base portion, wherein the base portion includes an attachment surface.

17. A system, when used in a method of detecting a urination event for a cow as claimed in claim 3, wherein the system includes the apparatus as claimed in any one of claims 8 to 18, the system also including a central processing station.

18. A system, when used in a method of detecting a urination event for a cow as claimed in claim 1, wherein the system includes an imaging apparatus and a central processing station.

19. The system as claimed in claim 18, wherein the imaging apparatus is mounted to a post or drone.

20. The system as claimed in any one of claims 18 to 19, wherein the imaging apparatus is a camera.