NZ547497A - A measurement and monitoring system - Google Patents

Abstract

A method for determining dairy stock and plant performance is disclosed. The method comprises the steps of: (a) making performance measurements using a measurement unit during a normal milking session; (b) collecting measurements to incrementally develop performance profiles during a normal milking session; (c) storing performance profiles in memory as a completed set at the end of a normal milking session; (d) using a processor, applying a best fit matching of a stored set to the current milking; (e) providing settings for user or pre-determined standards of performance; and (f) using a processor, providing performance assessment for measurements during the current milking session using assessment criteria that are in part determined by stored profiles.

Description

54 7 4 97 *10051205870* PATENTS FORM NO. 5 Our ref: JT226173NZPR NEW ZEALAND PATENTS ACT 1953 COMPLETE SPECIFICATION Divisional out of New Zealand Patent Application No. 537013 A MEASUREMENT AND MONITORING SYSTEM We, Radian Technology Limited a New Zealand Company of 1 Coolidge Street, Brooklyn, Wellington, New Zealand hereby declare the invention, for which we pray that a patent may be granted to us and the method by which it is to be performed, to be particularly described in and by the following statement: 100797053_1 .DOC: JT:gbo "1" | (Followed by page 1 a) J intellectual property office 2 6 MAY 2006 .RECEIVED -1a- A MEASUREMENT AND MONITORING SYSTEM FIELD OF INVENTION This invention relates to in-line measurements for quality control and management of dairy stock and plant.

BACKGROUND Approaches in the field can be divided into systems for mastitis detection and systems for volume yield.

Elevated ion levels in the harvested milk are symptomatic of mastitis. Measurement of the resulting increase in conductivity has formed the basis for 15 many mastitis detection systems but practical issues have detracted from useful deployment in a farm context.

A common method to measure conductivity is by positioning electrodes in the solution through the wall of a containing vessel.

A simple implementation of this approach is a hand held conductivity meter. This requires sample collection by the manual stripping of milk directly from the cow. Manual stripping usually is resisted by the cow and becomes a time consuming process poorly suited to high production milking.

Improvements have been proposed by incorporating electrodes directly into the claw of the milking apparatus. This has the potential of facilitating automatic inline detection with improved reliability using individual quarter conductivity measurements for comparative analysis. Obviously this approach dictates the 30 use of a specialist claw that is generally larger than the traditional arrangement. Both the lack of flexibility in choice of claw and the size are undesirable. In addition, sensor wiring and location on the harsh parlour floor environment and around animal hooves represents practical reliability issues due to incidents of breakage, water damage or wiring failure. 2 Some conductivity measuring systems have been placed at the top of the long milk tube above the parlour floor. These systems have been targeted at detecting the presence or absence of solution to determine when a cow is finished milking.

Accurate measurement for the purposes of mastitis detection has not yet been made practical with this approach.

Approaches with electrodes in solution are subject to electrode fouling or poisoning. This is due to the build up of coatings with poor solubility, preferential 10 plating of metal ions over time or the effect of cleaning agents used in the milking process. The result is calibration drift and measurement inaccuracy that can only be rectified with time consuming regular maintenance or replacement.

To overcome the difficulties associated with electrodes in direct contact with 15 solution, some systems have been proposed with sensing arrangements on the outside of a plastic wall containing the solution. Such arrangements necessarily use high frequency fields since plastic blocks direct or low frequency fields. In proposals to date that use fields of this type, both the effect of the containment wall and the dielectric behaviour of the solution dominate any measurement 20 result and overshadow any small effect due to solution conductivity. As a consequence, while sensing through a plastic containment wall is adequate in detecting the presence or absence of solution, it has not been able to measure conductivity to the accuracy required for mastitis detection in the practical milking situation.

A compounding issue for automatic in-line sensing is the mixed air and solution nature of the flow. The presence of uncertain amounts of air in the solution results in uncertainty in bulk measurements such as conductivity. Systems have been proposed that employ mechanical sampling arrangements that allow for the 30 solution to settle as discrete samples. These systems are complex and often involve moving parts that reduce reliability and increase cost. Many automatic sampling arrangements also require regular cleaning to ensure hygiene levels are maintained. 3 Approaches for in-line measurement of volume yield also depend on separating solution from air. With most arrangements separated solution is accumulated in a sampling reservoir. One arrangement uses a reservoir that samples a known proportion of solution by splitting a solution jet stream. Yield is determined from 5 total sample volume and the sample is either returned to the bulk solution or discarded. Other proposals involve counting of smaller reservoir samples as they continuously fill and empty. In both cases manual and automatic variations have been suggested.

As with comparable sampling systems for conductivity measurement, arrangements for yield measurement are complex and often involve moving parts that reduce reliability and increase cost. Here again they require regular cleaning to ensure hygiene levels are maintained.

Solutions integrated into management information systems have also been proposed. These systems typically include information in addition to conductivity and yield. They involve databases and provide for the manipulation of management information through computer workstations. The use of databases allows for the tracking of long-term trends that has the potential to improve the reliability of measurements and provide information for comparison overtime or between animals within a particular farm. However, systems of this type are expensive and are necessarily dependent on the underlying sensor technology with the accompanying difficulties described herein.

No cost effective solution suited to automatic performance management within practical tolerances during the normal milking process has been proposed at this time.

DISCLOSURE OF INVENTION In an effort to ameliorate the forgoing disadvantages or to at least provide the public with a useful choice, the present invention proposes methods and apparatus of use in monitoring and determining stock and plant performance during the normal milking session. 4 According to the first aspect of this invention there is provided a method for determining dairy stock and plant performance comprising the steps of: making performance measurements during a normal milking session; collecting measurements to incrementally develop performance profiles 5 during a normal milking session; storing performance profiles as a completed set at the end of a normal milking session; applying a best fit matching of a stored set to the current milking; providing settings for user or pre-determined standards of performance; and providing performance assessment for measurements during the current milking session using assessment criteria that are in part determined by stored profiles.

According to the second aspect of this invention there is provided an apparatus for determining dairy stock and plant performance comprising: a unit for making measurements during a normal milking session; a unit for collecting and developing measurement profiles, storing completed sets of profiles, matching stored profiles to the current milking and 20 calculating performance criteria to be compared with measurements during the current milking session from standards of performance; a unit for annunciation of performance during the current milking session; and a network for communicating information between units.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will now be described by example with reference to the accompanying drawings in which: Figure 1 shows a schematic diagram of a manifold assembly; Figure 2a shows an end view of a manifold assembly; Figure 2b shows a side view of a manifold assembly; Figure 3 shows a schematic diagram of a sensing system; Figures 4a and 4b show phasor diagrams produced by the circuit of Figure 3; Figure 5 Figure 6 Figure 7 DETAILED DESCRIPTION The present invention consists of methods and apparatus of use in the monitoring and determination of stock or plant performance during the normal milking process. The methods and apparatus may be used independently or used in 10 combination.

In one embodiment of the present invention the use of the combined methods and apparatus the results in a system that monitors and determines alerts for abnormal stock performance including indicators of milk with mastitis, milk flow, 15 milk volume yield and end of milking. It also monitors and determines alerts for abnormal plant performance including abnormal bail equipment operation and cleaning characteristics. In this embodiment of the present invention, the system comprises a Bail Unit for each milking cluster and a single central Command Unit. A common 2-wire bus for power distribution and communication connects all 20 units.

Of relevance to the first and second aspects of the present invention is a manifold that connects in series with the long milking tube from the milking cluster.

Solution is transported in the milking tube under differential vacuum and 25 subsequently flows through the manifold as shown in Figure 1.

Solution enters the manifold through the entry pipe (1) and passes into an entry chamber (2). Air and solution are separated in the entry chamber by a whirlpool and/or settling effect with solution outflow from the outer peripheral of the rotating 30 mass and/or from the bottom of the entry chamber. shows a sensor circuit for use in the system of Figure 3; shows a schematic diagram of a signal conditioning circuit; and shows an information flow diagram.

From the entry chamber outflow the separated solution passes into an accumulation chamber (3). Chamber walls or baffles are used to improve air solution flow (2a) and define the boundary between the entry and accumulation 6 chambers (3a). Solution is able to build-up in the accumulation chamber and provides short periods of continuous supply to an attached conductivity tube or tubes (6).

The conductivity tube or tubes define the physical arrangement for conductivity sensing. A high length to sectional area ratio and a co-ordinated external electrode configuration (12) is used to enhance solution conductivity effects and reduce solution dielectric effects. This allows for the extraction of a useful 10 conductivity measurement from the otherwise dominating solution capacitance. In one embodiment of the present invention useful performance was achieved with a single plastic tube of about length 120 mm, diameter 9 mm, and wall thickness 0.9 mm and electrodes of about 50 mm length. To ensure the same potential at each end of the tube to prevent erroneous effects from currents in solution 15 external to the tube, one of the electrodes was split into two parts located at each end of the tube and the other electrode was located in the centre of the tube.

The conductivity tube or tubes incorporate a mechanical or fluid dynamic exit restriction (8) to extend solution residence time under conditions of low flow. The 20 size of a mechanical restriction is a compromise between extended residence time and the ability to pass dense or solid components in the solution flow without blockage. In one embodiment of the present invention the exit restriction was achieved by the disturbed fluid dynamic flow where solution the conductivity and bypass tubes re-combine (10).

The conductivity tube or tubes also receive preferential supply from the accumulation chamber to extend performance under low flow conditions. The preferential supply is due to the separation and settling of solution within the entry (2) and accumulation (3) chambers and the lower position of conductivity tube 30 entry orifice (4) compared to entry orifice (5) of the alternative solution passageway through a bi-pass tube or tubes (7).

The bi-pass tube or tubes (7) carry overflow solution from the accumulation chamber mixed with the separated air stream from the entry chamber. The tube 7 dimensions including the entry (5) and exit (9) orifices ensure that the over-all manifold causes minimal head loss.

In a preferred embodiment of the present invention the bi-pass tube or tubes are 5 fitted with electrodes and used to measure fluid mass within the tube. This can be used together with the measurements from the conductivity tubes to determine flow rate and volume yield from the measured mass and mass time dependency, which indicates velocity. Here the tube requires a relatively high surface to sectional area ratio and a co-ordinated external electrode configuration (13) to 10 enhance both dielectric and conductivity effects associated with fluid mass. In one embodiment of the present invention useful performance was achieved with four parallel plastic tubes of about length 120 mm, diameter 12 mm, and wall thickness 0.9 mm and electrode of about 50mm length. One of the electrodes was split into two parts located at each end of the tube and the other electrode 15 was located in the centre of the tube as for the conductivity tube above.

Solution velocity is obtained from the time dependence of the measured mass. This is most easily illustrated by considering asymmetric electrodes as an example. Fluid often arrives in large "plugs" in sympathy with the pulsation of the 20 milking machine. With an asymmetric electrode, the "plug" first moves past the smaller electrode and then causes a time dependent ramp in mass measurement as it moves past the large electrode. The known electrode length divided by the ramp time gives velocity. Filtering over many "plugs" gives practical results with minimum uncertainty. Other similar methods for obtaining velocity using time 25 dependent mass measurement with symmetrical or split electrodes are equally as effective. In a preferred embodiment of the present device split electrodes were used.

An exit chamber (10) collects the outflow from both the conductivity and bi-pass 30 tube or tubes. The solution exits the manifold through an exit pipe (11) One physical implementation of a manifold arrangement is illustrated in Figures 2a and 2b where the entry pipe (1) is tangential to entry chamber (2) to produce the "whirlpool" effect. Like integers have like numbers to those used in figure 1. 8 Of relevance to the third and fourth aspects is conductivity sensing and signal conditioning circuitry that can operate in conjunction with the manifold and attached electrodes described herein or with some other arrangement.

Figures 3 and 4 show the signal sensing block diagram and phasor addition method used to improve the conductivity measurement.

The sensing system is driven by an oscillator (50) producing a high frequency 10 sine-wave excitation voltage (51). The oscillator frequency is selected for a particular fluid tube and electrode arrangement. In one embodiment of the present invention the manifold described herein a frequency of about 5 MHz was high enough to give a satisfactory performance. The excitation voltage (51) is fed to a conductivity measuring circuit section (53-60) for solution in the conductivity 15 tube or tubes (52). The excitation voltage is also fed to a duplicate conductivity measurement circuit section (53a-60a) and an additional low ion mass measuring circuit (62-69) section for solution in the bi-pass tube or tubes (61). For solutions with high ion concentration a compensated conductivity measurement is used to determine mass whereas for solutions with low ion conductivity the low ion mass 20 measurement is used.

The excitation voltage is impressed across a conductivity sensor assembly (52) by a coupling circuit (53). The sensor assembly represents a load that can be considered as a fixed capacitance due to the tube wall in series with a parallel 25 combination of fixed capacitance due to the solution dielectric and a variable resistance (conductance) due to ion solutes in the solution. However, because of the physical design of the conductivity tube and electrode assembly described herein, the capacitance due to the solution can be neglected and the load simplified to a fixed capacitance and variable resistance series circuit.

The current in this capacitance-resistance series circuit is shown in the phasor diagram of Figure 4(a). The reference phasor is the excitation voltage (101). The circuit current leads this voltage with a phase angle that depends on the variable resistance component. At high resistance (low ion concentration) the current is 9 small and has a small phase lead (102). At a standard resistance (typical ion concentration) the current is moderate with a phase lead of about 45 degrees (103). At low resistance (high ion concentration) the current is large and with a phase lead approaching 90 degrees (104). With the ion concentration range 5 found in healthy to mastitis infected cows the practical variation in phase is rather less than the 0 to 90 degree limit points above. In addition, the current amplitude is small for practical excitation voltage amplitudes and therefore sensitive to interference from electrical noise. The phase of the current does however represent conductivity and is the fundamental output of the sensor (54) (Figure 10 3).

In order to improve both the range of phase shift and amplitude of the conductivity sensor, a second development current is employed. This is derived from the excitation voltage (51) using a development current circuit (55). The 15 circuit is designed to produce a current that is equal in magnitude but opposite in phase to the imaginary component of the sensor current output at a standard ion concentration. The development current output (56) and conductivity sensor current output (54) are summed (57) to give the improved performance.

The phasor diagrams in Figure 4(a) and (b) illustrate the effect of employing the development current. The development current as designed lags the excitation voltage (101) by -90 degrees (105). When added to the sensor current the resulting output amplitude is maintained a moderate level and the phase shift range is doubled. At high resistance (low ion concentration) the current is 25 moderate with a lag approaching -90 degrees (107). At a standard resistance (typical ion concentration) the current is moderate and about the same phase as the excitation voltage (108). At low resistance (high ion concentration) the current is moderate with a lead approaching 90 degrees (109).

The sensor and development current summer (57) also converts the current to an ac coupled output voltage (58). This is fed to a phase detector circuit (59) to give an unfiltered conductivity measurement output (60).

In one embodiment of the present invention the excitation voltage coupling, current phase-shift response, development current summation and ac coupled output are performed by a compact circuit segment shown in Figure 5. A high frequency transformer (121 to 123) shown in the equivalent circuit form of an 5 ideal transformer (121,122) and magnetising inductance (23) is used to couple the excitation voltage (51) to the sensor arrangement (52). The sensor response current flows in the transformer secondary (121) and is reflected in the transformer primary (122). The development current flows in the magnetising inductance of the transformer (123). In a real transformer the primary and 10 magnetising inductance are one, and the response and development currents are summed intrinsically. A resistor (126) is used to convert the resulting current (125) to a voltage which is ac coupled using a coupling capacitor (127) to provide the required output voltage (58). The use of a transformer as a coupling device also provides a large common-mode impedance caused only by transformer 15 inter-winding and stray capacitance (124). This reduces spurious behaviour due to stray fields coupling into the sensor fluid and surrounding environment. For one implementation of the sensor arrangement discussed herein the transformer was designed with a magnetising inductance of about 13 uH.

In a preferred embodiment of the present invention the excitation voltage (51) is also impressed across a bi-pass tube or tubes and sensor assembly (61) by a coupling circuit (62). The sensor assembly represents a load that can be considered a capacitor in series with a parallel combination of capacitor and resistor as for the conductivity sensor. With a solution of low ion concentration, 25 the resistance due to the solution is small and the load can be simplified to a fixed capacitance and variable capacitance series circuit.

The current in this capacitance-capacitance series circuit leads the excitation voltage by 90 degrees and has an amplitude that depends on the solution mass 30 within the bi-pass tube. The current has zero amplitude for an empty tube and reaches a maximum when the tube is completely full. The current amplitude when the tube is full is relatively small for practical excitation voltage amplitudes and becomes increasingly sensitive to interference from electrical noise as the 11 tube empties. The current does however represent mass as amplitude and is the fundamental low ion concentration mass sensor output (63) (Figure 3).

To prevent interference from electrical noise an offset current is used. The offset 5 current is derived from the excitation voltage (51) using an offset current circuit (64). The circuit is designed to produce constant amplitude current in phase with the sensor current output. The offset current (65) and mass sensor current (63) are summed (66) so that an empty bi-pass tube gives a current equal to the offset current.

The mass sensor and offset current summer (66) also converts the current to an ac coupled output voltage (67). This is fed to an amplitude detector circuit (68) to give an unfiltered mass measurement (69). Mass flow, volume and yield are determined from the time dependence of mass within the sensor tube.

In one embodiment of the present invention the excitation voltage coupling, current amplitude response, offset current summation and ac coupled output are performed by the same compact circuit as with the conductivity measurement in Figure 5.

When the solution has a high ion concentration the situation is similar to that for the conductivity sensor assembly and a duplicate signal conditioning circuit is used to represent mass as phase shift as the unfiltered high ion concentration mass sensor output. Since this operates in identical fashion to the conductivity 25 sensor no further explanation is given here.

The measurement from the conductivity sensor defines the ion concentration of the solution and is used to determine the most appropriate mass sensor output.

In one embodiment of the present invention the phase detector (59) and amplitude detector (68) are as set out in Figure 5. The approach used gives an accurate and cost effective solution to detecting the phase and amplitude of the high frequency sine-wave signals.

The phase detector compares a zero phase reference output (72) to the improved conductivity output (68). The reference output is used to give a phase reference that tracks the improved conductivity output. The reference output is derived from the excitation voltage (51) using a reference current circuit (70). The output from the reference current circuit (71) is fed to a current to voltage converter with an ac coupled output (72) to give the required zero phase reference (73). In one embodiment of the present invention the zero phase reference output is obtained using a resistance-capacitance phase shift circuit and an ac coupling capacitor.

Subsequent signal conditioning for the zero phase output (72) and the improved conductivity output (68) is through identical pathways to ensure the relative phase relationships are accurately maintained. A precision squaring circuit is formed by the dc restoration circuit (74), high-speed comparator (75) and 50% duty cycle integrator (76). The squaring circuit operates as a feedback regulator for the dc restoration level that produces a square-wave output with a precise 50% duty cycle. Squared-up forms of the two input waveforms (56 and 54) are fed into a high-speed digital phase comparator (80) to give the unfiltered conductivity measurement (60).

The amplitude detector also utilises a precision squaring circuit. This is formed by the dc restoration circuit (81) high-speed comparator (82), duty cycle reference (83), and programmable duty cycle integrator (84). Again the squaring circuit operates as a feedback regulator for the dc restoration level but in this case produces a rectangular-wave output with a duty cycle equal to the duty cycle reference (83) which is set for a low value. When the amplitude of the mass sensor output (67) is small the output of the integrator will produce a dc value slightly less than the switching threshold of the high-speed comparator to maintain a low duty cycle. When the amplitude of the mass sensor output (67) is large the output of the integrator will produce a dc value significantly less than the switching threshold of the high-speed comparator to maintain the same low duty cycle. In this way the dc restoration level represents amplitude and is the unfiltered low ion mass output (69). 13 Of relevance to the third and fourth aspects is the use of information in the determination of stock and plant performance during a normal milking session. Figure 7 shows the information flow.

Multiple measurements are made by multiple units (150) during the milking process. Within each unit the measurements are filtered, combined and qualified by analysing trends as a cow is milked to provide derived parameters with improved reliability or accuracy. The derived parameters are made available over a communications network (152). The information from the communications 10 network is collected and used to incrementally construct statistical profiles for each derived parameter (152) and at the end of the milking process the set of profiles is saved for later use (154). At some stage near the start of the milking process a determination is made as to which of the stored set of profiles best matches the current milking situation (155). The best match set together with 15 user or pre-determined standards of performance (156) and the current incremental profiles (152) are used to calculate performance assessment criteria (157). These are made available on a communications network or used locally. The assessment criteria are used to determine performance by comparison with measurements during the milking (158).

The use of best match profiles allows performance to be monitored and assessed relative to the norms of a particular herd and plant. Performance determined in this way is of the most practical value as farm management is performed within parameters determined by local conditions. It allows performance to track with 25 local variation such as farm location, changes in feed, and stage of lactation. This prevents excessive alerts from absolute measurement not relevant to the local situation. In addition, the selection of best match profiles including frequency of occurrence distributions allows for management with performance standards defined by a number of stock. A standard of this kind can be used with the best 30 match profile distribution to calculate performance criteria for measurements to separate out the desired number of stock. This is in sympathy with on farm management practices and rather more useful than management based on an absolute thresholds. 14 In one embodiment of the present invention a Bail Unit is used for making measurements (151). The Bail Unit incorporates a microcontroller for signal processing of measurements, operation of local interface outputs and communication over a common power bus.

Signal processing combines conductivity and mass measurements with additional measurements of solution temperature, ambient temperature and time. Processing includes range and mean filtering to remove noise and improve accuracy, linearisation and scaling to give corrected values, and qualification by 10 analysing trends as a cow is milked to provide derived parameters with improved reliability or accuracy. Time and frequency domain analysis are used to determine rates and periodicity. Known heat transfer models can be used to relate temperature information to mixed air and solution to provide an alternative mass-flow estimation.

The information from the low level measurements and derived parameters from signal processing are used to define application level measurements of flow rate, indicative mastitis, volume yield, flow periodicity, flow temperature, cycle time, cycle state (start, mid, end), and cycle type (milking or cleaning).

The application level measurements are compared with calculated performance criteria (157) to determine stock and plant performance. Stock performance includes indicators of milk with mastitis, milk flow, milk volume yield and end of milking. Plant performance includes abnormal bail equipment characteristics 25 including milking time, air solution ratio, pulsator rate. Abnormal cleaning characteristics include correct hot or cold cycles, cleaning phases (rinse, wash) cleaning volume, cleaning temperature and detergent use.

In one embodiment of the present invention performance criteria and 30 measurements are compared in the Bail Unit and indicated through a local interface. One arrangement includes a display with lights (LED's) for performance level and parameter type and an audible sounder and relay output for an unsatisfactory performance alert. Alerts include mastitis (Mastitis alert), low yield (Yield alert), abnormal bail equipment (Plant alert) and abnormal cleaning (Clean alert). Application level parameters are also indicated through the local interface. These include current milk flow, mass, volume, yield and end of milking.

In one embodiment of the present invention the communications network (151) is 5 provided for using modems on a power distribution bus. Connected to the same bus is a Command Unit. In the Command Unit measurements are collected and parameter profiles built (152), sets of profiles stored (152), settings for user and pre-determined standard of performance are provided (158) and performance assessment criteria are calculated.

In one arrangement sets of profiles are stored for two previous milkings (normally the previous morning and evening). The best match profile set is that which corresponds to the current milking based on an elapsed time of about 24 hours. User input is set through a function (parameter type) and threshold (standard of 15 performance) switch stored in the Command Unit with a save switch. Lights (LED's) are used to give an indication of the current settings. Functions can be disabled by special threshold values.

The sensor technology described herein provides for a sufficiently accurate 20 measurement of milk conductivity under the conditions of low flow milk mixed with air to be of practical use in detecting mastitis. It is possible to make accurate conductivity measurements because of the sensor manifold design that provides at least partial separation of milk and air and a conductivity tube geometry with co-ordinated electrode arrangement for sensing through the tube wall that is 25 sensitive to conductivity. Measurements are further improved by signal conditioning that uses a phasor addition method to cancel unwanted effects caused by sensing through the tube wall. This increases output range and noise immunity. The bi-pass portion of the sensor manifold is used with similar electrode arrangements and sensing circuits enable the solution mass to be 30 measured and therefore mass, flow, volume and yield to be determined. The sensor manifold, electrode arrangement and signal conditioning circuits form a simple, reliable, and low cost sensor with no moving parts, no electrodes in solution, no hygiene traps and minimal head loss. 16 Placement in the top of the long milking tube provides protection from the harsh milking environment and enables automatic measurements. Preferred arrangements include additional sensing and signal processing that provides for additional measurements of stock and plant performance.

Implementation using multiple Bail Units communicating over a network to a single Command Unit allows for profiles of measurement parameters to be constructed and stored.

This enables standards of performance to be defined in terms of on farm stock 10 and plant norms using levels or number of stock in contrast to other systems that enforce absolute standards. Using standards within an on farm context is in keeping with farm management as is actually practised.

The Bail Unit and Command Unit implementation provides an effective low cost 15 solution well suited to the practical farm environment. The basic system can be extended for other dairy automation requirements including remote display, cow identification and counting, automatic cluster removal, supplement feeding and automatic cow drafting.

A preferred communications interface on the Command Unit together with support software enables remote access using standard telecommunications infrastructures including the internet to provide farm or industry based computer systems and services.

Where in the foregoing description reference has been made to integers or components having known equivalents then such equivalents are herein incorporated as if individually set forth.

Although this invention has been described by way of example it is to be 30 appreciated that improvements and/or modifications may be made thereto without departing from the scope or spirit of the present invention. 17

Claims (23)

WHAT WE CLAIM IS:

1. A method for determining dairy stock and plant performance comprising the steps of: 5 making performance measurements using a measurement unit during a normal milking session; collecting measurements to incrementally develop performance profiles during a normal milking session; storing performance profiles in memory as a completed set at the end of a 10 normal milking session; using a processor, applying a best fit matching of a stored set to the current milking; providing settings for user or pre-determined standards of performance; and 15 using a processor, providing performance assessment for measurements during the current milking session using assessment criteria that are in part determined by stored profiles.

2. The method according to claim 1 wherein measurements are made 20 automatically during the milking session.

3. The method according to any one of claims 1 to 2 wherein measurements are made for individual cows with or without retaining individual cow identification information. 25

4. The method according to any one of claims 1 to 3 wherein measurements are made or derived that represent stock performance including milk flow rate, volume, mass, yield and conductivity. 30 5. The method according to any one of claims 1 to 4 wherein measurements are made or derived that represent plant milking performance including pulsing action, milking time, and air or fluid flow rates or ratios.

INTELLECTUAL PROPERTY OFFICE OF N.Z. 2 f AUG 2007 RECEIVED 18

6. The method according to any one of claims 1 to 5 wherein measurements are made or derived that represent plant cleaning performance including hot and cold cycles, fluid volume, time and detergent use. 5

7. The method according to any one of claims 1 to 6 wherein performance profiles include distributions, averages, maximums, minimums for individual cows or parts or all of a herd.

8. The method according to any one of claims 1 to 7 wherein stored performance 10 profiles for stock can be representative of standards defining a number of cows or a level within the herd.

9. The method according to any one of claims 1 to 8 wherein stored profiles are matched on the basis of the last diurnal milking period corresponding to the 15 current milking.

10. The method according to any one of claims 1 to 9 wherein performance during the current milking can be categorised as satisfactory or unsatisfactory or represented as a relative value. 20

11. The method according to any one of claims 1 to 10 wherein measurements result in local annunciation.

12. The method according to any one of claims 1 to 11 wherein measurements 25 are used to determine end of milking or cluster removal for individual cows.

13. The method according to any one of claims 1 to 12 wherein individual cow identification system is incorporated. 30

14. The method according to any one of claims 1 to 13 wherein remote access is incorporated including access to a farm computer system, industry computer system or the internet.

15. An apparatus for determining dairy stock and plant performance comprising: INTELLECTUAL PROPERTY OFFICE OF N.Z 2 I AUG 2007 RECEIVED 19 a unit for making measurements during a normal milking session; a unit for collecting and developing measurement profiles, storing completed sets of profiles, matching stored profiles to the current milking and calculating performance criteria to be compared with measurements during the 5 current milking session from standards of performance; a unit for annunciation of performance during the current milking session; and a network for communicating information between units. 10

16. The apparatus according to claim 15 wherein measurements are made with a unit at each milking cluster.

17. The apparatus according to claim 15 or 16 wherein measurements are made at each milking cluster in-line with the long milk hose from the milking cluster or 15 the equivalent.

18. The apparatus according to any one of claims 15 to 17 wherein a commercial computer with appropriate adaptations and software or dedicated device with inbuilt computing capability is used for the unit collecting and developing 20 measurements to define and store performance profiles and determining performance standards for the current milking session.

19. The apparatus according to any one of claims 15 to 18 wherein the unit for annunciation is also used to input user selected standards or to announce 25 performance for the current milking session.

20. The apparatus according to any one of claims 15 to 19 wherein some'or all of the units perform their functions automatically. 30 21. The apparatus according to any one of claims 15 to 20 wherein announcement of performance for the current milking session is made with a unit at the milking cluster position as part of the measuring unit or a separate unit.

INTELLECTUAL PROPERTY OFFICE OF N.Z. 2 1 AUG 2007 RECEIVED 20

22. The apparatus according to any one of claims 15 to 21 wherein communication between units is performed using a standard telecommunication network system or a dedicated network including a power line modem. 5

23. The method according to any one of claims 1 to 14, implemented in an apparatus according to any one of claims 15 to 22. 10 RADIAN TECHNOLOGY LIMITED 15 By Their Attorneys Ellis I Verboeket I Terry vrfice of Nz 2 f AUG 2007 BMceiv Fp