Title: Method and system for electronically examining animals such as cows
This invention relates to a method for electronically examining milk-producing animals such as cows. This invention also relates to a system for carrying out such method.
A method and system for electronically examining milk-producing animals such as cows are known per se. Thus, it is known to measure the conductivity of raw milk from an animal with the aid of sensors with the objective of determining whether the animal is possibly suffering from a particular disease such as mastitis. If the conductivity of the milk exceeds a particular threshold, it can be assumed that an udder of the animal is infected. A disadvantage, however, is that it is not clear what type of bacterium is the cause of the infection. As a result, it is less efficiently possible to control the infection and it is not immediately clear whether the milk is suitable for consumption any longer. A further disadvantage is that the conductivity measurement is not an accurate indication of a possible infection.
The object of the invention is to provide a solution to the disadvantages mentioned.
Accordingly, the method according to the invention is characterized in that with the aid of at least one electronic nose an odor is detected of raw milk coming from the at least one animal to be examined, while the electronic nose determines a signal representing the detected odor; and that the signal is supplied to a computer for further processing and analysis of the detected odor. The invention is based on the insight that in this way in a highly accurate manner the animals can be examined, for instance, for an infection of the udder. In particular, it is thus possible to determine which bacteria are present in the raw milk. It can thus be accurately established, for instance, whether the animal is possibly suffering from mastitis and which bacteria are the cause of this.
In particular, at least one electronic nose is contacted with air coming from the raw milk of the animal for detecting the odor.
Preferably, with the aid of the at least one electronic nose the odor of milk is detected per quarter of the udder of the animal from which the milk originates, for further processing with the aid of the computer. In this way it is even possible to determine which quarter of the udder is possibly infected.
The method according to the invention is suitable for laboratory application, but in addition also particularly suitable to be carried out on the farm. The computer and the electronic nose can then be designed as a stand-alone unit by means of which the farmer can, periodically or whenever he has doubts about the health of the animal, in this case a cow, perform an odor analysis on the basis of raw milk from the cow. It is also possible, however, that the at least one electronic nose is incorporated in a milking plant or a milking robot. This enables on-line analysis of the milk from cows which have just been milked. The at least one electronic nose can then, for instance, be incorporated in a milk tank, milk line, milk claw or teat cup of the milking robot or milking plant. The system according to the invention is characterized in that the system at least comprises an electronic nose for detecting an odor from raw milk coming from the at least one animal to be examined, while the electronic nose determines a signal representing the detected odor; and that the system further comprises at least one computer to which the signal is fed, which computer is arranged for further processing and analysis of the detected odor.
The invention will presently be further elucidated with reference to the drawing. In the drawing:
Fig. 1 shows a system according to the invention for carrying out a method according to the invention.
The system 1 for electronically examining milk-producing animals such as cows at least comprises an electronic nose 2 and a computer 4 which is connected via a line 6 with the electronic nose 2.
The electronic nose 2 is an electronic nose that is known per se and can comprise, for instance, multiple gas sensors 8.1, 8.2, .... to 8.n each arranged to deliver a signal when they come into contact with a predetermined odor. The signals generated by the individual gas sensors 8.1 to 8.n are supplied via line 6 to the computer 4. In use, the gas sensor 2 is placed above raw milk 10 originating from the animal to be examined. It is also possible that air is extracted above the milk in question to be subsequently passed along the sensor. Accordingly, the at least one electronic nose is contacted with air coming from an amount of the raw milk coming from the animal, for detecting the odor. In this example, the raw milk 10 is located in a housing 12 which, according to the invention, can consist of a multiplicity of different objects and/or parts of machines, as will be set out hereinafter. By means of the electronic nose, therefore, an odor of the raw milk is detected, more particularly the presence of a number of predetermined odors.
Also, with the aid of the electronic nose 2, on line 6 a signal is generated which in this example comprises the signals from the individual gas sensors 8.1 to 8.n. This signal is supplied to the computer 4 for further processing and analysis of the odor of the raw milk as determined by means of the electronic nose.
Thus, for instance, mastitis infection in the udder of the cow can be established. Early diagnosis of a mastitis infection in dairy stock can lead to a high degree of prevention and amelioration of the quality. Low concentrations of mastitis causing bacteria can be detected in the milk. These mastitis causing bacteria cause the milk to acquire an odor which is known per se and depends on the type of bacterium. When thus in an early stage particular types of bacteria are detected by means of the
electronic nose, the animal can be subjected to preventive treatment in order that it does not get mastitis. When the infection is already present, the animal can be treated at an early stage.
The types of bacteria that can thus be detected are, for instance, Staphylococcus aureus, Escherichia coli, Streptococcus agalactiae, Streptococcus uberis and others.
Preferably, with the aid of the at least one electronic nose 4 the odor from milk is detected per quarter of the udder of the animal from which the milk derives, for further processing by means of the computer. Thus it can be established exactly which quarter of the udder of the animal is possibly infected with a particular bacterium. In that case, the milk sensor 2 can, for instance, be included in the line to the teat cup or in the teat cup of a milking robot or milking plant. The housing 12 schematically shows in Fig. 1 the line in question to the teat cup and the teat cup.
In particular, the odor of raw milk is analyzed per quarter and for each quarter of an udder, as shown in Fig. 2. In that example, preferably, in each milk line/cup 12.1 - 12.4 which is connected with a milk claw of a milking robot, a connection to the electronic nose 2.1 - 2.4 is included. The electronic noses 2.1 — 2.4 are each connected via lines 6 with the computer 4 and are, for instance, identical to the electronic nose 2 of Fig. 1.
It is also possible that the odor of milk coming from a plurality of quarters of an udder of the animal is detected with the aid of the electronic nose for further processing by means of the computer 2. In that case, the raw milk therefore consists of a mixture coming from each of the quarters of the udder of the animal. The housing 12 can then be formed, for instance, by a milk collecting chamber of the milking robot, known per se, as schematically shown in Fig. 1.
In another variant, the odor of milk coming from a plurality of animals is detected by means of the at least one electronic nose for further processing by the computer. In that case, the housing 12 consists, for instance, of a known milk tank of a milking robot in which the milk of the animals is collected.
The electronic nose 2, 2.1 - 2.4 consists, as set out hereinbefore, for instance, of a plurality of sensors 8.1 to 8.n. These sensors can be so designed that the sensors are respectively sensitive to mutually different odors. In this way, an odor spectrum of the raw milk can be determined. By means of the computer, on the basis of the signal determined by the electronic nose 2, 2.1 — 2.4, the presence of a number of predetermined odors can be determined. Thus, for instance, sensor 8.1 can generate the signals when a predetermined first odor is present, while the sensor 8.2 generates a signal when the predetermined second odor is present. All this holds analogously for the sensors 8.3 - 8.n. The computer 4 can be so arranged that it indicates which of the sensors 8.1 to 8.n generates a signal that is above particular predetermined values. If the signal of, for instance, sensor 8.i lies above a predetermined value, it can be concluded that the predetermined odor associated with the sensor 8.i is released by the raw milk.
Since particular types of bacteria release particular characteristic odors or spectra when they are contained in raw milk, it can be determined in this way with the computer which bacteria, or other relevant constituents in the milk, such as, for instance, acetone, progesterone, etc., are possibly present in the raw milk. Accordingly, with the computer, on the basis of the signal of the electronic nose, the presence of at least one predetermined type of bacterium or other constituents in raw milk can be determined. More particularly, it can thus be established whether the animal is infected and thus runs the risk of getting mastitis, or has mastitis already. When by means of the
odor sensor the raw milk is examined per quarter, it can be determined for each quarter whether it is infected. If it is, preventive measures can be taken to prevent mastitis or measures can be taken to cure the animal. As said, it is also possible that the raw milk is examined per udder of the animal. If the computer finds that particular types of bacteria are present in the raw milk coming from the respective udder, a further examination can tell which of the quarters is possibly infected. As said, it is also possible to examine the milk contained in a tank of the milking robot. If, with the aid of the odor sensor, this milk is found to contain particular types of bacteria, the cattle that has been milked can be further examined to determine which animal is infected.
The method can preferably be carried out on the farm. The apparatus as shown in Fig. 1 can then be part of a milking robot as described hereinbefore. In that example, the computer 4 can also be a part of a farm computer 14, known per se. In the farm computer 14, still other data of the animals are present, such as their feeding behavior, the milk yield per animal, the conductivity of the milk of the animal, etc. To obtain a still better picture of the health of the animals, these data can be processed in combination with the data of the computer 4. The apparatus according to Fig. 1 can also be used per milk stand.
In that case, each milk stand is provided with an apparatus according to Fig. 1, with the housing 12 being formed, for instance, by a milk claw of the respective milk stand. In this way, the examination of the animal can be carried out during milking. The invention is not limited in any way to the exemplary embodiments and forms of application outlined hereinbefore. The computer can, for instance, be arranged such that on the basis of the presence of one predetermined odor and/or presence of a predetermined combination of a predetermined number of odors, it is concluded that a particular bacterium and/or other relevant constituents in the milk, such
as, for instance, acetone, progesterone, etc., and/or a predetermined combination thereof are present in the raw milk. The odors to which the sensors 8.1 to 8.n are sensitive can be designated, for instance, by Gi, G2, to Gn. Further, it can be indicated by the number g, (i = 1, 2, ...n) in what measure the odor Gi is present, or the extent to which the odor Gi is present in excess of a particular threshold value. The detected odor spectrum can then be expressed in the vector G = (g1; g2, g3, gi, gn).
The types of bacteria to be possibly detected can be designated, for instance, by Bi, B2, B3, ...Bm. The number bj can indicate to what extent the bacterium B3 is present in the raw milk, or if the amount of bacteria Bj present in the raw milk exceeds a predetermined threshold value. The vector B = (bi, b2, b3, ...bm) can then be designated as the bacterial spectrum. The relation between the odor spectrum G and the bacterial spectrum B can be expressed in the following matrix multiplication: B = A . G
Wherein A is n x m matrix whose matrices Aij can be determined experimentally. The odor spectrum can also serve for detecting diseases, for instance, metabolic diseases, such as, inter alia, acetonemia (acetone smell), TB, etc. Also, odor spectra can give information about reproductive stages of the cow, for instance whether or not the animals are estrous/in heat. Another option is smelling other parameters that tell something about the status (health, reproduction, etc.) of the animal. The detector can therefore also be designed for obtaining information about the health and/or fertility of the animal. Further, the apparatus according to Fig. 1 may also be designed as a stand-alone unit. The farmer can then place the odor sensor 2 above the raw milk 10 to examine it. To that end, he can take a sample of raw milk and deposit it in a random reservoir 12, such as, for instance, a cup. Such variants are all understood to fall within the scope of the invention . The apparatus is also applicable outside the farm, as in a laboratory.