EP3107381A2

EP3107381A2 - Livestock-farming system

Info

Publication number: EP3107381A2
Application number: EP15705384.4A
Authority: EP
Inventors: Cornelis Christianus Franciscus Havermans; Serge Louis Loosveld
Original assignee: Lely Patent NV
Current assignee: Lely Patent NV
Priority date: 2014-02-17
Filing date: 2015-02-03
Publication date: 2016-12-28
Also published as: WO2015122763A2; NL2012275C2; WO2015122763A3

Abstract

Livestock-farming system (1), comprising an animal space (4), such as an animal shed or a meadow, for keeping livestock animals (9), and a detector system (17, 24) for detecting an entity (20) within a predefined detection zone (18), wherein the detector system comprises - a detector (17), and - a control unit (24) which comprises a memory (28) and is connected to the former, wherein detecting comprises receiving a signal from the detection zone using the detector, which signal is suitable for recognising the entity, and wherein the control unit processes the received signal to yield detection data relating to the entity, and stores these in the memory. In this way, it is automatically recorded which entities visit the animal space or at least come sufficiently close to it to have an effect thereon. It is then possible to take respective action, if desired automatically.

Description

Livestock-farming system

The invention relates to a livestock-farming system, comprising an animal space, such as an animal shed or a meadow, for keeping at least one livestock animal, and a detector system, wherein the detector system comprises a detector, and a control unit which comprises a memory and is operatively connected to the at least one detector.

Such a livestock-farming system comprises for example an animal shed with a security camera with recording facilities. The security camera comprises, for example, a movement sensor.

A drawback of such systems is the fact that they are not always able to find the required data. An example thereof is monitoring which contact there has been with the animals, for example in order to contain contagious diseases.

It is an object of the present invention to improve the system mentioned in the introduction, at least is more suitable to solve the abovementioned drawback.

The invention achieves this object by a livestock-farming system according to claim 1 , in particular comprising an animal space, such as an animal shed or a meadow, for keeping at least one livestock animal, and a detector system, wherein the detector system comprises a detector, and a control unit which comprises a memory and is operatively connected to the at least one detector, wherein the detector system is configured to detect an entity within a predefined detection zone which comprises a part of said animal space and/or a part of adjacent surroundings, wherein detecting comprises receiving a signal from the detection zone using the detector, which signal is suitable for recognising the entity, and wherein the control unit is configured to process the received signal to yield detection data relating to the entity, and to store the detection data in the memory. The idea behind the invention is that, in this way, it is automatically recorded which entities visit the animal space or at least come sufficiently close to it to have an effect thereon. This saves a lot of investigative work compared to the known system, where someone would have to watch, for example, all the footage of a continuous period. If desired, it is now also possible to immediately take suitable action instead of only warning an operator. All this will be explained in more detail in the following description.

It should be noted here that "a part of the animal space" may also comprise the entire animal space, such as the animal shed. Furthermore, the expression "signal suitable for recognising the entity" is understood to mean that the signal contains information relating to the entity, in particular in order to be able to recognise the entity. In this case, "recognise" in turn comprises determining the type of object, such as "human", "animal" etc., and possibly even the identity of the entity. In other words, the signal is representative of the entity, again in particular of at least a type of object thereof, and more particularly of an identity thereof. It will be clear that the detector cannot determine an entity from each arbitrary signal, but it will be configured to be able to recognise relevant entities. This will be explained in more detail below.

Particular embodiments are described in the attached subclaims, as well as in the following description.

In embodiments, the control unit is configured to store the received signal. This may be useful when checking or proving entities which have been observed, for example if the identity of the entity could not be determined automatically. If it has been possible, the stored signal can also be deleted or does not have to be stored per se.

In embodiments, the detector comprises a detector of electromagnetic signals. Electromagnetic signals comprise a very wide spectrum and, like all wave signals, they are immune to other signals outside their own frequency (range), so that they can often be accurately detected and traced back to an entity. In particular, the detector may comprise a GPS tracker. This is understood to mean a tracker for detecting and/or following a GPS transmitter or the like, as is often fitted to a vehicle (lorry, rental car) or a valuable object. The detector may also comprise a radio beacon detector. It should be noted that, in that case, the detector will then often have to be a dedicated detector in order to pick up the, often specific, (radio) signal. In that case, it is advantageous if the beacons or the like and the detectors are tuned to each other. Such signals are intended to be received and can therefore easily be linked to an entity in most cases. Furthermore, it is also possible for the detector or the control unit to be, preferably automatically, connected to a third-party database, which database contains information regarding entities of said party which are fitted with a GPS transmitter or similar transmitter. Thus, all entities (vehicles, people, objects) of said party which have been fitted with such a GPS transmitter or the like can also readily be recognised on the livestock-farming system, and their visits can automatically be recorded.

In embodiments, the detector comprises an optical camera for capturing optical images, wherein the control unit furthermore comprises an image-recognition system which is configured to analyse the received images and to recognise at least the object type of the detected entity, and in particular to determine an identity of the detected entity, in the analysed images by comparison with reference data stored in the image-recognition system. Such optical systems with image recognition are in theory already well-developed and reliable. Obviously, it is also possible to use a plurality of cameras, which will then, for example, have to be operatively connected to the control unit. Combinations of a plurality of detectors of different types, such as one or more GPS trackers and one or more cameras, are likewise possible.

In embodiments, the detector is attached to a part which is to be worn by one of the animals, such as a collar or ear tag. Such a camera which is being worn offers advantages when it is desired to be able to accurately check for each animal whether a person or another animal has been in contact with the animal wearing the detector. Furthermore, it is also possible to use the natural curiosity of animals, so that the detector is automatically directed at a noise, an intruding figure, etc. It should be noted that miniature cameras, as required for use in an ear tag or collar, have already been known per se for a long time.

In embodiments, the control unit is furthermore configured to store at least the recognised object type, and in particular the determined identity, together with or in the detection data. In this way, detected entities can be retrieved efficiently and quickly in the stored data, so that an investigation can also lead to results in an efficient and quick manner, without an investigator having to check all the stored signals afterwards in order to determine the data relating to object type and/or identity. Furthermore, it is possible to store additional data, such as in particular the time at which an entity, identity, etc. has been detected and where this took place. Other data are obviously also possible.

In embodiments, the reference data stored in the image-recognition system comprise one or more of the following data types: vehicle licence plates, faces of people, logos and barcodes. Such data types are already often being used as such for marking products and other objects, so that the technology in question has been reliably developed further. It should be noted here that "barcode" in this case comprises both the 1 -dimensional barcode and 2- dimensional barcodes, such as QR codes.

In embodiments, the livestock-farming system comprises at least one device for executing an animal-related operation on the animals, as well as a program for controlling the device, wherein the program is configured to control the device partly on the basis of the stored detection data. This reveals a significant advantage of the invention, as, due to the automatic recognition of (id)entities, it is also possible to automatically react thereto. In this case, "animal- related" means only "nonhuman/animal-related". In this case, the program may be incorporated in the control unit or in a separate control unit of, in particular, said device(s).

In embodiments, the at least one device comprises at least one autonomously movable part, such as a milking robot, an automatic feeding device, a manure slide or an autonomously moving vehicle. Important examples of the device are a milking robot, a manure slide, a feed pusher, such as the Lely Juno™, or an automatic feeding system, such as the Lely Vector™ system, comprising, for example, an automatic "feed kitchen" and an autonomously moving feed vehicle. The program may then, for example, be configured to switch to a "safe" mode, such as moving or driving slowly, when detecting an entity or even to come to a standstill, until the entity is no longer detected. It is also possible, for example, for the program to be configured to disregard a part of the stocks of livestock feed when formulating livestock feed after detecting an entity in the automatic "feed kitchen", comprising a plurality of stocks of livestock feed and an automatic gripping or distributing device. After all, there is a risk that the feed in parts of the stocks has been tampered with or that an entity such as a dog or child may still be present. In an embodiment, the device comprises a milking robot with a cleaning device, as well as a milk reservoir which is connected to the milking robot for collecting milk from the milking robot, wherein the control unit is configured to emit a cleaning start signal to the cleaning device for cleaning the milking robot, and is configured to emit a setting signal to the milk reservoir to prepare the milk reservoir for transferring milk in the milk reservoir to a milk tanker by pumping, and wherein the control unit is configured to automatically emit the cleaning start signal and the setting signal if, and preferably either as soon as or with a predefined time delay, the detection data correspond to a recognised milk tanker. In this case, the device is thus configured to recognise a milk tanker, for example by means of a permanent logo, one or more licence plates, a radio beacon, etc., as described above. As soon as the milk tanker approaches or passes the associated detector, such as at the entrance to the farm yard, the milk tanker is recognised and the control unit generates the two signals, so that cleaning of the milking robot and preparing of the milk reservoir for transferring the milk from the milk reservoir to the milk tanker by pumping are automatically started. The latter may comprise, for example, releasing the (negative) pressure from the milk reservoir, as well as turning one or more valves, such as an inlet valve from the milk line to the milk reservoir or even to a buffer tank. With this embodiment, as little time as possible is lost by not milking or by the milk tanker driver waiting.

The present invention will now be explained in more detail with reference to the drawing, in which the sole figure shows a diagrammatic top view of an embodiment of the livestock-farming system according to the invention.

Figure 1 shows a diagrammatic top view of an embodiment of a livestock-farming system 1 according to the invention. Within a yard enclosure 2, the livestock-farming system 1 comprises a farm yard 3 with a shed space 4 which has a wall 5 containing a gate 6 and a door 7. Inside, there is a shed space 8 containing livestock 9, a milking robot 10 with robot fences 1 1 , a clean room 12 and a feeding passage 13, separated from the shed space 8 by a feeding fence 14. In the feeding passage 13, a feed pusher 15 pushes back the roughage 16.

The figure furthermore shows cameras 17-1 to 17-5, with respective detection zones 18-1 to 18-5. The figure also shows a driveway 19 to the farm yard 3 with a vehicle 20 containing a licence plate 21 and a beacon 22. Reference numeral 23 denotes a beacon detector.

The figure furthermore diagrammatically shows a control unit 24, comprising a communication device 25, a data processor 26, a database 27 and a memory 28.

The livestock-farming system 1 in principle comprises an arbitrary livestock-farming system, that is to say a livestock farm for, for example, dairy animals, slaughter animals, laying hens, etc. It is an object of the present invention to monitor the livestock farm automatically and, in particular, automatically record visits by entities to the livestock-farming system. The livestock-farming system often comprises an animal shed 4 surrounded by a farm yard 3. Often, but not always, the farm yard is surrounded by a yard enclosure 2, which makes it easier to monitor visitors as access to the livestock-farming system is limited to the openings in the yard enclosure 2. However, in the case of the present system, a low yard partition 2 is assumed, or at least a detector system in second line, in which entrances to the animal shed 4 are monitored especially.

Said entrances to the animal shed 4 are in this case located at the gate 6 which is movable in the direction of the double arrow, as well as the door 7. The gate 6 serves to gain access to the feeding passage 13 in order to be able to provide roughage 16. Said roughage 16 is accessible to, in this case, dairy animals 9 via the feeding fence 14. In the feeding passage 13, the roughage 16 is pushed back by means of an autonomous vehicle 15, the feed pusher, such as the Lely Juno®. Obviously, the roughage 16 may also be swept back by hand or the like, but the invention offers advantages if functions on the livestock-farming system 1 are automated.

The wall 5 furthermore has a door 7, via which, for example, a farmer, but also livestock animals 9 may gain access to the shed space 8.

Access to the farm yard 3 is furthermore possible via the driveway 19, on which in this case there is a vehicle 20. The vehicle 20 has a licence plate 21 at the front, and obviously often also at the rear. Furthermore, the vehicle 20 is provided with a beacon 22, such as a GPS beacon, which is often used, for example, in rental cars and lorries. By means of such a beacon 22, it is possible to detect the vehicle from a distance. In order to monitor the livestock-farming system 1 , a detector system according to the invention is provided, which in this case comprises, inter alia, the cameras 17-1 to 17-5. These each have a detection zone 18-1 to 18-5. The first camera 17-1 is provided at the movable gate 6 and can thus monitor access via this gate 6. The camera 17-4 is provided at the door 7 and can thus monitor access via the door 7. Furthermore, a third camera 17-3 monitoring the milking robot 10 and a fourth camera 17-2 monitoring the shed space 8 are provided in the animal shed. Finally, the fifth camera 17-5 is provided which monitors the driveway 19. Miniature cameras can be provided on the animals 9, although they have not been shown.

Monitoring the livestock-farming system 1 not only serves the purpose of, for example, preventing theft. It may also be important to record which entities, in particular which individuals or animals, have visited the livestock- farming system. In case of contagious diseases, it may be very important to have an overview of these visits and contacts. Such a contact or visit may take place directly by entering through the door 7, but may also take place indirectly via the feed, by a visit via the gate 6 or by a visit on the farm yard via the driveway 19.

Each of the cameras 17-1 to 17-5 is connected to or provided with an image-recognition and image-recording system, also referred to here as the control unit 24. The images from, for example, camera 17-1 , are transmitted to the control unit 24 and received there by the communication device 25, such as a wifi connection or also a cable connection. In the control unit 24, the received images are processed by the data processor 26 which contains, for example, optical image-recognition software. In this case, the data processor 26 may use data stored in the database 27, such as with regard to shapes, colours and the like of known objects. Known objects include, for example, figures, colours, faces and the like of employees who work on the livestock-farming system 1 . Vehicles which are in use at this location may also be stored, such as one or more cars, tractors, autonomous vehicles, etc. In addition, data relating to regular visitors, such as suppliers, veterinarians, etc., may also be stored. In addition, general data relating to entities which could possibly pay a visit may be stored, such as people in general, other livestock animals, dogs, cats, foxes, etc. The latter data therefore relate more to objects or general shape data than to specific individuals and the like.

In addition to data regarding shape and colour, it is also possible to store specific data, such as a licence plate of a vehicle, a logo of a supplier, barcodes or other codes. In the illustrated example, the vehicle 20 has a licence plate 21 . The fifth camera 17-5 can record the licence plate 21 and the control unit 24 is then able to recognise the vehicle in question. Thus, a visit by the vehicle 20 can be registered, including the time of arrival and/or departure. Alternatively or additionally, the vehicle may for example be provided with a beacon 22 which can be detected by the beacon detector 23. Thus, a visit by the vehicle in question an also or alternatively be registered using the beacon detector 23. For example, the vehicle 20 may be a milk tanker of a company which is known to the farmer, with a known logo and/or with a known licence plate 21 . To this end, corresponding reference data may be stored in the control unit. After detecting this vehicle 20, the control unit may emit a cleaning start signal to the milking robot 10 in order to initiate cleaning of the milking robot by a cleaning device (not shown here), and a setting signal to prepare a milk reservoir (likewise not shown here) which is connected to the milking robot 10 for transferring the milk to the milk tanker 20 by pumping. As a result thereof, a time saving is achieved, both for the milking robot and for the milk tanker. It should be noted that, if it takes the milk tanker a certain amount of time to reach a position for pumping over after detection, for example due to the large distance between the driveway and milk reservoir, the two signals may also be emitted after a predefined period of time.

If a person, animal or the like has been able to gain access to the farm yard 3 without having been noticed by, for example, the fifth camera 17-5, the respective entrances to the shed space 8 are still monitored by the first camera 17-1 and the fourth camera 17-4. These cameras can also process the used images and/or transmit them to the control unit 24. This provides a second check with regard to access to animals and the shed space 8 in general.

If a person, animal or the like still gains access to the shed space 8 without having been detected, there are still the second camera 17-2 and the third camera 17-3 which can perform a third check on visits by said person, animal or the like. Obviously, different numbers of cameras may be provided, as known per se from burglar alarms and the like.

If a person or animal or another object is detected which cannot be identified, that is to say whose unique identity cannot be determined because no licence plate, logo, beacon or recognised face or the like is present, then it is still possible to register the type of object, such as for example "human", "dog", etc. To this end, it is possible, for example, to provide certain ranges relating to the detected data, such as height, speed, etc.

If the type of object cannot be detected either, but also, in addition, in all other cases where this is indeed possible, the images or such signals may be stored in, for example, the memory 28. Not only may, for example, an operator attempt to subsequently determine a possible identity, but it may also serve as evidence in case an investigation into, for example, a contagious disease or theft is conducted. Storing data is known per se, for example from security systems.

As has been described above, the detector system is configured to recognise objects, people and the like, either by means of individual identification or as a type of object. The livestock-farming system, in particular the control unit 24, is preferably furthermore configured to perform one or more animal-related operations on the basis of the received signals and/or the detected object types or identities. An example of such an operation is stopping the feed pusher 15 if, for example, the first camera 17-1 has detected a visit by a person or the like via the gate 6. This obviously serves to provide safety to the visitor in question.

Another measure relates, for example, to the operation of the milking robot 10, for example in order to stop it for the same reason as that mentioned above. Alternatively, the robot fences 1 1 may be opened, for example if an animal or the like visits the shed space 8, which could result in undesirable unrest among the livestock 9, in particular an animal in the milking robot 10. By opening the robot fences 1 1 , an animal is then able to escape from the milking robot 10, and thus injuries and damage are prevented.

Claims

1 . Livestock-farming system, comprising an animal space, such as an animal shed or a meadow, for keeping at least one livestock animal, and a detector system, wherein the detector system comprises

- a detector, and

- a control unit which comprises a memory and is operatively connected to the at least one detector,

wherein the detector system is configured to detect an entity within a predefined detection zone which comprises a part of said animal space and/or a part of adjacent surroundings,

wherein detecting comprises receiving a signal from the detection zone using the detector, which signal is suitable for recognising the entity,

and wherein the control unit is configured to process the received signal to yield detection data relating to the entity, and to store the detection data in the memory.

2. Livestock-farming system according to Claim 1 , wherein the control unit is configured to store the received signal.

3. Livestock-farming system according to any of the preceding claims, wherein the detector comprises a detector of electromagnetic signals, in particular a GPS tracker.

4. Livestock-farming system according to Claim 3, wherein the detector comprises an optical camera for capturing optical images, wherein the control unit furthermore comprises an image-recognition system which is configured to analyse the received images and to recognise at least the object type of the detected entity, and in particular to determine an identity of the detected entity, in the analysed images by comparison with reference data stored in the image-recognition system.

5. Livestock-farming system according to Claim 4, wherein the detector is attached to a part which is to be worn by one of the animals, such as a collar or ear tag.

6. Livestock-farming system according to Claim 3, 4 or 5, wherein the control unit is furthermore configured to store at least the recognised object type, and in particular the determined identity, together with or in the detection data.

7. Livestock-farming system according to Claim 4, 5 or 6, wherein the reference data stored in the image-recognition system comprise one or more of the following data types: vehicle licence plates, faces of people, logos and barcodes.

8. Livestock-farming system according to any of the preceding claims, wherein the livestock-farming system comprises at least one device for executing an animal-related operation on the animals, as well as a program for controlling the device, wherein the program is configured to control the device partly on the basis of the detection data.

9. Livestock-farming system according to Claim 8, wherein the at least one device comprises at least one autonomously movable part, such as a milking robot, an automatic feeding device, a manure slide or an autonomously moving vehicle.

10. Livestock-farming system according to Claim 8, wherein the device comprises a milking robot with a cleaning device, as well as a milk reservoir which is connected to the milking robot for collecting milk from the milking robot, wherein the control unit is configured to emit a cleaning start signal to the cleaning device for cleaning the milking robot, and is configured to emit a setting signal to the milk reservoir to prepare the milk reservoir for transferring milk in the milk reservoir to a milk tanker by pumping, and wherein the control unit is configured to automatically emit the cleaning start signal and the setting signal if, and preferably either as soon as or with a predefined time delay, the detection data correspond to a recognised milk tanker.