A robotic arm maneuvers a teat preparation cup and executes instructions from a robotic arm controller. The controller comprises an interface, a memory, and a processor. The processor instructs the sensor to perform a first scan. If the first scan discovers a first set of teats, the processor moves the robotic arm a first distance and instructs the sensor to perform a second scan. If the second scan discovers a second set of teats, the processor moves the robotic arm to a location under the first teat, and instructs the sensor to perform a third scan. The processor determines if the third scan discovers a third set of teats. If each of the first set, second set, and third set of discovered teats comprises the first teat, the processor instructs the robotic arm to attach the preparation cup to the first teat.

A method for use with a multibody system includes two body sections assembled via a joint, one of the body sections including a camera which is moveable between a first position and a second position, includes detecting a reference object with the camera situated in the first position, determining a reference value of the reference object, requesting movement of the camera from the first position into the second position, initiating movement of the camera from the first position towards the second position, while iteratively sampling image of the reference object, determining a current object value of the reference object, based on the sampled image, matching the reference value with the current object value, and determining successfulness of the camera movement, based on the matching.

A robotic attacher retrieves a preparation cup from an equipment area located behind a dairy livestock and attaches and detaches the preparation cup to the teats of the dairy livestock in sequence. The sequence comprises attaching and detaching the preparation cup to the left front teat, the right front teat, the right rear teat, and the left rear teat.

In picking up tools in an automatic milking arrangement, the positions of the tools are determined by registering, via a camera at an origin location, three-dimensional image data representing the tools whose positions are to be determined. Using an algorithm involving matching the image data against reference data, tool candidates are identified in the three-dimensional image data. A respective position is calculated for the tools based on the origin location and data expressing respective distances from the origin location to each of the identified tool candidates. It is presumed that the tools have predefined locations relative to one another. Therefore, any second tool candidate is disregarded, which is detected at a separation distance from a first tool candidate when the separation distance exceeds a first threshold distance in relation to the predefined relative locations.

The positions of the tools in an automatic milking arrangement are determined by registering, via a camera at an origin location, three-dimensional image data representing the tools whose positions are to be determined. Using an algorithm involving matching the image data against reference data, tool candidates are identified in the three-dimensional image data. A respective position is calculated for the tools based on the origin location and data expressing respective distances from the origin location to each of the identified tool candidates. It is presumed that the tools are arranged according to a spatially even distribution relative to one another. Therefore, any tool candidate is disregarded, which is detected at such a position that the position for the candidate deviates from the spatially even distribution.

Systems and methods are described for selecting a sound type of interest from a first (e.g., master/global) machine learning library comprising information derived from reference audio stream data acquired from a plurality of farm animal operation reference sound monitoring events, including from a first farm animal operation monitoring event of a first farm animal operation, wherein the sound type of interest is associated with a condition state of interest of a first collection of farm animals. Further, information associated with the selected sound type of interest can be included a second machine learning library, wherein the second machine learning library is operational on an edge computing device located in proximity to a second farm animal operation. Audio stream data can be acquired from the second farm animal operation in a second farm animal operation monitoring event, and processed using the second machine learning library information to determine whether the sound type of interest is present in the acquired audio stream data, thereby generating information associated with the presence or absence of the condition during the second farm animal operation monitoring event.

A milking system with a milking device, a milking control, a milk line in fluid connection with the milking device, and a sampling and analysis device to take a sample of the milk from the milk line and to analyse milk from the sample are disclosed. The milking control is arranged to control the milking device on the basis of the analysis. The sampling and analysis device includes a control unit, a tape reel provided with a tape that is lengthwise provided with a series of consecutive reagent pads that provide a detectable response in the presence of a substance in the sample, a tape mover to move the tape, a dosing device to supply a part of the sample onto a reagent pad on the tape, and a camera to obtain an image of the reagent pad supplied with the droplet, and an analysis device to analyse the obtained images to provide to the milking control device an indication of a presence or concentration of said substance. The camera has a field of view that contains a plurality of reagent pads of the series of consecutive reagent pads. This allows to observe the reaction in the reagent pad for a much longer time. In turn, this allows to use much less reagent material, such as expensive enzymes, in the pads. It is particularly useful when observing double layer reagent types.

Systems and methods are described for selecting a sound type of interest from a first (e.g., master/global) machine learning library comprising information derived from reference audio stream data acquired from a plurality of farm animal operation reference sound monitoring events, including from a first farm animal operation monitoring event of a first farm animal operation, wherein the sound type of interest is associated with a condition state of interest of a first collection of farm animals. Further, information associated with the selected sound type of interest can be included a second machine learning library, wherein the second machine learning library is operational on an edge computing device located in proximity to a second farm animal operation. Audio stream data can be acquired from the second farm animal operation in a second farm animal operation monitoring event, and processed using the second machine learning library information to determine whether the sound type of interest is present in the acquired audio stream data, thereby generating information associated with the presence or absence of the condition during the second farm animal operation monitoring event.

A system, apparatus, and/or method of determining a caloric requirement of an animal may be provided. In an aspect, animal characteristic data comprising a breed of an animal or a condition of the animal may be received. Durational parameters of the animal for a first predetermined time duration may be received. The durational parameters may include at least one of an activity level of the animal, a weight of the animal, or an amount of calories consumed by the animal during the first predetermined time period. A caloric requirement of the animal during a second predetermined time period may be determined. The caloric requirement of the animal may be based on the animal characteristic data of the animal and/or the durational parameters of the animal during the first predetermined time period. The caloric requirement of the animal during the second predetermined time period may be displayed via a display device.

Systems and methods are described for selecting a sound type of interest from a first (e.g., master/global) machine learning library comprising information derived from reference audio stream data acquired from a plurality of farm animal operation reference sound monitoring events, including from a first farm animal operation monitoring event of a first farm animal operation, wherein the sound type of interest is associated with a condition state of interest of a first collection of farm animals. Further, information associated with the selected sound type of interest can be included a second machine learning library, wherein the second machine learning library is operational on an edge computing device located in proximity to a second farm animal operation. Audio stream data can be acquired from the second farm animal operation in a second farm animal operation monitoring event, and processed using the second machine learning library information to determine whether the sound type of interest is present in the acquired audio stream data, thereby generating information associated with the presence or absence of the condition during the second farm animal operation monitoring event.