A method for operating a robotic arm, comprises determining a speed of rotation of a rotary milking platform, the rotary milking platform having a stall for a dairy livestock. The method continues by moving a carriage along a track positioned adjacent to the rotary milking platform at a rate that is based at least in part upon the determined speed of rotation of the rotary milking platform, wherein the carriage moves independently of the rotary milking platform and in a direction corresponding to a direction of rotation of the rotary milking platform. The method continues by extending a robotic arm that is coupled to the carriage between the legs of the dairy livestock, wherein the robotic arm remains extended between the legs of the dairy livestock for a period of time as the stall rotates adjacent to the robotic arm. The method concludes by retracting the robotic arm from between the legs of the dairy livestock as the stall rotates adjacent to the robotic arm.

A spray tool coupled to a robotic arm that extends between the legs of dairy livestock, comprises a linear member and a plurality of spray nozzles. The linear member rotates about an axis that is perpendicular to the robotic arm, and has a perimeter that lies within an outer perimeter of the robotic arm when the robotic arm extends between the legs of a dairy livestock. The plurality of spray nozzles are coupled to the linear member.

A method for operating a robotic arm comprises determining a speed of rotation of a rotary milking platform, the rotary milking platform having a stall for a dairy livestock. The method further includes moving a carriage along a track positioned adjacent to the rotary milking platform at a rate that is based at least in part upon the determined speed of rotation of the rotary milking platform. The method further includes extending a robotic arm that is coupled to the carriage between the legs of the dairy livestock.

A spray tool coupled to a robotic arm includes a linear member, a first spray nozzle and a second spray nozzle. The linear member rotates about an axis that is perpendicular to the robotic arm. The linear member has a perimeter that lies within an outer perimeter of the robotic arm when the robotic arm extends between the legs of a dairy livestock. The first spray nozzle is coupled to the linear member proximate a first end of the linear member. The second spray nozzle is coupled to the linear member proximate a second end of the linear member.

A system includes a camera, a controller and a robotic arm. The camera generates an image of an udder of a dairy livestock. The controller determines a position of the udder of the dairy livestock based at least in part upon the image. The controller further determines a spray position based at least in part upon the determined position of the udder of the dairy livestock. Determining the spray position includes processing the accessed image to determine a tangent at the rear of the udder and a tangent at the bottom of the udder. The spray position is a position relative to the intersection of the two tangents. The robotic arm is communicatively coupled to the controller and positions a spray tool at the spray position.

A controller controls an end-effector of a milking robot to move to a desired position (p.sub.set) according to a desired velocity profile via: a feedforward module producing predicted control signal(s) (c.sub.pred) based on a set vector (v.sub.set) specifying the desired velocity profile, a closed-loop controller, based on a modified position (p), producing primary control signal(s) (c.sub.prim) for controlling the end-effector to the desired position (p.sub.set), and first and second summation modules deriving modified control signal(s) (c.sub.input) to be fed to the milking robot and deriving the modified position (p) respectively. The feedforward module contains a trained artificial neural network with an input layer configured to obtain the set vector (v.sub.set), an output layer configured to provide the at least one predicted control signal (c.sub.pred), and a number of hidden layers interconnecting the input layer and the output layer. The respective nodes in said layers have weights that were assigned through a training process in which output signals (p.sub.out) from the robot were used as training data and registered control signals for controlling the end-effector of the milking robot were used as reference data.