Automatic milking of cows has been around for more than 30 years, making the milking process more efficient, less labour intensive and more cow friendly. Before a cow can be milked, three stages of udder pretreatment have to be completed: cleaning, stimulation and premilking. There is no clearly superior pretreatment technique for automatic milking systems and the three main manufacturers of milking robots (Lely, DeLaval and GEA) each use different systems. Lely uses two counter rotating brushes that the teat is inserted in between, with good cleaning and stimulation results. However, their R&D department is looking for innovative alternatives. The goal of this thesis is therefore to find an innovative alternative for Lely’s current pretreatment system with at least as good cleaning and stimulating properties. Preferably the premilking stage is also included, which Lely’s current system is unable to perform in contrast to its competitors. A morphological chart is used to design three concepts. Prototypes are then evaluated based on a field test with a real cow and a cleaning test using a fake test-teat. Based on a weighted Harris profile, the Origami Vacuum Gripper concept was chosen to further iterate into a final design and prototype. The device is essentially a gripper that functions like a foldable milking cup that sprays, grips and then brushes the teat by being pulled downward off the teat. The folding is based on the Waterbomb origami pattern and is actuated using vacuum and an airtight skin that is wrapped around it. A brush liner is positioned inside the gripper and the dirt is removed at the bottom through a vacuum tube. A final field test showed that the device fulfilled all requirements, being able to properly clean and stimulate the teats and leaving them dry enough to attach the milking cups. Unfortunately, the device was not able to premilk the cow. The compression of the teat by the gripper prevented the milk from being pulled out by the vacuum inside the gripper. Finally, a gripping force test is performed to see the effect of different origami skeleton geometries on the gripping force applied to the teat. It was concluded that the skeleton with shorter folding edges applies a larger force to the teat with the same vacuum applied to the airtight skin. More research is needed on the folding geometries in order to find an optimal design for the gripper.