In mid 1988 the research council and the UK Ministry of Agriculture set up a robotic milking project, partly out of concerns that systems being developed would damage animal health and that they were not well adapted to the UK’s grazing systems. Toby Mottram sketched out a vision (see right) of how a system could look which was very different from the direction taken by other groups.
A Review and Design of Robotic Milking Systems, Mottram, 1992 listed the challenges we anticipated, some of which (cow and milk hygiene, mastitis detection, cow traffic management) still remain. You can see how we overcame the problems in the video Milking time below.
As the 1990s progressed various systems were launched and gained grudging acceptance within the industry, particularly in the Netherlands and Denmark where the farm size and management is suited to minimal grazing and animals housed for many months each year. In 1997 a committee of the IDF reviewed the situation and reported field experience Automatic Milking Review IDF2000.
One of the key problems was how to get cows to the robot. In Motivation of Cows to be Milked in a Robotic Milking System, published 1997 Neville Prescott and Toby Mottram showed conclusively the main motivation for cows is to eat and even cows with full udders will not walk to milking unless food is offered in or after milking. Giving cows feeding during milking causes them to move more than if there is no feed a simple solution is to reserve the feed for the exit area of the parlour. Food-Type-and-Location-Effect-on-Robotic-Milker-Attendance-Effect-of-Feeding-During-Milking-on-Behaviour-1998
We chose to build a lightweight one armed robot that could track cow movements, we calculated the transfer functions for the robot response in a neat little experiment TeatCupAttachment Frost 1993.
This technique has shown itself to be highly adaptable and with the launch of the deLaval robot on a rotary the ideas we demonstrated are coming to fruition.
Maintaining clean milk from clean cows is a major problem we proposed using an Optical Teat Inspection for Robotic Milking 1995 to detect chlorophyll left on cows teats by looking for the optical absorption but sadly no company provides a cow inspection system.