Epsilon Archive for Student Projects

Abstract

The trend in milk production is similar in Sweden and across the world; the farms are getting fewer but increasingly bigger. The costs for hiring workers are increasing, which results in more cows per animal keeper. Sensors can then be a great tool for keeping track of animals which need special attention. With a reported decreasing fertility, sensors that can detect oestrus even at low levels can be especially helpful in bigger herds to pinpoint the cows that may be in oestrus. The number of farms with dairy cows that install AMS, Automatic Milking Systems, are increasing, and that also increases the need for sensors to control milk quality and composition when there is no visual control of the milk by a milker before milking starts.

For the sensor data to be useful you have to have software designed to analyze and interpret the data to get relevant information that the farmer can use. A lot of companies that make sensors and systems for the sensors also design a management system to gather information from all systems for a better overall view and herd management.

The project has been conducted as a review where information and research reports on sensors and systems has been gathered, in order to make a background for further research at the Department of Rural Buildings and Animal Husbandry.

There are many sensors on the market today. Most of the sensors used today are designed to detect mastitis, abnormal milk or oestrus. Since the major reasons for culling cows are udder health related problems or low fertility rate, the economic gain in research and development for sensors that can detect the problem early on and diminish the costs are obvious. A management system that is being introduced is Smart Farming, by DeLaval, that besides milking, feeding and fertility control, which are obvious in a dairy farm management system, also focuses on environmental sustainability and optimize the farms production to get a more effective company. It also holds possibilities for advisors to gain access to several farms data and help the farmers to get better production and profitability.

Many of the companies that sell milking equipment have their own sensors and analyzing systems that can be combined in their respective management systems. The components in the respective companies’ management systems are similar, indicating the demand from the farmers.

Critical parts of how well the system will perform are the settings of threshold values and how easy the system is to use. With low threshold value you get more false positives but with a high threshold value you risk missing early signs of for example mastitis. User friendly systems would probably determine how much the farmer will use the system and how much information one would choose to get from the system.

The conclusions are that there are different kinds of limitations in the sensors. The limitations can be roughly divided as technical, system related and economic. There is a constant development and tomorrows sensors will hopefully have fewer limitations. More sensors to detect lameness and metabolical disorders are desired. The systems that are being launched are of a network structure where farmers and advisors can work together for a better milk production. Sensors can never replace a good animal keeper, but it can help them pinpoint the cows that need extra attention.