The dairy industry has continued to innovate to meet the needs of the consumers' specification of milk at a low price, of good hygienic quality and with rising expectations of animal welfare. The introduction of robotic milking offers the opportunity for the cost effective deployment of novel sensors for a variety of milk analytes. Traditional methods of monitoring health changes in animals are based entirely on the human senses. However, in modern milking systems humans rarely have enough time to see the cows to observe for signs of ill health, the extreme case is that of robotic milking. Novel sensors will allow closed loop control systems where the early detection of deviations from optimal performance will enable the farm manager to make management decisions before damage to potential milk yields is irreversible. Where a biological model already exists, for example, in the prediction of ovulation with milk progesterone analysis, rapid progress is being made towards an automated prediction system. Integrated management systems for dairy cows will not only have the traditional goals of efficient milk production but can also be tuned to reduce polluting outputs of ammonia, phosphorus and methane. The main metabolic markers in milk to be monitored are urea, fat, ketones and protein. The detection of mastitis can be achieved by the development of sensor systems to detect enzyme markers of inflammatory response such as Nagase. Multi-disciplinary research is needed to develop integrated management systems drawing all the different elements of dairy cow management into a single system. The major cause of death in dairy cows is dystocia and monitoring systems are needed to ensure that parturition is better managed.