We are living in adigital world, where digital devices are increasingly affecting ourlives, by handlingdifferent tasks and making our lives easier. Wearables have also been a part ofthe digitalization trend and process in the last years, not only for humans butfor animals as well. The automation of processes can especially help dairyfarmers to reduce physical labour and labour costs, considering the increase of farmsizes and the need of higher-yielding cattle. (de Koning, 2010) This review delivers an organised overview of thewearable sensors developed to assist with three different health issues ofdairy animals like mastitis, fertility and locomotion. Before starting with theoverview and comparison of the sensor systems developed for each health issue,it is useful to describe the framework with the four levels of a sensor system,where the comparison is based on. The first level is technique, which describes the different types of sensors thathave already been developed. In the second level named data interpretation, is shown if there is an existing algorithm in the system that processes the data gatheredfrom the sensor and produces information by determining the changes in it.
Thethird level, integration of information, presents a decision support model whichcombines the information produced by the system with external data coming forexample from the advisor or farmer and adviseshow to act upon the detected event. The fourth and last level, decision makingdetermines if the decision is taken by the farmer or autonomously by theadvanced system sensor. (Rutten, 2013) Starting with the firstdisease, mastitis is inflammation of the mammary gland in the breast due tobacterial infection.
It is the most common and most costly disease of dairycattle.(McGill, 2003) For sensor systems detecting mastitis, 31 publicationswere found where 37 sensor systems are described. 48% of the sensors found useelectrical conductivity, 23 % combine electrical conductivity and milk colour sensors – inline and 16 % are biosensorswhich detect different enzymes. 92 % of the sensors found interpret the gathered data, but they do notintegrate external information for the decision making part. 43 % of them give a mastitis alert to the farmer, 32 % showthe probability of the disease as well, whereas only 5 % degree and classify the disease. (Rutten, 2013)The second health issue thatis being contemplated is fertility.
61% of the sensors study the activity ofthe cow, 15% of them are biosensors and measure the progesterone levels in milkand another 15% consider their mounting behaviour. 75% of the sensor systemsdeveloped include data interpretation, but there is no integration of externalinformation for decision making. 60 % of them give an estrus alert, but only15% of them show the probability of the estrus alert as well.
(Rutten, 2013) Theobjective of sensors developed for fertilityissues is to prevent missing cases of true estrus caused by using visualobservation. (Firk et al., 2002) The oldest and most studied sensor is thepedometer. (Williams et al., 1981) A pedometer counts each step the animaltakes by detecting the motion of its hips and legs. Pedometers can detectestrus accurately and appear to be a promising tool for prediction of ovulationand hence could be a tool for improving fertilization rates. (Roelofs,2005)Lameness is considered to be thethird most influential health issue, after mastitis and fertility issues.
Researchershave observed the frequency of this disease in 21 farms and the physicaleffects caused by it on 2183 cows. The final result showed that a lame cowcauses a loss of 110 € per foot per year. (Enting, 1997) Pedometers & 3Daccelerometers are the most studied sensors. (Mol, 2009) Accelerometers measurethe overall dynamic body acceleration. (Wilson, 2006) 45 % of the sensorsdeveloped to detect lameness study the weight distribution between the legs, 42% study their walking behaviour and 13 %study their walking activity.
However, only half of the systems can interpretthe data gathered from the sensors, whereas the rest shows only raw and unclearresults. (Rutten, 2013) Inconclusion, the development of wearables for dairy animals is still in itsearly stages. There have been a lot of research and development done on thetechnique and data interpretation levels, especially for mastitis and fertilityissues, whereas sensors helping with locomotion detection are still in the technique development phase. Therefore, the alertscoming from the wearables for mastitis and estrus are more informative than thelocomotion alerts. (Rutten, 2013) It is also still unclear if the locomotion sensors can detect lameness at anearly stage.
(Juarez et al., 2003). Based on the findings ofthis literature review, we can observe that this topic is still quiteinnovative and there is still a lot of room for improvements and new developments.