Search

Die Autorin wurde zwei Jahre durch ein Promotionsstipendium der Universität Kassel gefördert.

Milking personnel have been affected by musculoskeletal disorders for many years. In parlor workers the shoulders, wrists and lower back are most affected. As the procedure of milking cluster attachment has been reported to be the most strenuous during milking, we took a three- step approach to reduce the physical strain of this task. In the first step, we used the computer-assisted recording and long-term analysis system (CUELA) to record flexion angles of multiple joints during milking. The posture of 30 milkers was evaluated on 15 farms. Milking parlor types evaluated included the Herringbone 30°, the Herringbone 50°, and the Parallel as well as the Rotary parlor. The 5th, 50th and 95th percentiles of the data recordings were classified against ISO Norms and it was found that joint flexion angles were concerning. The statistical analysis revealed a significant interaction between milking parlor type and a working height coefficient, which reflected the ratio between the subject’s height and the effective udder height (udder height + depth of pit). By using model predictions, we calculated working height coefficients that could improve joint flexion angles. These working height coefficients were calculated for each parlor type and used within the newly developed “milking health formula” to calculate the ideal depth of pit, under the consideration of the milker’s height, milking parlor type and udder height. As the working heights recommended within the milking health formula were relatively low for all parlor types, and the recommendations made for the Herringbone 30° were broad, we aimed to further validate our findings by using surface electromyography to monitor muscle contraction intensities of 16 milkers (nine females, seven males). The second step of this thesis was performed in a laboratory setting where the milking cluster was attached to an artificial udder. It was important to ensure that the milking health formula enabled a consistent setting of working heights for milkers of different body heights, as well as ensuring that lower working heights reduced muscle contraction intensities of the upper limb and shoulder muscles. The results showed that lower working heights decreased muscle contraction intensities of the shoulder muscles, but not of the lower and upper arms. Further, since the subjects body height had no effect on muscle contraction intensities, it can be concluded that the formula offers an effective way to set comparable working heights for milkers of different body heights. Posture of milkers is not only affected by working heights, but also by the horizontal reaching distance between the milker and the cows’ udder. It has recently been assumed that milking stall dimensions are currently too small for dairy cows and that they should be increased to ensure their welfare. This could however increase the reaching distance between the udder and the cow and thus negatively affect ergonomics. In the third step of the thesis, we therefore used surface electromyography, in both a Herringbone 30° and a Side by Side milking parlor, to investigate the effect of increased milking stall dimensions on muscle contraction intensities of the upper limb and shoulder muscles during milking. Nine male subjects milked 30 cows twice per parlor type, where the milking stall dimensions were large on one side of the milking parlor and standard sized on the other. Milking stall dimensions had no effect on muscular contraction intensities in the Side by Side parlor and a controversial effect in the Herringbone 30° parlor. The contraction intensities in the right lower and upper arm were higher when cows were milked in standard sized milking stalls, but were higher in the left upper arm when cows were milked in large milking stalls. The effect of milking stall dimensions on the work environment should therefore be further investigated. In conclusion, the current project has developed a method to calculate beneficial working heights for a variety of milking parlor types. These derived recommendations have been further validated and it was shown that lower working heights reduced muscular load of the shoulder muscles. Seit vielen Jahren werden bei Melkern und Melkerinnen Erkrankungen des Bewegungsapparates festgestellt. Bei Melkern, die in Melkständen arbeiten sind die Schultern, Handgelenke und der untere Rücken am stärksten von diesen Beschwerden betroffen. Das Ansetzen des Melkzeugs ist beim Melken die anstrengendste Tätigkeit. Deshalb versuchten wir die Arbeitsbelastung des Melkpersonals bei dieser Tätigkeit in einem dreistufigen Ansatz zu reduzieren. In einem ersten Schritt verwendeten wir das computerunterstützte Erfassung- und Langzeit-Analyse (CUELA) System, um die Beugewinkel verschiedener Gelenke beim Melken zu erfassen. Auf 15 Landwirtschaftsbetrieben wurde hierzu die Haltung von insgesamt 30 melkenden Personen untersucht. Die Untersuchung umfasste die Melkstandtypen Fischgräte 30°, Fischgräte 50°, Side by Side und Melkkarussell. Das 5te, 50te und 95te Perzentil der erfassten Daten wurde jeweils nach ISO-Normen klassifiziert. Hierbei wurden besorgniserregende Beugewinkel der Gelenke festgestellt. Die statistische Analyse ergab eine signifikante Wechselwirkung zwischen dem Melkstandtypen und einem Arbeitshöhekoeffizienten, der sich aus der Körpergrösse der melkenden Person und der effektiven Höhe des Euters (Euterhöhe + Tiefe der Melkgrube) ergibt. Mit Hilfe von Modellschätzungen bestimmten wir Arbeitshöhekoeffizienten, durch die die Beugewinkel der Gelenke verbessert werden konnten. Diese Arbeitshöhekoeffizienten wurden spezifisch für die einzelnen Melkstandtypen berechnet und konnten in der neu entwickelten “Milking Health Formel” eingesetzt werden, um die ideale Melkgrubentiefe in Abhängigkeit der Körpergrösse der melkenden Person, des Melkstandtyps und der Euterhöhe zu berechnen. Da die durch die Milking Health Formel empfohlenen Arbeitshöhen für alle Melkstandtypen relativ tief und die Empfehlungen für den Fischgräten 30° Melkstand noch unpräzise waren, prüften wir in einem zweiten Schritt unsere Ergebnisse durch den Einsatz von Oberflächen-Elektromyographie. Wir untersuchten hierzu die Intensität der Muskelkontraktion von 16 melkenden Personen (9 Frauen, 7 Männer) unter Laborbedingungen. Die Probanden setzten das Melkzeug hierzu auf drei unterschiedlichen Arbeitshöhen an einem Kunsteuter an. Die Milking Health Formel ermöglichte eine konsistente Einstellung der Arbeitshöhen für melkende Personen unterschiedlicher Körpergrössen. So konnte untersucht werden, ob die niedrigeren Arbeitshöhen zu einer Entlastung der oberen Extremitäten und Schultern des Melkers führen. Die Ergebnisse zeigten, dass die tieferen Arbeitshöhen die Intensität der Muskelkontraktion der Schultern, nicht jedoch der Unter- und Oberarme senkten. Da die Körpergrösse des Melkers weiterhin keinen Einfluss auf die Intensität der Muskelkontraktionen hatte, konnten wir schlussfolgern, dass die Formel es effektiv ermöglicht vergleichbare Arbeitshöhen für Personen unterschiedlicher Körpergrösse einzustellen. Im dritten Teil dieser Dissertation untersuchten wir mit Hilfe von Oberflächen-Elektromyographie, sowohl im Fischgräten 30° Melkstand als auch im Side by Side Melkstand, den Einfluss vergrösserter Melkplatzabmessungen auf die Intensität von Muskelkontraktionen in den oberen Extremitäten und Schultern. Dieser Versuch wurde durchgeführt, da ein Zusammenhang zwischen dem horizontalen Abstand der melkenden Person zum Kuheuter und der Ergonomie in Melkständen festgestellt wurde. Zudem wurde die Vermutung geäussert, dass die derzeitigen Melkstandabmessungen für die Kühe zu klein sind und im Hinblick auf das Wohlbefinden der Milchkühe vergrössert werden sollten. Grössere Melkplätze könnten jedoch dazu führen, dass die horizontale Entfernung zwischen Melker und Kuh sich vergrössert und die Ergonomie des Melkers somit beeinflusst. In der dritten Studie melkten daher neun Personen 30 Kühe je zwei Mal im Fischgräten 30° und zwei Mal im Side by Side Melkstand. Hierbei waren die Melkplatzabmessungen der Kühe jeweils auf einer Seite des Melkstands gross, während Sie auf der anderen Seite Standardmass hatten. Die Melkplatzmessungen hatten beim Side by Side Melkstand keinen und beim Fischgräten 30° Melkstand einen widersprüchlichen Einfluss auf die Intensität der Muskelkontraktionen. Die Intensität der Muskelkontraktionen im rechten Unter- und Oberarm waren höher, wenn die Kühe in Standard Melkplätzen gemolken wurden, während im linken Oberarm die Intensität der Muskelkontraktionen höher waren, wenn die Kühe in grossen Melkplätzen gemolken wurden. Der Einfluss der Melkplatzabmessungen auf die Arbeitsbedingungen im Melkstand sollte daher noch genauer untersucht werden. In der vorliegenden Arbeit wurde eine Methode entwickelt, mit der ergonomisch vorteilhafte Arbeitshöhen für unterschiedliche Melkstandtypen berechnet werden können. Diese Empfehlungen wurden validiert und es wurde gezeigt, dass tiefere Arbeitshöhen die Belastung der Schultermuskulatur reduzierten.

During machine milking, farmers often encounter milking problems even though milking machine constructions generally comply with the required international standards. The first aim of this thesis was to investigate the effect of two different durations of the closing and closed phase (c- and d-phase, respectively) on physical processes in the milking cluster. The second aim was to examine the effect of these c- and d-phases on milk removal and hind-leg activity. It was hypothesized that a prolonged c-phase (i.e. slower liner closing) might be gentler and more comfortable for the dairy cow than a short c-phase (i.e. faster liner closing). Consequently, this would lead to optimized milk removal and calmer dairy cow behavior in the milking parlor. Therefore, dairy cows were confronted randomly with two types of pulsation chamber cycles (Treatments A and B) for 12 milkings. The treatments differed in the durations of c- and d-phases. In Treatment A, the c-phase lasted 70 ms and the d-phase 330 ms, whereas in Treatment B, the c-phase lasted 130 ms and the d-phase 270 ms. Using a vacuum measuring device (MT52, BEPRO AG, Güttingen, Switzerland), measurements were taken during milking proceedings. Milk flow characteristics were recorded using milk flow meters (LactoCorder®, WMB AG, Balgach, Switzerland). Hind-leg activity was recorded during milking using accelerometers attached on the hind-legs (RumiWatch® pedometer, ITIN+HOCH GmbH, Liestal, Switzerland). Treatment B showed the following effects on parameters measured in the milking cluster compared with Treatment A: -Durations of open liner were on average 26.2 ms longer. -Durations of closing liner were on average 23.4 ms longer. -Pressure sum was on average 1 kPa lower. Treatment B showed the following effects on milk flow and hind-leg activity compared with Treatment A: -Cows produced 0.21 kg higher total milk yield. -Peak flow rate was 1.04 kg/min higher. This thesis showed that c-phase durations influence physical processes in the milking cluster and milk flow characteristics. A prolonged c-phase is not only beneficial for dairy cow’s health but also from an economic point of view. Statements regarding welfare improvements are rather difficult to make. In addition to the first and second aim of this thesis, the third aim was to examine the correlations of hind-leg activity with accelerometers attached to the hind-leg and to the milking cluster with direct observations. As restlessness during milking is considered as an important parameter for impairments of dairy cow welfare, a standardized measuring procedure for this behavior may be severe for detection of deficient milking conditions. It was assumed that the milking cluster is set in motion as hind-legs of dairy cows move, because the milking cluster is freely suspended on the cow’s udder. Therefore, an accelerometer attached to the milking cluster may measure the hind-leg activity of the dairy cow indirectly. This method could replace laborious attaching of accelerometers on the hind-legs of dairy cows and provide a standardized on-line measuring procedure. Therefore, measurements with accelerometers on hind-legs of dairy cows and on the milking cluster were taken once during morning and evening milkings for every cow. In addition, direct observations of dairy cows’ motion behavior during milking were made. The differentiation of hind-leg activity took place between active phases and inactive phases. Data from morning milkings were used to create an algorithm to validate data automatically from evening milkings. The correlation measurements of hind-leg activity were as follows: - The algorithm is defined as the number of observations using mean values + standard deviation + 0.2 m/s2. -Correlation “visually counted number of active phases from graphs” with “number of observations calculated with the algorithm from the milking cluster” was 97 % (morning milkings). -Correlation “number of active phases recorded with milking cluster using the algorithm” with “number of active phases using direct observations” was 74 %. -Correlation “number of active phases recorded with the hind-leg using the algorithm” with “number of active phases using direct observations” was 91 %. It was possible to develop a hind-leg activity measuring method, without attaching accelerometers on the hind-legs of dairy cows. With this novel method, costs and labor can be minimized and objective examination of animal behavior can be guaranteed. In a next step, it can be implemented in the milking parlor or in the automatic milking device as a diagnostic tool providing valuable information to the farmer and consultant in a management program. Obwohl die Installationen der marktverfügbaren Melkanlagen generell mit den Normen des internationalen Standards übereinstimmen, bekunden Landwirte oftmals Melkprobleme. Das erste Ziel dieser vorliegenden Arbeit war die Analyse des Einflusses zweier unterschiedlich andauernden Belüftungs- und Druckphasen (C- und D-Phasen) im Pulszyklus, auf die physikalischen Vorgänge im Melkaggregat. Das zweite Ziel war die Untersuchung der Auswirkung dieser C- und D-Phasen auf den Milchentzug und dem Kuhkomfort. Es wurde angenommen, dass eine verlängerte C-Phase sanfter und angenehmer für das Milchvieh ist als ein schnellschliessender Zitzengummi. Eine verlängerte C-Phase könnte zu einem optimierten Milchentzug und zu ruhigeren Milchkühen während des Melkens führen. Um dies zu untersuchen wurden die Milchkühe in zufälliger Reihenfolge mit zwei unterschiedlichen Pulskurven (Behandlung A und B) während 12 Melkungen konfrontiert. In Behandlung A betrug die C-Phase 70 ms und die D-Phase 330 ms, während in Behandlung B die C-Phase 130 ms und die D-Phase 270 ms andauerte. Mit einem Vakuummessgerät (MT52, BEPRO AG, Güttingen, Schweiz) wurden Messungen an einem Melkzeug während des Melkens durchgeführt. Milchflussparameter wurden mit einem Milchflussmessgerätes (LactoCorder®, WMB AG, Balgach, Schweiz) erfasst, und die Aktivität der Hinterbeine wurde mit Beschleunigungssensoren (RumiWatch® Pedometer, ITIN+HOCH GmbH, Liestal, Schweiz) untersucht. Behandlung B zeigte verglichen mit Behandlung A folgende Auswirkungen auf die physikalischen Vorgänge im Melkaggregat: -Die Dauer des offenen Zitzengummis war im Durchschnitt 26.2 ms länger. -Die Dauer des schliessenden Zitzengummis war im Durchschnitt 23.4 ms länger. -Die Drucksumme war im Durchschnitt 1 kPa niedriger. Behandlung B zeigte verglichen mit Behandlung A folgende Auswirkungen auf den Milchfluss und die Aktivität der Hinterbeine: -Die Kühe erzielten eine durchschnittlich 0.21 kg höhere Gesamtmilchmenge. -Der „höchste Milchfluss“ war im Durchschnitt 1.04 kg/min höher. Es wird in dieser Studie deutlich, dass die Dauer der c-Phase einen Einfluss auf die physikalischen Vorgänge im Melkaggregat und den Milchfluss hat. Die verlängerte C-Phase ist nicht nur für die Tiergesundheit, sondern auch wirtschaftlich gesehen vorteilhaft. Aussagen über das Wohlbefinden der Kühe sind jedoch schwer zu formulieren. Zusätzlich zum ersten und zweiten Ziel dieser Studie, galt es als drittes Ziel, die Korrelationen von der direkt beobachteten Hinterbeinaktivität mit Beschleunigungssensoren an den Hinterbeinen und am Melkaggregat zu überprüfen. Es wurde vermutet, dass das Melkaggregat aufgrund der Bewegung der Hinterbeine in Bewegung gesetzt wird, da es frei am Kuheuter hängt. Um die Hinterbeinaktivität am Melkaggregat indirekt zu erfassen, wurde ein Beschleunigungssensor am Melkaggregat angebracht. Diese neue Methode der Hinterbeinaktivitätserfassung könnte ein mühsames Anbringen von Beschleunigungssensoren an den Hinterbeinen der Milchkühe überflüssig machen. Zur Datenaufnahme wurden Messungen mit Beschleunigungssensoren an den Hinterbeinen der Milchkühe und am Melkaggregat während der Morgen- und Abendmelkungen einmal je Tier durchgeführt. Zusätzlich wurde die Hinterbeinaktivität direkt beobachtet. Die Hinterbeinaktivität wurde in aktive und inaktive Phasen unterteilt. Die Daten der Morgenmelkungen wurden dazu benutzt, einen Algorithmus zu erstellen, welcher auf die Daten der Abendmelkungen angewendet wurden, um eine automatisierte Hinterbeinaktivitätserfassung zu testen. Die Korrelationsrechnungen der Aktivitätsmessung an den Hinterbeinen waren wie folgt: -Der Algorithmus definierte sich aus „Anzahl Beobachtungen aus Mittelwert + Standardabweichung + 0.2 m/s2“. -Die Korrelation von der „Anzahl gezählter aktiver Phasen aus der grafischen Auswertung“ und der „Anzahl aktive Phasen berechnet aus dem Algorithmus am Melkaggregat“ war 97 % (Morgenmelkung). -Die Korrelation von der „Anzahl aktiver Phasen des Melkaggregats errechnet aus dem Algorithmus“ und der „Anzahl aktive Phasen Direktbeobachtung“ war 74 %. -Die Korrelation „Anzahl aktiver Phasen der Hinterbeine errechnet aus dem Algorithmus“ und der „Anzahl aktive Phasen Direktbeobachtung“ war 91 %. Mit dem Erstellen eines Algorithmus gelang es eine Methode zur Hinterbeinaktivitätsmessung zu entwickeln, ohne Beschleunigungssensoren an den Hinterbeinen anbringen zu müssen. Mit diesem neuartigen Ansatz könnten Kosten und Arbeitseinsatz minimiert werden. Es ist vorstellbar, diese Methode als Diagnosewerkzeug in den Melkstand oder in das automatische Melksystem zu implementieren um wichtige Kuhsignale noch besser zu erkennen und an den Landwirt und Berater durch ein Managementprogramm zu senden.

This report presents the main results of the second meeting of the European Grassland Federation Working Group "Grazing" which was held in Lublin, Poland on 3 June 2012. The theme of the meeting was "Innovations in Grazing".

This report presents the main results of the second meeting of the European Grassland Federation Working Group "Grazing" which was held in Lublin, Poland on 3 June 2012. The theme of the meeting was "Innovations in Grazing".

The dairy cow's teat is the first line of defence against mastitis pathogens. The milking process may affect the teat's condition, increasing the risk of mastitis. It is well-proven that teat-ends with severe erosions or broken skin will have an increased risk of mastitis. However, more common changes in teat condition because of milking have not been related to udder health problems. The focus of this thesis was on the relationship between teat-end condition, machine milking and occurrence of mastitis. In this thesis, two types of changes of teat-end condition were distinguished: callosity rings around the orifice and machine-induced teat swelling. A classification system of the callosity rings around the orifice was defined: the teat-end callosity (TEC) classification system. In this system, a distinction is made between roughness of the callosity ring (TECR) and thickness of the callosity ring (TECT). The developed TEC classification system was used in a 1½ year longitudinal field study on 15 farms to examine the relationship between TEC and the incidence of clinical mastitis. Teats with a thin and smooth TEC ring showed the lowest incidence risk of clinical mastitis. To evaluate TEC in the field, a simplified 4-category scoring system is suggested and used in an observational study on 200 dairy farms. Variation in %ROUGH between farms is explained by cow factors such as teat-end shape and machine-on time and milking machine factors such as the liner and the vacuum. In order to measure machine-induced teat swelling, a methodology, using ultrasound, has been developed. Using this method, the changes of teat tissue in relation to machine milking and the recovery time of teat tissue after milking were evaluated. 8 h after milking, the teat-end width and the teat-canal length still differed from before milking. The teat-wall thickness and the teat-cistern width were recovered after 6 and 8 h. The overall conclusion of this thesis is that a healthy teat of a dairy cow has a good balance between the physiological reaction to machine milking and maintaining its first line of defence mechanism against invading mastitis pathogens. Increasing rates of IMI were related to one or more of the following: a high degree of machine-induced swelling, a high level of TECT, a high level of TECR and the absence of TEC. Pathways through which these machine-induced changes lowered the resistance of the teat to bacterial invasion are the openness of the teat canal, harbouring of pathogens in TEC, and significantly increased or decreased level of keratin regeneration rate. Part of the impaired reaction of the teat to machine milking may lay in the peak milkflow rate. Suggestions are made to adjust the characteristics of machine milking to the milk flow profile of an individual cow. This can minimise machine-induced teat condition problems. Teat condition changes can be used as an early warning signal for enhanced risk of clinical mastitis. Classification of teat condition is an essential tool in milking machine research and a useful monitoring tool of the quality of milking in the field. Protocols for systematic evaluation of teat condition are available De spenen van melkkoeien zijn de eerstelijns afweer tegen mastitis veroorzakende bacteriën. Het melken kan de speenconditie aantasten. Het is bewezen dat ernstige speenpunt-beschadigingen lijden tot meer mastitis. Maar de meer algemeen voorkomende veranderingen in de conditie van de speen door het melken zijn niet gerelateerd aan uiergezondheidsproblemen. Dit proefschrift was gericht op het verband tussen speenconditie, melken en het optreden van mastitis. In dit werk zijn 2 typen van speenconditie-veranderingen bekeken: speenpuntvereelting en zwelling van spenen door het melken

The dairy cow's teat is the first line of defence against mastitis pathogens. The milking process may affect the teat's condition, increasing the risk of mastitis. It is well-proven that teat-ends with severe erosions or broken skin will have an increased risk of mastitis. However, more common changes in teat condition because of milking have not been related to udder health problems. The focus of this thesis was on the relationship between teat-end condition, machine milking and occurrence of mastitis. In this thesis, two types of changes of teat-end condition were distinguished: callosity rings around the orifice and machine-induced teat swelling. A classification system of the callosity rings around the orifice was defined: the teat-end callosity (TEC) classification system. In this system, a distinction is made between roughness of the callosity ring (TECR) and thickness of the callosity ring (TECT). The developed TEC classification system was used in a 1½ year longitudinal field study on 15 farms to examine the relationship between TEC and the incidence of clinical mastitis. Teats with a thin and smooth TEC ring showed the lowest incidence risk of clinical mastitis. To evaluate TEC in the field, a simplified 4-category scoring system is suggested and used in an observational study on 200 dairy farms. Variation in %ROUGH between farms is explained by cow factors such as teat-end shape and machine-on time and milking machine factors such as the liner and the vacuum. In order to measure machine-induced teat swelling, a methodology, using ultrasound, has been developed. Using this method, the changes of teat tissue in relation to machine milking and the recovery time of teat tissue after milking were evaluated. 8 h after milking, the teat-end width and the teat-canal length still differed from before milking. The teat-wall thickness and the teat-cistern width were recovered after 6 and 8 h. The overall conclusion of this thesis is that a healthy teat of a dairy cow has a good balance between the physiological reaction to machine milking and maintaining its first line of defence mechanism against invading mastitis pathogens. Increasing rates of IMI were related to one or more of the following: a high degree of machine-induced swelling, a high level of TECT, a high level of TECR and the absence of TEC. Pathways through which these machine-induced changes lowered the resistance of the teat to bacterial invasion are the openness of the teat canal, harbouring of pathogens in TEC, and significantly increased or decreased level of keratin regeneration rate. Part of the impaired reaction of the teat to machine milking may lay in the peak milkflow rate. Suggestions are made to adjust the characteristics of machine milking to the milk flow profile of an individual cow. This can minimise machine-induced teat condition problems. Teat condition changes can be used as an early warning signal for enhanced risk of clinical mastitis. Classification of teat condition is an essential tool in milking machine research and a useful monitoring tool of the quality of milking in the field. Protocols for systematic evaluation of teat condition are available

The objectives of the workshop are to share our knowledge, to develop a research framework for future co-operative research between Israel and the Netherlands, and to explore the possibilities for funding of this research. As a starting point, a seminar is held in which available knowledge on robotic milking and on heat stress is shared and discussed. This proceedings gives a view of the knowledge presented during this one-day seminar.

Detection models for oestrus and mastitis in dairy cows were developed, based on sensors for milk yield, milk temperature, electrical conductivity of milk, cow's activity and concentrate intake, and on combined processing of the sensor data. The detection model generated alerts for cows, that need the farmer's attention, because of a possible case of oestrus or mastitis. A first detection model for cows, milked twice a day, was based on time series models for the sensor variables, where the parameters were fitted on-line for each cow after each milking by a Kalman filter. This model was tested during two years on two experimental farms, and under field conditions on four farms during several years. A second detection model, for cows milked in an automatic milking system (AMS), was based on a generalisation of the first model. Two data sets (one small, one large) were used for testing. The results of both models for oestrus detection were good, for mastitis varying. Fuzzy logic was used for the classification of mastitis and oestrus alerts with both detection models, to reduce the number of false positive alerts. Input for the fuzzy logic model were alerts from the detection models and additional information. The number of false positive alerts decreased considerably, while keeping the number of detected cases at the same level. The models make automated detection possible in practice.

Detection models for oestrus and mastitis in dairy cows were developed, based on sensors for milk yield, milk temperature, electrical conductivity of milk, cow's activity and concentrate intake, and on combined processing of the sensor data. The detection model generated alerts for cows, that need the farmer's attention, because of a possible case of oestrus or mastitis. A first detection model for cows, milked twice a day, was based on time series models for the sensor variables, where the parameters were fitted on-line for each cow after each milking by a Kalman filter. This model was tested during two years on two experimental farms, and under field conditions on four farms during several years. A second detection model, for cows milked in an automatic milking system (AMS), was based on a generalisation of the first model. Two data sets (one small, one large) were used for testing. The results of both models for oestrus detection were good, for mastitis varying. Fuzzy logic was used for the classification of mastitis and oestrus alerts with both detection models, to reduce the number of false positive alerts. Input for the fuzzy logic model were alerts from the detection models and additional information. The number of false positive alerts decreased considerably, while keeping the number of detected cases at the same level. The models make automated detection possible in practice.

This thesis describes new methods and knowledge-based systems for the analysis of technical and economic accounting data from the year-end records of individual dairy farms to support the management and, after adaptation, for policy analysis.A new method for farm comparison, the farm-adjusted standard, which makes it able to compare a dairy farm with similar farms, is described. Two methods for the acquisition, representation, and presentation of knowledge from experts are developed. These methods, IMAGINE and FUZZY-DETECTOR, can be used for different types of data and are especially characterised by fuzzy boundaries, compensatory mechanisms, and fast and easy knowledge acquisition.Two knowledge-based systems have been developed where these methods are used. GLOBAL-DETECTOR performs a global analysis of year-end results concerning aspects of gross margin. ENVIRONMENT-DETECTOR gives suggestions for a reduction of nitrogen surplus while maintaining the income. Proposed is a new method for sector responses on government options (APPROXI). A model is presented as an example of this method.The developed methods and knowledge-based systems for this thesis offer many opportunities for application in other domains.

Afhankelijk van de bedrijfsgrootte worden de baten berekend van het gebruik van het rundergroeihormoon BST per koe per jaar

In this paper, consumer behaviour in the Netherlands in respect of milk is investigated using a model based on the EKB model, a so-called integrated model of consumer behaviour. The objectives of the study are: to gain insight into the factors that influence buying and consumption behaviour with respect to milk and into the applicability of integrated models to consumer behaviour regarding generic products, such as liquid milk. It was established that liquid milk was perceived as a neutral drink: not ordinary or luxury, as a food not a drink for pleasure, and not as being refreshing. Liquid milk was perceived differently at varying times of consumption during the day. The most important socioeconomic variables explaining individual differences in consumer beliefs regarding milk are: age, level of education, and residential area. These and other conclusions from the analysis are useful for market segmentation. The results suggest that empirical models of the EKB type can contribute to the understanding of consumer behaviour in respect of generic food products. The specific extensions of the general EKB model made in this study may be relevant for the analysis of consumer behaviour in regard to other food products. Via het geintegreerde model voor consumentengedrag van Engel, Kollat en Blackwell (EKB-model) wordt het consumentengedrag in Nederland met betrekking tot melk benaderd