S. van Mourik

Automated body weight prediction of dairy cows using 3-dimensional vision

The objectives of this study were to quantify the error of body weight prediction using automatically measured morphological traits in a 3-dimensional (3-D) vision system and to assess the influence of various sources of uncertainty on body weight prediction. In this case study, an image acquisition setup was created in a cow selection box equipped with a top-view 3-D camera. Morphological traits of hip height, hip width, and rump length were automatically extracted from the raw 3-D images taken of the rump area of dairy cows (n = 30). These traits combined with days in milk, age, and parity were used in multiple linear regression models to predict body weight. To find the best prediction model, an exhaustive feature selection algorithm was used to build intermediate models (n = 63). Each model was validated by leave-one-out cross-validation, giving the root mean square error and mean absolute percentage error. The model consisting of hip width (measurement variability of 0.006 m), days in milk, and parity was the best model, with the lowest errors of 41.2 kg of root mean square error and 5.2% mean absolute percentage error. Our integrated system, including the image acquisition setup, image analysis, and the best prediction model, predicted the body weights with a performance similar to that achieved using semi-automated or manual methods. Moreover, the variability of our simplified morphological trait measurement showed a negligible contribution to the uncertainty of body weight prediction. We suggest that dairy cow body weight prediction can be improved by incorporating more predictive morphological traits and by improving the prediction model structure.

Predicting hairline fractures in eggs of mature hens

&NA; Eggshell damage poses a serious problem for the consumption egg industry. Increasing the maximum age of laying hens will increase eggshell damage due to loss of shell strength. This poses a serious problem for automatic collection, packing, and transport. We performed a model based study focused on hairline fractures in eggs of 88‐week‐old hens, and simulated side collisions on 1,235 eggs using a specially designed pendulum. The kinetic energy at the moment of impact was related to the accelerations measured by an electronic egg going through the transport chain. Further, several egg mechanical properties were measured. For collisions with a realistic impact, fracture occurrence correlated negatively with dynamic stiffness (14%), mass (15%), shape index (9%), and damping ratio (12%). We manipulated the data set to investigate the influence of improving egg properties. Removing the least favorable 50% of the eggs based on stiffness and mass resulted in a moderate reduction of fracture occurrence, from 7.7% down to 4.4%. The peak acceleration of an egg running through the transport chain lies typically in the range of 15 to 45 g. Our model predicts that a moderate decrease from 30 g down to 20 g will result in a drastic reduction of fracture occurrence from 7.7% down to 0.3 to 1% (95% confidence region), whereas an increase to 40 g will increase fracture occurrence to 42 to 55%. The model predicts that severe collisions pose a relatively high risk for eggshell damage, which suggests that a reduction of collision severity is of first priority when increasing the age of laying hens.

Indicators of resilience during the transition period in dairy cows: A case study

The transition period is a demanding phase in the life of dairy cows. Metabolic and infectious disorders frequently occur in the first weeks after calving. To identify cows that are less able to cope with the transition period, physiologic or behavioral signals acquired with sensors might be useful. However, it is not yet clear which signals or combination of signals and which signal properties are most informative with respect to disease severity after calving. Sensor data on activity and behavior measurements as well as rumen and ear temperature data from 22 dairy cows were collected during a period starting 2 wk before expected parturition until 6 wk after parturition. During this period, the health status of each cow was clinically scored daily. A total deficit score (TDS) was calculated based on the clinical assessment, summarizing disease length and intensity for each cow. Different sensor data properties recorded during the period before calving as well as the period after calving were tested as a predictor for TDS using univariate analysis of covariance. To select the model with the best combination of signals and signal properties, we quantified the prediction accuracy for TDS in a multivariate model. Prediction accuracy for TDS increased when sensors were combined, using static and dynamic signal properties. Statistically, the most optimal linear combination of predictors consisted of average eating time, variance of daily ear temperature, and regularity of daily behavior patterns in the dry period. Our research indicates that a combination of static and dynamic sensor data properties could be used as indicators of cow resilience.

Integrating two patterning processes in the flower

Spatial organ arrangement plays an important role in flower development. The position and the identity of floral organs is influenced by various processes, in particular the expression of MADS-box transcription factors for identity and dynamics of the plant hormone auxin for positioning. We are currently integrating patterning processes of MADS and auxin into our computational models, based on interactions that are known from experiments, in order to get insight in how these define the floral body plan. The resulting computational model will help to explore hypothetical interactions between the MADS and auxin regulation networks in floral organ patterning.

Modelling and controller design for distributed parameter systems via residence time distribution

For chemical reactors with non-linear fluid dynamics, a linear model realisation is proposed. The inputs are the ingoing concentration of a certain component in the fluid, and the reaction rate. The output is the outgoing concentration. The realisation makes use of a first-order reaction equation, and the residence time distribution of the fluid particles inside the reactor. Also dead time is incorporated in the modelling. The method is tested on two non-linear models for which the residence time distributions are known analytically. The first model is a series of mixed tanks, and it is shown by simulation that the method gives an accurate approximation of the original model. The second model is a UV disinfection reactor, which has a dead time. For this model, the residence time distribution is first fitted by a form that is suitable for our realisation method. Simulations show that for realistic disturbances a high-performance linear controller can be designed. After that, the residence time distribution of a real life UV reactor (for which we have no model) is fitted by a suitable form. The fit is of the same quality as for the UV reactor model. This indicates that also for the real life UV reactor a high-performance controller can be designed.