Advanced embedded systems and sensor networks for animal environment monitoring

Advancements in sensing and monitoring of air quality parameters within confined animal feeding operations have been realized through the application of embedded systems and advanced networking. The development of an embedded vibration sensor to detect the presence of ventilation fan activity provided researchers with an improved method to monitor ventilation from high capacity CAFO facilities. Experiments revealed over estimation errors common to the majority of passive ventilation sensors. Analysis of ventilation sensor systems resulted in proposed limits to overall measurement error by minimizing the modulus of fan on-time and sampling time.; Controller Area Networks were found to be a visible means to link multiple analog and digital sensors through a multi-master based embedded network. It was found that signal attenuation was significantly as bus lengths increased to a maximum of 600 meters. This attenuation was counteracted by reducing the baud rate of the communication and allowing for longer bit times. Signal reflection of the individual bits was another major factor of transmission error caused by the mismatch of impedance between the signal wire and the termination resistor.; Wireless sensor networks were also evaluated for their potential to act as the data communication network within a multi-point sampling system inside a CAPO. Results from experimental path loss studies found many factors including antenna orientation, enclosure thickness, free space, antenna height, animal cages, and concrete floor separations to all be statistically relevant factors in determining the overall system path loss. It was found that linear separation within an aisle and number of cage separations provided the highest levels of signal attenuation.; A model was developed to predict the path loss at any point within a poultry layer facility based on the aisle and cage separation terms. The model was able to predict 86% of the system variability and was able to produce an average error of -0.7 dB for all combined points. The verification of all theoretical path loss models indicates that when applied to new systems not representative of poultry layer facilities, fundamental laws can be used to create initial predictions for path loss.