No abstract

Summary Precise localization and reliable navigation are of crucial importance in order to fully automatize tasks in agriculture using field robots in an outdoor environment. However, under certain conditions, absolute localization accuracy of outdoor positioning systems, mainly global navigation satellite systems (GNSS), might drop under a critical threshold. Thus, field robots must rely on indoor positioning methods such as laser scanners and adaptive Monte Carlo localization (AMCL) in order to maintain necessary localization accuracy. Here, the localization accuracy of the field robot “Mathilda” using such an indoor positioning method is evaluated in a realistic scenario, in this case, an outdoor laboratory with plant pots aligned in rows. For this, the positioning error of the robot was determined using a motion capture system. The results showed a mean absolute distance error over all positions of 198.9 mm and a mean angular error over all positions of 4.9°. Most likely, limitations by the differential drive system, the large pneumatic tires, and unsatisfactory path planning are responsible for this large deviation.

Grassland vegetation typically comprises the species groups grasses, herbs, and legumes. These species groups provide different functional traits and feed values. Therefore, knowledge of the botanical composition of grasslands can enable improved site-specific management and livestock feeding. A systematic approach was developed to analyze vegetation of managed permanent grassland using hyperspectral imaging in a laboratory setting. In the first step, hyperspectral images of typical grassland plants were recorded, annotated, and classified according to species group and plant parts, that is, flowers, leaves, and stems. In the second step, three different machine learning model types—multilayer perceptron (MLP), random forest (RF), and partial least squares discriminant analysis (PLS-DA)—were trained with pixel-wise spectral information to discriminate different species groups and plant parts in individual models. The influence of radiometric data calibration and specific data preprocessing steps on the overall model performance was also investigated. While the influence of proper radiometric calibration was negligible in our setting, specific preprocessing variants, including smoothening and derivation of the spectrum, were found to be beneficial for classification accuracy. Compared to extensively preprocessed data, raw spectral data yielded no statistically decreased performance in most cases. Overall, the MLP models outperformed the PLS-DA and RF models and reached cross-validation accuracies of 96.8% for species group and 88.6% for plant part classification. The obtained insights provide an essential basis for future data acquisition and data analysis of grassland vegetation.

Summary Although crop residues contribute to erosion control, the influence of the tillage depth (TD) on their incorporation has not been studied extensively. The main objective of this study was to determine the differences in the amount and distribution of incorporated crop residues and surface residue coverage if the TD of a cultivator is varied (0.10, 0.20, or 0.30 m). The experiment was carried out on a chernozem soil with winter barley residues in 2016 in Groß-Enzersdorf (Lower Austria). Individual soil cores, each 0.05 m long, were removed using a special device. No significant differences were observed for the incorporated crop residues up to a depth of 0.35 m between the three TDs. The mean values of the incorporated crop residues at a TD of 0.10, 0.20, or 0.30 m were 11.64, 13.30, and 10.82 t/ha, respectively. The distribution of crop residues in the individual depth segments (DSs) showed a main concentration of more than 90% at a depth of 0.10 m and a significant decrease in deeper layers. This stratification was independent of the TD. Therefore, a shallower TD is sufficient for straw management on a chernozem soil in the production area of Marchfeld, which also enables a reduction in draft and, consequently, fuel consumption and processing costs.

No abstract