No abstract

Long-term situation prediction plays a crucial role in the development of intelligent vehicles. A major challenge still to overcome is the prediction of complex downtown scenarios with multiple road users, e.g., pedestrians, bikes, and motor vehicles, interacting with each other. This contribution tackles this challenge by combining a Bayesian filtering technique for environment representation, and machine learning as long-term predictor. More specifically, a dynamic occupancy grid map is utilized as input to a deep convolutional neural network. This yields the advantage of using spatially distributed velocity estimates from a single time step for prediction, rather than a raw data sequence, alleviating common problems dealing with input time series of multiple sensors. Furthermore, convolutional neural networks have the inherent characteristic of using context information, enabling the implicit modeling of road user interaction. Pixel-wise balancing is applied in the loss function counteracting the extreme imbalance between static and dynamic cells. One of the major advantages is the unsupervised learning character due to fully automatic label generation. The presented algorithm is trained and evaluated on multiple hours of recorded sensor data and compared to Monte-Carlo simulation.

Ultrasonic guided waves have been used successfully in structural health monitoring systems to detect damage in isotropic and composite materials with simple and complex geometry. A limitation of current research is given by a lack of freely available benchmark measurements to comparatively evaluate existing methods. This article introduces the extendable online platform Open Guided Waves ( http://www.open-guided-waves.de ) where high-quality and well-documented datasets for guided wave-based inspections are provided. In this article, we describe quasi-isotropic carbon-fiber-reinforced polymer plates with embedded piezoelectric transducers as a first benchmark structure. Intentionally, this is a structure of medium complexity to enable many researchers to apply their methods. In a first step, ultrasound and X-ray measurements were acquired to verify pristine conditions. Next, mechanical testing was done to determine the stiffness tensor and sample density based on standard test procedures. Guided wave measurements were divided into two parts: first, acoustic wave fields were acquired for a broad range of frequencies by three-dimensional scanning laser Doppler vibrometry. Second, structural health monitoring measurements in the carbon-fiber-reinforced polymer plate were collected at constant temperature using a distributed transducer network and a surface-mounted reversible defect model. Initial results serving as validation are presented and discussed.

Superabsorbent polymers (SAP) are used, inter alia, as soil amendment to increase the water holding capacity of soils. Biodegradability of soil conditioners has become a desired key characteristic to protect soil and groundwater resources. The present study characterized the biodegradability of one acrylate based SAP in four agricultural soils and at three temperatures. Mineralisation was measured as the (13)CO₂ efflux from (13)C-labelled SAP in soil incubations. The SAP was either single-labelled in the carboxyl C-atom or triple-labelled including additionally the two C-atoms interlinked in the SAP backbone. The dual labelling allowed estimating the degradation of the polyacrylate main chain. The (13)CO₂ efflux from samples was measured using an automated system including wavelength-scanned cavity ring-down spectroscopy. Based on single-labelled SAP, the mean degradation after 24 weeks varied between 0.45% in loamy sand and 0.82% in loam. However, the differences between degradation rates in different soils were not significant due to a large intra-replicate variability. Similarly, mean degradation did not differ significantly between effective temperature regimes of 20° and 30 °C after 12 weeks. Results from the triple-labelled SAP were lower as compared to their single-labelled variant. Detailed results suggest that the polyacrylate main chain degraded in the soils, if at all, at rates of 0.12-0.24 % per 6 months.

For many automated driving functions, a highly accurate perception of the vehicle environment is a crucial prerequisite. Modern high-resolution radar sensors generate multiple radar targets per object, which makes these sensors particularly suitable for the 2D object detection task. This work presents an approach to detect 2D objects solely depending on sparse radar data using PointNets. In literature, only methods are presented so far which perform either object classification or bounding box estimation for objects. In contrast, this method facilitates a classification together with a bounding box estimation of objects using a single radar sensor. To this end, PointNets are adjusted for radar data performing 2D object classification with segmentation, and 2D bounding box regression in order to estimate an amodal 2D bounding box. The algorithm is evaluated using an automatically created dataset which consist of various realistic driving maneuvers. The results show the great potential of object detection in high-resolution radar data using PointNets.

No abstract

No abstract

No abstract