In a worldwide collaborative effort, 19,630 Y-chromosomes were sampled from 129 different populations in 51 countries. These chromosomes were typed for 23 short-tandem repeat (STR) loci (DYS19, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, DYS385ab, DYS437, DYS438, DYS439, DYS448, DYS456, DYS458, DYS635, GATAH4, DYS481, DYS533, DYS549, DYS570, DYS576, and DYS643) and using the PowerPlex Y23 System (PPY23, Promega Corporation, Madison, WI). Locus-specific allelic spectra of these markers were determined and a consistently high level of allelic diversity was observed. A considerable number of null, duplicate and off-ladder alleles were revealed. Standard single-locus and haplotype-based parameters were calculated and compared between subsets of Y-STR markers established for forensic casework. The PPY23 marker set provides substantially stronger discriminatory power than other available kits but at the same time reveals the same general patterns of population structure as other marker sets. A strong correlation was observed between the number of Y-STRs included in a marker set and some of the forensic parameters under study. Interestingly a weak but consistent trend toward smaller genetic distances resulting from larger numbers of markers became apparent.

Hydrothermal carbonization (HTC) of stabilized and non-stabilized sewage sludge was investigated in a 25 L vessel as a pretreatment for sewage sludge before incineration, and the composition and properties of the obtained HTC coal and process water were studied. The observed values for H/C and O/C in HTC coal from stabilized and non-stabilized sewage sludge were shown to be higher than in natural coal and rather close to typical values for cellulose. The upper heating value of the stabilized sewage sludge was increased from 3.4 to 6.5%, and the upper heating value of the non-stabilized sludge was increased from 5.8 to 11.0%, after carbonization. The carbon efficiency ranged from 62 to 71% for stabilized sewage sludge and from 60 to 66% for non-stabilized sewage sludge, and the dry matter (DM) loss after carbonization was 31 and 42% for stabilized sludge and 34 and 44% for non-stabilized sludge. After carbonization, the mechanical dewaterability was increased from 30 to 70% DM content for non-stabilized sludge and from 37 to 52% for stabilized sludge. The drying process of sewage sludge including HTC needs a clearly lower energy input than established drying techniques to produce a fuel. For the drying process of 1 ton of nonstabilized sewage sludge with 9% DM, the calculated energy consumption was lowered by 99.6 kWh of thermal energy and 8.5 kWh of electric energy by introducing HTC. The results of these experiments show the feasibility of the HTC process as part of the drying process of sewage sludge and the fate of key elements in the process on a laboratory scale. However, the process has to be further optimized and developed on an industrial scale. Further important development steps include recovery steps for the carbon in the process water and adapted process water treatments.

A significant leap forward in the performance of remote sensing models can be attributed to recent advances in machine and deep learning. Large data sets particularly benefit from deep learning models, which often comprise millions of parameters. On which part of the data a machine learner focuses on during learning, however, remains an open research question. With the aid of a notion of label uncertainty, we try to address this question in local climate zone (LCZ) classification. Using a deep network as a feature extractor, we identify data samples that are seemingly easy or hard to classify for the model and base our experiments on the relatively more uncertain samples. For training of the network, we make use of distributional (probabilistic) labels to incorporate the voter confusion directly into the training process. The effectiveness of the proposed uncertainty-guided representation learning is shown in context of active learning framework where we show that adding more certain data to the training pool increases model performance even with the limited data.

This paper discusses the framework for the development of an Energy Toolbox (ETB). The aim of the ETB is to support the design of domestic Zero Emission Buildings (ZEBs), according to the concept of net zero-energy buildings during the early architectural design and planning phases. The ETB concept is based on the calculation of the energy demand for heating, cooling, lighting, and appliances. Based on a building's energy demand, technologies for the onsite conversion and production of the specific forms and quantities of final and primary energy by means of renewable energy carriers can be identified. The calculations of the ETB are based on the building envelope properties of a primary building design, as well as the physical and climate parameters required for the calculation of heat transfer coefficients, heat gains, and heat losses. The ETB enables the selection and rough dimensioning of technologies and systems to meet, and, wherever possible, reduce the thermal and electric energy demand of a building. The technologies included comprise green facades, adaptable dynamic lighting, shading devices, heat pumps, photovoltaic generators, solar thermal collectors, adiabatic cooling, and thermal storage. The ETB facilitates the balancing of the energy consumption and the production of renewable energies of a primary building design.

In the scope of this study, a pilot facility for the recycling of laundry effluent was developed and tested. With the aim to enable nearly complete energy and water self-sufficiency, the system is powered by a photovoltaic plant with second-life batteries, treats the wastewater within the unit and constantly reuses the treated wastewater for washing in a closed cycle. The technology for wastewater treatment is based on a low-tech approach consisting of a physical/mechanical pre-treatment and biological treatment in trickling filter columns. The treatment process is operated in batch mode for a capacity of five washing cycles per day. During five weeks of operation water quality, energy consumption and production, water losses and washing performance were monitored. The system recovered 69% of the used water for the washing machine while treating the wastewater to the necessary water quality levels. The average COD removal rate per cycle was 92%. Energy analysis was based on modelled data of the monitored energy consumption. With the current set-up, an internal consumption rate of 80% and self-sufficiency of 30% were modelled. Future developments aim at increasing water and energy self-sufficiency and optimizing the water treatment efficiency.

Technological and computational advances continuously drive forward the broad field of deep learning. In recent years, the derivation of quantities describing the uncertainty in the prediction -which naturally accompanies the modeling processhas sparked general interest in the deep learning community. Often neglected in the machine learning setting is the human uncertainty that influences numerous labeling processes. As the core of this work, label uncertainty is explicitly embedded into the training process via distributional labels. We demonstrate the effectiveness of our approach on image classification with a remote sensing data set that contains multiple label votes by domain experts for each image: The incorporation of label uncertainty helps the model to generalize better to unseen data and increases model performance. Similar to existing calibration methods, the distributional labels lead to better-calibrated probabilities, which in turn yield more certain and trustworthy predictions.Preprint. Under review.

No abstract

No abstract