Search

A spectacular measurement campaign was carried out on a real-world motorway stretch of Hungary with the participation of international industrial and academic partners. The measurement resulted in vehicle based and infrastructure based sensor data that will be extremely useful for future automotive R&D activities due to the available ground truth for static and dynamic content. The aim of the measurement campaign was twofold. On the one hand, road geometry was mapped with high precision in order to build Ultra High Definition (UHD) map of the test road. On the other hand, the vehicles—equipped with differential Global Navigation Satellite Systems (GNSS) for ground truth localization—carried out special test scenarios while collecting detailed data using different sensors. All of the test runs were recorded by both vehicles and infrastructure. The paper also showcases application examples to demonstrate the viability of the collected data having access to the ground truth labeling. This data set may support a large variety of solutions, for the test and validation of different kinds of approaches and techniques. As a complementary task, the available 5G network was monitored and tested under different radio conditions to investigate the latency results for different measurement scenarios. A part of the measured data has been shared openly, such that interested automotive and academic parties may use it for their own purposes.

The first chapter of this book explained what machine learning is and why it is needed. This chapter now gives an in-depth overview of the subject. The most important machine learning algorithms (models) are explained in detail and several important questions are answered: Which algorithm should we select for the task? What are the advantages and disadvantages of the model? This chapter focuses on the practical uses of machine learning; the mathematical background is only explained when it is really necessary—typically in separate ‘expert sections’ to aid comprehension and to allow interested readers to dive deeper into the subject. Several examples are provided to help explain the different applications of machine learning.

In the earlier chapters of this book we have seen how machine learning works and what the different machine learning techniques are. This chapter will explain how to apply these machine learning techniques to real-world problems: automatic classification (clustering) of an unknown dataset, dimensionality reduction of large datasets, recommendation models, anomaly detection, root cause analysis, engineering applications of supervised learning regression, predictive maintenance, image recognition, detection of different diseases, business process improvement, etc. Not only the specific applications will be explained but also tips and tricks and important sector-dependent machine learning recommendations. After reading this chapter you will be able to apply the machine learning techniques explained in the previous chapters to your own machine learning problems, even if they are different from the examples presented here.

Machine learning models (algorithms) learn from data. This chapter explains how to collect, store and preprocess the data for machine learning. We will also look at the data strategy and the connection to machine learning data strategy. All data related subjects are very important because the quality of the machine learning model depends greatly on the data quality! We will look at data pre-processing (cleaning, transformation, sampling, resampling, feature selection, normalization, etc.) in more detail because of its importance.

It is not always clear to people, especially if they are new to the subject, what we mean by machine learning and when and why we need it. A lot of people are aware of artificial intelligence (AI) from science fiction but they may not really understand the reality and the connection to machine learning. This chapter will explain in clear lay terms what machine learning and AI are, and it will also introduce the three major forms of machine learning: supervised, unsupervised and reinforcement learning. The aim is that after reading this chapter you will understand what, exactly, machine learning is and why we need it.

After learning about the most important machine learning models in the previous chapter, this chapter explains how to evaluate the performance of machine learning models. Evaluating the performance means that we determine how well the model learns the given subject (data) and how well it applies this learning in practice. Each major form of machine learning has its own performance measures and they are often application dependent. This chapter explains the different types of performance evaluations and how and where to apply them. After reading this chapter you will be able to understand the performance measures reported by machine learning algorithms.

In this chapter you will learn how the AI-TOOLKIT works and how to use it. The full version of the AI-TOOLKIT is free, for non-commercial purposes, and therefore you can experiment with all of the examples in this book and apply the knowledge you have learned about machine learning in the previous chapters without any programming! The AI-TOOLKIT supports all three major forms of machine learning: supervised, unsupervised and reinforcement learning! No programming skills are needed at all to build and use state-of-the-art machine learning models!

Recent concepts of environmental sustainability have focused on narrative economic and societal aspects rather than quantitative ones. Many key sustainability indicators also lack a consistent definition of sustainability, have perspectives that are too short-term, and are unable to model the dynamics of complex environmental utilization which can then result in inappropriate projection of long-term sustainability and/or sustainability indication. Here I propose a generalized quantitative framework of environmental sustainability requiring that (1) environmental capacities and utilization rates are identified, (2) their complex temporal dynamics are quantitatively modeled or estimated (3) while also adjusting for uncertainties, and finally, (4) using one of three options, determining which cumulative utilization pathways can be sustained for a (usually well-defined) period of time. Using the example of wood volume and its growth as capacities and harvest as utilization, and the example of global greenhouse gas emissions as the utilization component and the capacity of the air to absorb these emissions, I demonstrate how the proposed framework can be applied in practice, how sustainability indicators could be developed, and also how they can inform policies and measures to ensure sustainability.

This article reports experiences gathered in 12 developing countries in Latin America and Africa with capacity development for national GHG inventory systems. The national systems and their ability to deliver on measurement, reporting and verification (MRV) requirements is assessed using a scorecard that covers the transparency, accuracy, completeness, consistency, and comparability of the GHG inventory as well as its institutional set-up and management aspects. The analysis shows that a quantitative assessment of GHG inventory systems with a scorecard is feasible and useful, and could commonly be used for tracking progress in MRV capacity development.Policy relevanceA large number of capacity development initiatives are underway to advance national GHG inventory systems in developing countries, particularly in the land-use sector. These aim to promote the reliable MRV of GHG emissions and removals, which is expected to underpin developing countries’ contributions to global mitigation efforts, including t...

The "Allometric, Biomass and Carbon factors" database (ABC factors database) was designed to facilitate the estimation of the biomass carbon stocks of forests in order to support the development and the verification of greenhouse gas inventories in the LULUCF sector (Land Use, Land Use Change and Forestry). The database contains several types of expansion, conversion and combined factors, by various tree species or species groups that can be used to calculate biomass or carbon of forests of Eurasian region from proxy variables (e.g., tree volume) that may come from forest inventories. In addition to the factors, and depending on the information that was available in the cited source, the database indicates: (1) the biomass compartments involved when the factor was developed; and (2) the possible applicability of the factor, e.g. by country or by ecological regions. The applicability of the factors is either suggested by the source itself, or the type of source (e.g. National Greenhouse Gas Inventory Report), or was based on the expert judgement by the compilers of the database. Finally, in order to facilitate the selection of the most appropriate of the data, the web-based interface provides the possibility to compare several factors that may come from different sources. The ABC factors database is freely available at the URL: http://afoludata.jrc.it/data_fs.cfm, in the web site AFOLU-DATA, funded and hosted by the Joint Research Centre (European Commission, DG-JRC)., JRC.H.2-Climate change

This paper presents a framework that encompasses a full range of options for including land use, land-use change, and forestry (LULUCF) within future agreements under the United Nations Convention on Climate Change (UNFCCC). The intent is to provide options that can address the broad range of greenhouse gas (GHG) emissions and removals as well as to bring the broadest possible range of nations into undertaking mitigation efforts. We suggest that the approach taken for the Kyoto Protocol's first commitment period is only one within a much larger universe of possible approaches. This larger universe includes partially or completely “de-linking” LULUCF commitments from those in other sectors, and allowing commitments specified in terms other than tonnes of greenhouse gases. Such approaches may provide clarity and transparency concerning the role of the various sectors in the agreements and encourage participation in agreements by a more inclusive, diverse set of countries, resulting in a more effective use of LULUCF in addressing climate change.

Forest biomass and its change over time have been measured at both local and large scales, an example for the latter being forest greenhouse gas inventories. Currently used methodologies to obtain stock change estimates for large forest areas are mostly based on forest inventory information as well as various factors, referred to as biomass factors, or biomass equations, which transform diameter, height or volume data into biomass estimates. However, while forest inventories usually apply statistically sound sampling and can provide representative estimates for large forest areas, the biomass factors or equations used are, in most cases, not representative, because they are based on local studies. Moreover, their application is controversial due to the inconsistent or inappropriate use of definitions involved. There is no standardized terminology of the various factors, and the use of terms and definitions is often confusing. The present contribution aims at systematically summarizing the main types of biomass factors (BF) and biomass equations (BE) and providing guidance on how to proceed when selecting, developing and applying proper factors or equations to be used in forest biomass estimation. The contribution builds on the guidance given by the IPCC (Good practice guidance for land use, land-use change and forestry, 2003) and suggests that proper application and reporting of biomass factors and equations and transparent and consistent reporting of forest carbon inventories are needed in both scientific literature and the greenhouse gas inventory reports of countries.

Background Pasteurella multocida causes numerous economically relevant diseases in livestock including rabbits. Immunisation is only variably effective. Prophylactic antibiotics are used in some species but are contra-indicated in rabbits, due to their adverse effects on the rabbit microbiota. There is therefore a substantial need for alternative forms of infection control in rabbits; we investigated the effect of oral β-glucan on P. multocida infection in this species. Results Thirthy-five New Zealand White rabbits were randomly divided into five groups of seven animals. Three groups were inoculated with Pasteurella multocida intranasally (in.), a physiologically appropriate challenge which reproduces naturally acquired infection, and received either (1–3), (1–6) β-glucans or placebo. Four other groups were inoculated both in. and intramuscularly (im.), representing a supra-physiological challenge, and received either (1–3), (1–6) β-glucans, antibiotic or placebo. β-glucans given prophylactically were highly effective in protecting against physiological (in.) bacterial challenge. They were less effective in protecting against supra-physiological bacterial challenge (in. and im.), although they extended survival times. This latter finding has practical relevance to breeders as it extends the window in which heavily infected and symptomatic animals can be salvaged with antibiotics. Conclusions In our study, (1–3), (1–6) β-glucans were highly effective in protecting against a model of naturally acquired P. multocida infection and extended survival times in the supra-physiological model. Enrofloxacin was effective in protecting against supra-physiological infection. We are currently reviewing the use of combined prophylaxis. Electronic supplementary material The online version of this article (doi:10.1186/s12917-014-0276-6) contains supplementary material, which is available to authorized users.

International audience; This paper is the outcome of a group discussion held at the Savonlinna meeting Management of forest ecosystems and its impact on the GHG budget'. The aim of this break-out group discussion was to Characterize forest management impacts on the GHG budget of forest ecosystems in different European regions'. In this paper we briefly characterize different options that a forest owner has in order to maintain or maximize forest carbon pools and carbon sequestration. These hectare scale descriptions of measures are then regarded in connection to the current state of European forests and how they can be combined with ongoing management trends and local issues. We point out the various possibilities that exist in European forests, where they are located, and where they could possibly be combined with adaptation. We identify these hotspots for largest growing stocks, largest peat areas, and, e.g. largest risks for loss of carbon due to fire or urban sprawl. We conclude that one common strategy cannot be designed. Within each region, local solutions have to be found that optimize goals and aim at integrated and sustainable land use.

The complexities inherent in land use, land-use change and forestry (LULUCF) activities have led to contentious and prolonged debates about the merits of their inclusion in the 2008–2012 first commitment period of the Kyoto Protocol. Yet the inclusion of these activities played a key role in agreement on the general framework of the Kyoto Protocol, and LULUCF will likely continue to play a substantial part in negotiations on national commitments post-2012. The Marrakech Accords dictate which LULUCF activities are to be included under the Kyoto Protocol and provide rules on how they are to be accounted in the first commitment period. However, these rules have limitations and drawbacks that may be avoided in the structure of future commitments beyond 2012. Through adherence to the objectives of the United Nations Framework Convention on Climate Change (UNFCCC), and the incorporation of several critical features, a future framework can more effectively address the mitigation challenges and opportunities of this sector.