The XPM project aims to integrate explanations into Artificial Intelligence (AI) solutions within the area of Predictive Maintenance (PM). Real-world applications of PM are increasingly complex, with intricate interactions of many components. AI solutions are a very popular technique in this domain, and especially the black-box models based on deep learning approaches are showing very promising results in terms of predictive accuracy and capability of modelling complex systems. However, the decisions made by these black-box models are often difficult for human experts to understand – and therefore to act upon. The complete repair plan and maintenance actions that must be performed based on the detected symptoms of damage and wear often require complex reasoning and planning processes, involving many actors and balancing different priorities. It is not realistic to expect this complete solution to be created automatically – there is too much context that needs to be taken into account. Therefore, operators, technicians and managers require insights to understand what is happening, why it is happening, and how to react. Today’s mostly black-box AI does not provide these insights, nor does it support experts in making maintenance decisions based on the deviations it detects. The effectiveness of the PM system depends much less on the accuracy of the alarms the AI raises than on the relevancy of the actions operators perform based on these alarms. In the XPM project, we will develop several different types of explanations (anything from visual analytics through prototypical examples to deductive argumentative systems) and demonstrate their usefulness in four selected case studies: electric vehicles, metro trains, steel plant and wind farms. In each of them, we will demonstrate how the right explanations of decisions made by AI systems lead to better results across several dimensions, including identifying the component or part of the process where the problem has occurred; understanding the severity and future consequences of detected deviations; choosing the optimal repair and maintenance plan from several alternatives created based on different priorities, and understanding the reasons why the problem has occurred in the first place as a way to improve system design for the future.

Access to a diverse spectrum of food resources ensures appropriate nutrition and is thus crucial for animal health and fitness. Bees obtain nearly all nutrients from flowers. Their population dynamics are therefore largely determined by the availability, composition and diversity of flowering plants. Alarmingly, many bee populations are in decline in contemporary landscapes, likely due to the loss of floral resource diversity and abundance and to a decrease in the nutritional quality of floral resources. However, the actual link between floral diversity and composition, the nutritional composition of floral resources and bee health is still unclear, particularly in wild bees, which are considered even less resilient to environmental changes than honeybees. In NutriB2, eleven scientists from seven different countries will combine their expertise in taxonomy, nutritional & chemical ecology, physiology, behavior, epidemiology, biostatistics and modeling to, in a synergistic effort, clarify the link between floral biodiversity, nutrition and bee health. We will further reveal critical nutrients and/or ratios and thus key plant species and compositions of plant species that cover the nutritional needs and support health of a large fraction of bee species. This knowledge is crucial for our understanding of how floral composition and diversity structure bee communities through nutritionally mediated health effects. It is also essential for designing and/or identifying and restoring habitats that support wild bee populations. Our results will be shared and processed with different stakeholders (i.e. seed companies, beekeeping and farmers’ organizations, regional to international conservation groups, schools with programs to actively promote bee diversity, other business representatives and policymakers) based on already established contacts and networks. Our aim is to determine feasible ways of a) restoring and/or maintaining semi-natural habitats with nutritionally highly valuable plant species and b) designing and implementing nutritionally balanced floral seed mixes. Nutritionally appropriate and diverse floral communities will not only benefit diverse bee species, but also others animals depending on plants as well as higher trophic levels. NutriB2 will therefore not only shed light on the mechanisms underlying the known positive correlation between floral biodiversity and bee health, but also enable us to design better strategies for conserving or restoring floral diversity for bees and thus mitigate the ongoing wild bee and biodiversity decline.

Transitioning towards FAIR data has been notoriously challenging in the long tail of science. This is mainly because the long tail consists of many independently assembled datasets different in structure and management, yet collected with a similar purpose. This proposal builds on the recently established SPI-Birds Network and Database on long-term population studies of birds, hosting and standardising individual-level data. Our aim is to develop this network as a model for other (ecological) research domains, with particular emphasis on the currently under-addressed stages of data processing, data analysis, and data preservation. First, we will develop a repository with citable datasets that can be linked to publications, analytical codes, and other output, thereby closing the research lifecycle in the long tail of science. Second, we will expand the user base with respect to both data producers (expanding geographic and ecological coverage of the datasets), researchers within and outside the SPI-Birds community, and users outside research such as educators, journalists, environmental professionals or artists and designers. Central to these first two objectives is the continued development and refinement of an existing standard data format with rich metadata (and accessible metadata guides) according to common standards such as the Darwin Core. Third, we will establish a peer-reviewed archive (including tests by selected users) for data processing and analytical codes (that can be applied to the datasets hosted in the database, but also beyond), promoting efficiency and replicability of large-scale data analyses. Our final goal is a fully transparent FAIR research landscape of datasets, analytical tools and publications that are connected by digital identifiers and rich metadata.