Abstract
This article discusses how a business model based on traditional maintenance can evolve to generate servitization strategies, with the help of remote maintenance support. The application of cyber-physical systems and cloud technologies play a key role for such maintenance purposes. In fact, the utilization of large quantities of data collected on machines and their processing by means of advanced techniques such as machine learning enable novel techniques for condition-based maintenance. New sensor solutions that could be used in maintenance and interaction with cyber-physical systems are also presented. Here, data models are an important part of these techniques because of the huge amounts of data that are produced and should be processed. These data models have been used in a real case, supported by the Machinery Information Management Open System Alliance Open System Architecture for Condition-Based Maintenance standard architecture, for streamlining the modeling of collected data. In this context, an industrial use case is described, to enlighten the application of the presented concepts in a working pilot. Finally, current and future directions for application of cyber-physical systems and cloud technologies to maintenance are discussed.

Abstract
Points of Interest (POI) are widely used in many applications nowadays mainly due to the increasing amount of related data available online, notably from volunteered geographic information (VGI) sources. Being able to connect these data from different sources is useful for many things like validating, correcting and also removing duplicated data in a database. However, there is no standard way to identify the same POIs across different sources and doing it manually could be very expensive. Therefore, automatic POI matching has been an attractive research topic. In our work, we propose a novel data-driven machine learning approach based on an outlier detection algorithm to match POIs automatically. Surprisingly, works that have been presented so far do not use data-driven machine learning approaches. The reason for this might be that such approaches need a training dataset to be constructed by manually matching some POIs. To mitigate this, we have taken advantage of the Crosswalk API, available at the time we started our project, which allowed us to retrieve already matched POI data from different sources in US territory. We trained and tested our model with a dataset containing Factual, Facebook and Foursquare POIs from New York City and were able to successfully apply it to another dataset of Facebook and Foursquare POIs from Porto, Portugal, finding matches with an accuracy around 95%. These are encouraging results that confirm our approach as an effective way to address the problem of automatically matching POIs. They also show that such a model can be trained with data available from multiple sources and be applied to other datasets with different locations from those used in training. Furthermore, as a data-driven machine learning approach, the model can be continuously improved by adding new validated data to its training dataset. © Springer Nature Switzerland AG 2018.

Abstract

Abstract
The agribusiness volatility is related to the uncertainty of the environment, rising demand, falling prices and new technologies. However, generation of agriculture data has increased over past years and can be used for a growing number of applications of data mining techniques in agriculture. The multidisciplinary approach of integrating computer science with agriculture will support the necessary decisions to be taken in order to mitigate risks and maximize profits. The present study analyzes different methods of regression applied in the study case of grapes production forecast. The selected methods were multivariate linear regression, regression trees, lasso and random forest. Their performance were compared against the predictions obtained by the company through the mean squared error and the coefficient of variation. The four regression methods used obtained better predictive results than the method used by the company with statistical significance < 0.5%. © Springer International Publishing AG, part of Springer Nature 2018.

Abstract
This paper presents a control method for modular multilevel converters (MMCs) as an interface between renewable energy sources and the grid. With growing penetration of renewable energy sources in the power grid, the developments in converter technologies and controller designs become more prominent. In this regard, dynamic and steady state analysis of the proposed model for an MMC use in a renewable energy based power system are provided through dc, 1st, and 2nd harmonic models of the converter in dq reference frame. This detailed configuration is then used to accomplish converter modulation and controller design. The first novel contribution of this control method is to provide an accurate pulse width modulation (PWM) strategy based on network and converter parameters, in order to achieve a stable operation for the interfaced MMC during connection of renewable energy sources into the power grid. In addition, the proposed method is able to mitigate the converter circulating current by inserting a second harmonic reference in the modulation process of the MMC, which is the second contribution this paper provides over other control techniques. A capacitor voltage balancing algorithm is also included in this control method to adjust each sub-module (SM) voltage within an acceptable range. Finally, converter's maximum stable operation range is determined based on the dynamic equations of the proposed model. The functionality of the proposed control method is demonstrated by detailed mathematical analysis and comprehensive simulations with MATLAB/Simulink.

Abstract
This paper proposes an adaptive Biased Random-key Genetic Algorithm (A-BRKGA), a new method with on-line parameter control for combinatorial optimization problems. A-BRKGA has only one problem-dependent component, the decoder and all other parts can be reused. To control diversification and intensification, a novel adaptive strategy for parameter tuning is introduced. This strategy is based on deterministic rules and self adaptive schemes. For exploitation of specific regions of the solution space we propose a local search in promising communities. The proposed method is evaluated on the Capacitated Centered Clustering Problem (CCCP), which is an NP-hard problem where a set of n points, each having a given demand, is partitioned into m clusters each with a given capacity. The objective is to minimize the sum of the Euclidean distances between the points and their geometric cluster centroids. Computational results show that the A-BRKGA with local search is competitive with other methods of literature.