Abstract
This paper describes a two-month summer intensive course designed to introduce participants with a hands-on technical craft on robotics and to acquire experience in the low-level details of embedded systems. Attendants started this course with a brief introduction to robotics; learned to draw, design and create a personalized 3D structure for their mobile robotic platform and developed skills in embedded systems. They were familiarize with the practices used in robotics, learning to connect all sensors and actuator, developing a typical application on differential kinematic using Arduino, exploring ROS features under Raspberry Pi environment and Arduino – Raspberry Pi communication. Different paradigms and some real applications and programming were addressed on the topic of Artificial Intelligence. Throughout the course, participants were introduced to programming languages (including Python and C++), advanced programming concepts such as ROS, basic API development, system concepts such as I2C and UART serial interfaces, PWM motor control and sensor fusion to improve robotic navigation and localization. This paper describes not just the concept, layout and methodology used on RobotCraft 2017 but also presents the participants knowledge background and their overall opinions, leading to focus on lessons learned and suggestions for future editions. © 2018 The Authors

Abstract
The predawn leaf water potential (psi(pd)) is an eco-physiological indicator widely used for assessing vines water status and thus supporting irrigation management in several wine regions worldwide. However, the.pd is measured in a short time period before sunrise and the collection of a large sample of points is necessary to adequately represent a vineyard, which constitute operational constraints. In the present study, an alternative method based on hyperspectral data derived from a handheld spectroradiometer and machine learning algorithms was tested and validated for assessing grapevine water status. Two test sites in Douro wine region, integrating three grapevine cultivars, were studied for the years of 2014, 2015, and 2017. Four machine learning regression algorithms were tested for predicting the psi(pd) as a continuous variable, namely Random Forest (RF), Bagging Trees (BT), Gaussian Process Regression (GPR), and Variational Heteroscedastic Gaussian Process Regression (VH-GPR). Three predicting variables, including two vegetation indices (NRI554,561 and WI900,970) and a time-dynamic variable based on the psi(pd) (psi(pd_0)), were applied for modelling the response variable (psi(pd)). Additionally, the predicted values of psi(pd) were aggregated into three classes representing different levels of water deficit (low, moderate, and high) and compared with the corresponding classes of.pd observed values. A root mean square error (RMSE) and a mean absolute error (MAE) lower or equal than 0.15 MPa and 0.12 MPa, respectively, were obtained with an external validation data set (n= 71 observations) for the various algorithms. When the modelling results were assessed through classes of values, a high overall accuracy was obtained for all the algorithms (82-83%), with prediction accuracy by class ranging between 79% and 100%. These results show a good performance of the predictive models, which considered a large variability of climatic, environmental, and agronomic conditions, and included various grape cultivars. By predicting both continuous values of.pd and classes of psi(pd), the approach presented in this study allowed obtaining 2-levels of accurate information about vines water status, which can be used to feed management decisions of different types of stakeholders.

Abstract
Hyperspectral data collected through a handheld spectroradiometer (400-1010 nm) were tested for assessing the grapevine predawn leaf water potential (psi(pd)) measured by a Scholander chamber in two test sites of Douro wine region. The study was implemented in 2017, being a year with very hot and dry summer, conditions prone to severe water shortage. Three grapevine cultivars, 'Touriga Nacional', 'Touriga Franca' and 'Tinta Barroca' were sampled both in rainfed and irrigated vineyards, with a total of 325 plants assessed in four post-flowering dates. A large set of vegetation indices computed with the hyperspectral data and optimized for the psi(pd) values, as well as structural variables, were used as predictors in the model. From a total of 631 possible predictors, four variables were selected based on a stepwise forward procedure and the Wald statistics: irrigation treatment, test site, Anthocyanin Reflectance Index Optimized (ARI(opt_656,647)) and Normalized Ratio Index (NRI711,700). An ordinal logistic regression model was calibrated using 70 % of the dataset randomly selected and the 30 of the remaining observations where used in model validation. The overall model accuracy obtained with the validation dataset was 73.2 %, with the class of psi(pd) corresponding to the high-water deficit presenting a positive prediction value of 79.3 %. The accuracy and operability of this predictive model indicates good perspectives for its use in the monitoring of grapevine water status, and to support the irrigation tasks.

Abstract
The advances achieved during the last 30 years demonstrate the aptitude of the remote sensing-based vegetation indices (VI) for the assessment of crop evapotranspiration (ETc) and irrigation requirements in a simple, robust and operative manner. The foundation of these methodologies is the well-established relationship between the VIs and the basal crop coefficient (K-cb), resulting from the ability of VIs to measure the radiation absorbed by the vegetation, as the main driver of the evapotranspiration process. In addition, VIs have been related with single crop coefficient (K-c), assuming constant rates of soil evaporation. The direct relationship between VIs and ET is conceptually incorrect due to the effect of the atmospheric demand on this relationship. The rising number of Earth Observation Satellites potentiates a data increase to feed the VI-based methodologies for estimating and mapping either the K-c or K-cb, with improved temporal coverage and spatial resolution. The development of operative platforms, including satellite constellations like Sentinels and drones, usable for the assessment of K-cb through VIs, opens new possibilities and challenges. This work analyzes some of the questions that remain inconclusive at scientific and operational level, including: (i) the diversity of the K-cb-VI relationships defined for different crops, (ii) the integration of K-cb-VI relationships in more complex models such as soil water balance, and (iii) the operational application of K-cb-VI relationships using virtual constellations of space and aerial platforms that allow combining data from two or more sensors.

Abstract
Land cover maps obtained at high spatial and temporal resolutions are necessary to support monitoring and management applications in areas with many smallholder and low-input agricultural systems, as those characteristic in Mozambique. Various regional and global land cover products based on Earth Observation data have been developed and made publicly available but their application in regions characterized by a large variety of agro-systems with a dynamic nature is limited by several constraints. Challenges in the classification of spatially heterogeneous landscapes, as in Mozambique, include the definition of the adequate spatial resolution and data input combinations for accurately mapping land cover. Therefore, several combinations of variables were tested for their suitability as input for random forest ensemble classifier aimed at mapping the spatial dynamics of smallholder agricultural landscape in Vilankulo district in Mozambique. The variables comprised spectral bands from Landsat 7 ETM+ and Landsat 8 OLI/TIRS, vegetation indices and textural features and the classification was performed within the Google Earth Engine cloud computing for the years 2012, 2015, and 2018. The study of three different years aimed at evaluating the temporal dynamics of the landscape, typically characterized by high shifting nature. For the three years, the best performing variables included three selected spectral bands and textural features extracted using a window size of 25. The classification overall accuracy was 0.94 for the year 2012, 0.98 for 2015, and 0.89 for 2018, suggesting that the produced maps are reliable. In addition, the areal statistics of the class classified as agriculture were very similar to the ground truth data as reported by the Servicos Distritais de Actividades Economicas (SDAE), with an average percentage deviation below 10%. When comparing the three years studied, the natural vegetation classes are the predominant covers while the agriculture is the most important cause of land cover changes.

Abstract
The impact of climate on wine production (WP) temporal cycles in Douro (DR) and Vinhos Verdes (VVR) wine regions for a period of about 80 years, characterized by strong technological trend and climate variability, was modelled. The cyclical properties of WP, and which cycles are determined by spring temperature (ST) and soil water during summer (SW), were identified. It was achieved by applying a time-frequency approach, which is based on Kalman filter in the time domain. The time-varying autoregressive model can explain more than 67% (DR) and 95% (VVR) of the WP' variability and the integration of the ST and mainly SW increase the models' reliability. The results were then transferred into the frequency domain, and can show that WP in both regions is characterized by two cycles close to 5-6 and 2.5 years around the long run trend. The ST and SW showed great capacity to explain the cyclicality of WP in the studied regions being the coherence temporarily much more stable in VVR than in the DR, where a shift of the relative importance away from ST to SW can be recognized. This could be an indicator of lower impact of the foreseen hot and dry climate scenarios on WP in the regions with a maritime climate, such as the VVR, compared with hot and dry wine regions. Despite the marked differences in the two studied regions on ecological, viticulture practices and technological trend, the modelling approach based on time-frequency proved to be an efficient tool to infer the impact of climate on the dynamics of cyclical properties of regional WP, foreseeing its generalized use in other regions. This modelling approach can be an important tool for planning in the wine industry as well as for mitigation strategies facing the scenarios that combine technological progress and climate change.