The energy sector is currently undergoing a significant shift, driven by the growing integration of renewable energy sources and the decentralization of electricity markets, which are now extending into local communities. This transformation highlights the pivotal role of prosumers within these markets, and as a result, the concept of Renewable Energy Communities is gaining traction, empowering their members to curtail reliance on non-renewable energy sources by facilitating local energy generation, storage, and exchange. Also in a community, management efficiency depends on being able to predict future consumption to make decisions regarding the purchase, sale and storage of electricity, which is why forecasting the consumption of community members is extremely important. This study presents an innovative approach to manage community energy balance, relying on Machine Learning (ML) techniques, namely eXtreme Gradient Boosting (XGBoost), to forecast electricity consumption. Subsequently, a decision algorithm is employed for energy trading with the public grid, based on solar production and energy consumption forecasts, storage levels and market electricity prices. The outcomes of the simulated model demonstrate the efficacy of incorporating these techniques, since the system showcases the potential to reduce both the community electricity expenses and its dependence on energy from the centralized distribution grid. ML-based techniques allowed better results specially for bi-hourly tariffs and high storage capacity scenarios with community bill reductions of 9.8%, 2.8% and 5.4% for high, low, and average photovoltaic (PV) generation levels, respectively.

Forecasting road flow has strong importance for both allowing authorities to guarantee safety conditions and traffic efficiency, as well as for road users to be able to plan their trips according to space and road occupation. In a summer resort, such as beaches near cities, traffic depends directly on weather conditions, variables that should be of great impact on the quality of forecasts. Will the use of a dataset with information on transit flows enhanced with meteorological information allow the construction of a precise traffic flow forecasting model, allowing predictions to be made in advance of the traffic flow in suitable time? The present work evaluates different machine learning methods, namely long short-term memory, autoregressive LSTM, and a convolutional neural network, and data attributes to predict traffic flows based on radar and meteorological sensor information. The models trained to predict the traffic flow have shown that weather conditions were essential for this forecast, and thus, these variables were employed in the evaluated deep-learning models. The results pointed out that it is possible to forecast the traffic flow at a reasonable error level for one-hour periods, and the CNN model presented the lowest prediction error values and consumed the least time to build its predictions.

In the current economy, companies are increasingly interested in optimizing their logistics operations to reduce costs and increase agility. Transport logistics is one of the processes to be optimized, since companies have a limited fleet of heterogeneous vehicles, with particular capacities and costs, and have to attend to their customers within restricted periods. These features characterize the problem as a Heterogeneous Fixed Fleet Vehicle Routing Problem with Time Window (HFVRPTW). To solve this problem, this work proposes a method based on the Adaptive Large Neighborhood Search (ALNS) metaheuristic particularly focused on selecting vehicles that reduce the costs of the used fleet. The experiments showed that the proposed method improved \(69.6\%\) of the benchmark instances compared with the literature state-of-the-art, with \(0.44\%\) of average reduction in the total cost. Besides that, the implemented ALNS algorithm was around \(35\) times faster to run than the most relevant compared work.

The amount of spatiotemporal data collected by gadgets is rapidly growing, resulting in increasing costs to transfer, process and store it. In an attempt to minimize these costs several algorithms were proposed to reduce the trajectory size. However, to choose the right algorithm depends on a careful analysis of the application scenario. Therefore, this paper evaluates seven general purpose lossy compression algorithms in terms of structural aspects and performance characteristics, regarding four transportation modes: Bike, Bus, Car and Walk. The lossy compression algorithms evaluated are: Douglas-Peucker (DP), Opening-Window (OW), Dead-Reckoning (DR), Top-Down Time-Ratio (TS), Opening-Window Time-Ratio (OS), STTrace (ST) and SQUISH (SQ). Pareto Efficiency analysis pointed out that there is no best algorithm for all assessed characteristics, but rather DP applied less error and kept length better-preserved, OW kept speed better-preserved, ST kept acceleration better-preserved and DR spent less execution time. Another important finding is that algorithms that use metrics that do not keep time information have performed quite well even with characteristics time-dependent like speed and acceleration. Finally, it is possible to see that DR had the most suitable performance in general, being among the three best algorithms in four of the five assessed performance characteristics.