Abstract
The ability to predict the maintenance needs of machines is generating increasing interest in a wide range of indus-tries as it contributes to diminishing machine downtime and costs while increasing efficiency when compared to traditional maintenance approaches. Predictive maintenance (PdM) methods, based on state-of-the-art Internet of Things (IoT) systems and Artificial Intelligence (AI) techniques, are heavily dependent on data to create analytical models capa-ble of identifying certain patterns which can represent a mal-function or deterioration in the monitored machines. There-fore, a realistic and representative dataset is paramount for creating, training, and validating PdM techniques. This pa-per introduces a new dataset, which integrates real-world data from home appliances, such as refrigerators and wash-ing machines, suitable for the development and testing of PdM algorithms. The data was collected on various home ap-pliances at a repair center and included readings of elec-trical current and vibration at low (1 Hz) and high (2048 Hz) sampling frequencies. The dataset samples are filtered and tagged with both normal and malfunction types. An ex-tracted features dataset, corresponding to the collected work-ing cycles is also made available. This dataset could bene- fit research and development of AI systems for home ap-pliances' predictive maintenance tasks and outlier detection analysis. The dataset can also be repurposed for smart-grid or smart-home applications, predicting the consumption pat-terns of such home appliances.(c) 2023 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )

Abstract
Machine Learning has been overtaken by a growing necessity to explain and understand decisions made by trained models as regulation and consumer awareness have increased. Alongside understanding the inner workings of a model comes the task of verifying how adequately we can model a problem with the learned functions. Traditional global assessment functions lack the granularity required to understand local differences in performance in different regions of the feature space, where the model can have problems adapting. Residual Analysis adds a layer of model understanding by interpreting prediction residuals in an exploratory manner. However, this task can be unfeasible for high-dimensionality datasets through hypotheses and visualizations alone. In this work, we use weak interpretable learners to identify regions of high prediction error in the feature space. We achieve this by examining the absolute residuals of predictions made by trained regressors. This methodology retains the interpretability of the identified regions. It allows practitioners to have tools to formulate hypotheses surrounding model failure on particular regions for future model tunning, data collection, or data augmentation on critical cohorts of data. We present a way of including information on different levels of model uncertainty in the feature space through the use of locally fitted Model Agnostic Prediction Intervals (MAPIE) in the identified regions, comparing this approach with other common forms of conformal predictions which do not take into account findings from weak segment identification, by assessing local and global coverage of the prediction intervals. To demonstrate the practical application of our approach, we present a real-world industry use case in the context of inbound retention call-centre operations for a Telecom Provider to determine optimal pairing between a customer and an available assistant through the prediction of contracted revenue. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Abstract
Storing energy efficiently is one of the main factors of a more sustainable world. The battey management system in energy storage plays an extremely important role in ensuring these systems' efficiency, safety, and performance. This battery management system is capable of estimating the battery states, which are used to give better efficiency, a long life cycle, and safety. However, these states cannot be measured directly and must be estimated indirectly using battery models. Therefore, accurate battery models are essential for battery management systems implementation. One of these models is the nonlinear grey box model, which is easy to implement in embedded systems and has good accuracy when used with a good parameter identification method. Regarding the parameter identification methods, the nonlinear least square optimization is the most used method. However, to have accurate results, it is necessary to define the system's initial states, which is not an easy task. This paper presents a two-outputs nonlinear grey box battery model. The first output is the battery voltage, and the second output is the battery state of charge. The second output was added to improve the system's initial states identification and consequently improve the identified parameter accuracy. The model was estimated with the best experiment design, which was defined considering a comparison between seven different experiment designs regarding the fit to validation data, the parameter standard deviation, and the output variance. This paper also presents a method for defining a weight between the outputs, considering a greater weight in the output with greater model confidence. With this approach, it was possible to reach a value 1000 times smaller in the parameter standard deviation with a non-biased and little model prediction error when compared to the commonly used one-output nonlinear grey box model.

Abstract

Abstract
We live in an era in which the preservation of the environment is being widely discussed, driven by growing concerns over climate issues. One major factor contributing to this situation is the lack of attention societies give to maintaining high sustainability levels. Data plays a crucial role in understanding and assessing sustainability impacts in both urban and rural areas. However, obtaining comprehensive data on a country's sustainability is challenging due to the lack of simple and accessible sources. Existing solutions for sustainability analysis are limited by high costs and implementation difficulties, which restrict their spatial coverage. In this paper, we propose a solution using low-cost hardware and open-source technologies to collect data about the movement of people and vehicles. This solution involves low-cost video-based meters that can be flexibly deployed to various locations. Specifically, we developed a prototype using Raspberry Pi and YOLO which is able to correctly classify 91% of the vehicles by type, and 100% of the events (entering of leaving). The results indicate that this system can effectively and affordably identify and count people and vehicles, allowing for its implementations namely in remote sensitive areas such as natural parks, in which the access of people and vehicles must be controlled and monitored. © 2023 The authors and IOS Press.

Abstract
Regardless of the goal of a game, it should be a pleasant and fun experience for its players. For some games to be enjoyable, the level of difficulty must be carefully calibrated, otherwise, players will feel bored or frustrated. Multiplayer scenarios in particular, where one player's satisfaction might not translate to the enjoyment of other players and poses extra challenges in balancing the difficulty. The performance of one player is relative to the opponent, versus single-player scenarios where we can fully control the environment. We propose an AI automation framework for difficulty balancing in two-player games, where balancing is seen as a Reinforcement Learning task. A Game Master (GM) agent learns how to use handicap game mechanics, signaled by a reward function that evaluates a weighted combination of aesthetic criteria that encourages dramatization and allows a player in the lead to go back and a player in the rear to catch up, creating the desired rubber banding effect that balances out skill gaps. The quality of the games with the trained GM embedded is examined by measuring the same aesthetic criteria on the resulting games, and by analyzing the resulting changes in the game.