Abstract
Slight variations in user interface response times can significantly impact the user experience provided by an interface. Load testing is used to evaluate how an application behaves under increasing loads. For interactive applications, load testing can be done by directly calling services at the business logic or through the user interface. In modern web applications, there is a considerable amount of control logic on the browser side. The impact of this logic on applications' behaviour is only fully considered if the tests are done through the user interface. Capture reply tools are used for this, but their use can become costly. Leveraging an existing model-based testing tool, we propose an approach to automate load testing done through the user interface.

Abstract
The exploration of the ocean has got an increasing interest, including activities such as offshore wind farms and deep-sea mining. However, the ocean environment and the high cost of operations, namely for manned missions, have led to the development of Autonomous Underwater Vehicles (AUVs) and other sensing platforms. AUVs play a vital role in these environments, relying on communications systems to operate and exchange sensor data. Yet, reliable and energy-efficient broad-band wireless communications underwater remain an unsolved challenge, despite the recent advances in the field. We present a novel multimodal approach, named DURIUS, that considers the movement of the AUV to convey the sensor data and selects the most suitable underwater wireless communications technology - acoustic, optical or radio - according to the underwater context, targeting maximum performance and minimum energy consumption. Our analytical results show that DURIUS increases data throughput and reduces energy consumption when compared with the state of the art approaches.

Abstract

Abstract
This work considers the problem of segmenting heart sounds into their fundamental components. We unify statistical and data-driven solutions by introducing Markov-based Neural Networks (MNNs), a hybrid end-toend framework that exploits Markov models as statistical inductive biases for an Artificial Neural Network (ANN) discriminator. We show that an MNN leveraging a simple onedimensional Convolutional ANN significantly outperforms two recent purely data-driven solutions for this task in two publicly available datasets: PhysioNet 2016 (Sensitivity: 0.947 +/- 0.02; Positive Predictive Value : 0.937 +/- 0.025) and the CirCor DigiScope 2022 (Sensitivity: 0.950 +/- 0.008; Positive Predictive Value: 0.943 +/- 0.012). We also propose a novel gradient-based unsupervised learning algorithm that effectively makes the MNN adaptive to unseen datum sampled from unknown distributions. We perform a cross dataset analysis and show that an MNN pre-trained in the CirCor DigiScope 2022 can benefit from an average improvement of 3.90% Positive Predictive Value on unseen observations from the PhysioNet 2016 dataset using this method.

Abstract
This study presents a novel approach to cope with the human behaviour uncertainty during Human-Robot Collaboration (HRC) in dynamic and unstructured environments, such as agriculture, forestry, and construction. These challenging tasks, which often require excessive time, labour and are hazardous for humans, provide ample room for improvement through collaboration with robots. However, the integration of humans in-the-loop raises open challenges due to the uncertainty that comes with the ambiguous nature of human behaviour. Such uncertainty makes it difficult to represent high-level human behaviour based on low-level sensory input data. The proposed Fuzzy State-Long Short-Term Memory (FS-LSTM) approach addresses this challenge by fuzzifying ambiguous sensory data and developing a combined activity recognition and sequence modelling system using state machines and the LSTM deep learning method. The evaluation process compares the traditional LSTM approach with raw sensory data inputs, a Fuzzy-LSTM approach with fuzzified inputs, and the proposed FS-LSTM approach. The results show that the use of fuzzified inputs significantly improves accuracy compared to traditional LSTM, and, while the fuzzy state machine approach provides similar results than the fuzzy one, it offers the added benefits of ensuring feasible transitions between activities with improved computational efficiency.

Abstract
A área de conhecimento associada à gestão de ativos potencia a extração de valor de ativos físicos, incorporando uma abordagem integrada que envolve riscos, custos e desempenho. Este trabalho propõe uma nova abordagem para selecionar projetos de investimento em infraestruturas no âmbito da gestão de ativos. A métrica padrão utilizada para a seleção de projetos em ativos múltiplos é o Índice de Valor de Infraestruturas (IVI), que tem sido utilizado para caracterizar ativos do setor de águas em diversos países. Este trabalho desenvolveu dois modelos de programação linear inteira mista. O primeiro modelo de otimização trata de um problema de alocação de orçamento de capital selecionando projetos de investimento com vista à maximização da condição dos ativos. O segundo modelo de otimização trata de um problema de planeamento de orçamento, visando minimizar o capital necessário para manter os ativos nas condições desejadas. A aplicação dos modelos é ilustrada por um estudo de caso de uma empresa que opera no setor da água em Portugal. Os resultados obtidos nos diferentes cenários analisados são discutidos com vista a apoiar o desenho de políticas de gestão de ativos mais eficientes.