Abstract

Abstract
In this work a hybrid simulation-optimization approach is presented to support decision-making towards improved resiliency and sustainability in pharmaceutical supply chain (PSC) operations. In a first step, a simulation model is used to assess the PSC performance under a set of disruptive scenarios to select the best inventory-based strategy for enhanced resiliency. Disruptions addressed in this work are mainly related to unpredicted medium-term production stoppages due to unexpected high-impact events such as accidents in production and transportation, or natural disasters. In a second step, a multi-objective mixed integer linear programming (MO-MILP) model is developed to optimize the selected inventory-based strategy regarding the economic, social, and environmental dimensions. In particular, the social and environmental aspects are introduced by anticipating the expected waste generation of close to expire medicines, redirecting them into a donation scheme. The proposed approach is applied to a representative PSC, with preliminary results showing the relevance of this tool for decision-makers to assess the trade-offs associated to the economic and social dimensions, as well as their impacts on waste generation.

Abstract
The practical application of knowledge acquired during undergraduate studies is crucial for students to address real-world problems and seek solutions. The SMAR PD3 didactic plant provides a conducive environment for experiments in systems such as level and flow, common in various industrial sectors. Cascade control, an approach that sequentially uses two or more controllers, stands out as a promising strategy to enhance precision and stability in industrial processes. This work proposes a study on cascade control in flow and level systems, demonstrating its application in the didactic plant. The process involved system identification, tuning of conventional and cascade PI and PID controllers, followed by the implementation of the Successive Loop Closure technique. Results, in line with specialized literature, indicate that the implementation of cascade controllers in the industry can improve specific processes affected by disturbances or changes in variables, directly impacting the overall functioning of the process.

Abstract
Nowadays, the industrial market is characterised by high levels of competition, with customers increasingly demanding in terms of quality, delivery times, costs, etc.. However, with increasing demand and the need to increase productivity, many companies in recent years have dedicated themselves to decentralising their factories, thus moving to distributed production. Today's manufacturing systems are distributed in the sense that there are several jobs that have to be carry out on machines located in different factories. This paper proposes a multi-objective distributed job shop scheduling model with unrelated parallel machines and sequence-dependent setup times. The transport time of raw materials to carry out a given job to a factory is also taken into account. Small instances of the problem were solved using NSGA-III with the aim of simultaneously minimising two objectives: the makespan and average completion time. Preliminary results show the validity of this approach.

Abstract
Medical image analysis is crucial for the efficient diagnosis of many diseases. Typically, hospitals maintain vast repositories of images, which can be leveraged for various purposes, including research. However, access to such image collections is largely restricted to safeguard the privacy of the individuals whose images are being stored, as data protection concerns come into play. Recently, the development of solutions for Automated Medical Image Analysis has gained significant attention, with Deep Learning being one solution that has achieved remarkable results in this area. One promising approach for medical image analysis is Federated Learning (FL), which enables the use of a set of physically distributed data repositories, usually known as nodes, satisfying the restriction that the data do not leave the repository. Under these conditions, FL can build high-quality, accurate deep-learning models using a lot of available data wherever it is. Therefore, FL can help researchers and clinicians diagnose diseases and support medical decisions more efficiently and robustly. This article provides a systematic survey of FL in medical image analysis, specifically based on Magnetic Resonance Imaging, Computed Tomography, X-radiography, and histology images. Hence, it discusses applications, contributions, limitations, and challenges and is, therefore, suitable for those who want to understand how FL can contribute to the medical imaging domain.

Abstract
An optical fiber pH sensor based on a long-period fiber grating (LPFG) is reported. Two oppositely charged polymers, polyethylenimine (PEI) and polyacrylic acid (PAA), were alternately deposited on the sensing structure through a layer-by-layer (LbL) electrostatic self-assembly technique. Since the polymers are pH sensitive, their refractive index (RI) varies when the pH of the solution changes due to swelling/deswelling phenomena. The fabricated multilayer coating retained a similar property, enabling its use in pH-sensing applications. The pH of the PAA dipping solution was tuned so that a coated LPFG achieved a pH sensitivity of (6.3 +/- 0.2) nm/pH in the 5.92-9.23 pH range. Only two bilayers of PEI/PAA were used as an overlay, which reduces the fabrication time and increases the reproducibility of the sensor, and its reversibility and repeatability were demonstrated by tracking the resonance band position throughout multiple cycles between different pH solutions. With simulation work and experimental results from a low-finesse Fabry-Perot (FP) cavity on a fiber tip, the coating properties were estimated. When saturated at low pH, it has a thickness of 200 nm and 1.53 +/- 0.01 RI, expanding up to 310 nm with a 1.35 +/- 0.01 RI at higher pH values, mostly due to the structural changes in the PAA.