Applied deep learning and data science with a human-centric and data-centric approach

The term Artificial intelligence acquired a lot of baggage since its introduction and in its current incarnation is synonymous with Deep Learning. The sudden availability of data and computing resources has opened the gates to myriads of applications. Not all are created equal though, and problems might arise especially for fields not closely related to the tasks that pertain tech companies that spearheaded DL. The perspective of practitioners seems to be changing, however. Human-Centric AI emerged in the last few years as a new way of thinking DL and AI applications from the ground up, with a special attention at their relationship with humans. The goal is designing a system that can gracefully integrate in already established workflows, as in many real-world scenarios AI may not be good enough to completely replace its humans. Often this replacement may even be unneeded or undesirable. Another important perspective comes from, Andrew Ng, a DL pioneer, who recently started shifting the focus of development from “better models” towards better, and smaller, data. He defined his approach Data-Centric AI. Without downplaying the importance of pushing the state of the art in DL, we must recognize that if the goal is creating a tool for humans to use, more raw performance may not align with more utility for the final user. A Human-Centric approach is compatible with a Data-Centric one, and we find that the two overlap nicely when human expertise is used as the driving force behind data quality. This thesis documents a series of case-studies where these approaches were employed, to different extents, to guide the design and implementation of intelligent systems. We found human expertise proved crucial in improving datasets and models. The last chapter includes a slight deviation, with studies on the pandemic, still preserving the human and data centric perspective.

Innovation in conceptual design of chemical processes

Nowadays, the chemical industry has reached significant goals to produce essential components for human being. The growing competitiveness of the market caused an important acceleration in R&D activities, introducing new opportunities and procedures for the definition of process improvement and optimization. In this dynamicity, sustainability is becoming one of the key aspects for the technological progress encompassing economic, environmental protection and safety aspects. With respect to the conceptual definition of sustainability, literature reports an extensive discussion of the strategies, as well as sets of specific principles and guidelines. However, literature procedures are not completely suitable and applicable to process design activities. Therefore, the development and introduction of sustainability-oriented methodologies is a necessary step to enhance process and plant design. The definition of key drivers as support system is a focal point for early process design decisions or implementation of process modifications. In this context, three different methodologies are developed to support design activities providing criteria and guidelines in a sustainable perspective. In this framework, a set of key Performance Indicators is selected and adopted to characterize the environmental, safety, economic and energetic aspects of a reference process. The methodologies are based on heat and material balances and the level of detailed for input data are compatible with available information of the specific application. Multiple case-studies are defined to prove the effectiveness of the methodologies. The principal application is the polyolefin productive lifecycle chain with particular focus on polymerization technologies. In this context, different design phases are investigated spanning from early process feasibility study to operative and improvements assessment. This flexibility allows to apply the methodologies at any level of design, providing supporting guidelines for design activities, compare alternative solutions, monitor operating process and identify potential for improvements.

Upper-Limb Exoskeleton Design for Firefighter

Every day, firefighters are involved in emergency response tasks, which are both physically and psychologically exhausting. According to the National Fire Protection Association (NFPA), the number of firefighters who are injured or die while performing their job is incredibly high. When firefighters are injured, they must follow a rehabilitation therapy program to physically recover and depending on the severity of their injuries they may not fully recuperate at all. If they sustain a permanent injury that they cannot recover from, they may be out of work for the rest of their career. This research focuses on studying and developing a special device, known as an exoskeleton, aimed at assisting and preventing potential injuries among firefighters. Nowadays, the usage of human exoskeletons is becoming more common in a variety of fields. In fact, it is currently being researched and developed for soldiers, athletes, and critical care patients around the world. Most of the existing exoskeletons have been developed for the assistance of the lower human body. The research that I have done in my thesis instead relates to mobility of the upper body. Many of the existing exoskeletons have been analyzed and compared to each other and the human body, such as the study of human arm parts and their movements around three principal joints: shoulder, elbow, and wrist. The correct design of the shoulder exoskeleton join is still a big challenge for designers because of the complexity of biomechanical human movements. The exoskeleton must fit perfectly to the human body, otherwise it could be harmful for both the recovery and the safety of the user. The goal of this thesis is to design an upper-body arm exoskeleton worn by firefighters and develop and test a PID control system to prevent the risk of injuries while performing their job.

Riprogettare un corso di studi: bestpractices, progettazione human-centered e applicazioni dell'Università di Bologna

Questa tesi tratterà di come progettare un corso di studi in modo che sia Human-centered, dal punto di vista degli stakeholder coinvolti nel processo. Il processo inizia tramite benchmarking, per questo motivo inizialmente si racconteranno di ricerche sul benchmarking e quanto ciò sia ritenuto importante nei processi di questo tipo. Inoltre, si racconterà dello strumento utilizzato in questa tesi ossia il Design Thinking e di come sia stato utilizzato in letteratura in modo da sottolinearne l’importanza. In seguito, si vedrà come si crea un corso di studio nell’Università di Bologna e i problemi affrontati dai diversi attori che lo coinvolgono. In particolare, sono state mappate le procedure attuali e la loro applicazione nei diversi corsi di studio. In questo modo è stata costruita una BPMN finale in cui sono evidenziati i diversi attori del processo e le varie fasi che lo compongono, insieme al periodo di svolgimento del processo. Successivamente, sono state raccolte le interviste degli stakeholder coinvolti insieme ai loro bisogni e alle problematiche riscontrate nel processo sopra descritto. Da questa analisi è stata rappresentata una stakeholder map con gli attori assieme alle loro necessità. In un secondo tempo, si è pensato a come riprogettare un corso di studio affinché sia più human centered. Per questo motivo, il tema è affrontato dal punto di vista dell’università in modo da capire come poter cambiare un corso di studio o di laurea per favorire quanto descritto precedentemente. In questo scritto sono inoltre stabilite delle linee guida di questo processo in modo da risolvere alcuni dei problemi trattati, nel migliore dei modi. Sono anche raccolte delle interviste in cui il to be del processo è riproposto ai diversi stakeholder e le loro impressioni sul processo. Infine, sono tratte le conclusioni e le best practice finali del processo passate al vaglio dai diversi stakeholder.

Design and implementation of an optimal motion cueing algorithm for the Stuttgart Driving Simulator

Driving simulators emulate a real vehicle drive in a virtual environment. One of the most challenging problems in this field is to create a simulated drive as real as possible to deceive the driver's senses and cause the believing to be in a real vehicle. This thesis first provides an overview of the Stuttgart driving simulator with a description of the overall system, followed by a theoretical presentation of the commonly used motion cueing algorithms. The second and predominant part of the work presents the implementation of the classical and optimal washout algorithms in a Simulink environment. The project aims to create a new optimal washout algorithm and compare the obtained results with the results of the classical washout. The classical washout algorithm, already implemented in the Stuttgart driving simulator, is the most used in the motion control of the simulator. This classical algorithm is based on a sequence of filters in which each parameter has a clear physical meaning and a unique assignment to a single degree of freedom. However, the effects on human perception are not exploited, and each parameter must be tuned online by an engineer in the control room, depending on the driver's feeling. To overcome this problem and also consider the driver's sensations, the optimal washout motion cueing algorithm was implemented. This optimal control-base algorithm treats motion cueing as a tracking problem, forcing the accelerations perceived in the simulator to track the accelerations that would have been perceived in a real vehicle, by minimizing the perception error within the constraints of the motion platform. The last chapter presents a comparison between the two algorithms, based on the driver's feelings after the test drive. Firstly it was implemented an off-line test with a step signal as an input acceleration to verify the behaviour of the simulator. Secondly, the algorithms were executed in the simulator during a test drive on several tracks.

Analysis and development of interface components and procedures for the design and commissioning toolchains in automation systems

The objective of the thesis project, developed within the Line Control & Software Engineering team of G.D company, is to analyze and identify the appropriate tool to automate the HW configuration process using Beckhoff technologies by importing data from an ECAD tool. This would save a great deal of time, since the I/O topology created as part of the electrical planning is presently imported manually in the related SW project of the machine. Moreover, a manual import is more error-prone because of human mistake than an automatic configuration tool. First, an introduction about TwinCAT 3, EtherCAT and Automation Interface is provided; then, it is analyzed the official Beckhoff tool, XCAD Interface, and the requirements on the electrical planning to use it: the interface is realized by means of the AutomationML format. Finally, due to some limitations observed, the design and implementation of a company internal tool is performed. Tests and validation of the tool are performed on a sample production line of the company.

Modelli informativi per la valutazione BIM del livello di reversibilità e circolarità di componenti edilizi: il Design for Disassembly

La Tesi presenta un metodo per valutare il livello di reversibilità e circolarità dei componenti edili. Il concetto cardine dello studio è il Design for Disassembly (DfD), definibile come un approccio alla progettazione di organismi edilizi secondo criteri costruttivi volti a facilitare successivi cambiamenti e smantellamenti (completi o parziali), consentendo il riutilizzo dei componenti, al fine di ridurre l’impatto ambientale degli interventi. Attualmente, diverse ricerche in ambito scientifico si dedicano all’approfondimento di questa tematica, correlandola ad altri concetti come la metodologia del Building Information Modeling (BIM), che consente di digitalizzare il processo progettuale, la sua realizzazione e la sua gestione attraverso modelli. Dopo l’analisi dello stato dell’arte, il lavoro è giunto alla definizione di un insieme di parametri idonei per essere introdotti in un modello informativo, in grado di rappresentare la circolarità del componente in termini di DfD. Per ogni elemento del componente analizzato viene assegnato un valore numerico (variabile da 0,1 a 1) a ogni parametro. Tramite l’utilizzo di una formula elaborata nell’ambito di precedenti ricerche svolte dal Dipartimento di Architettura dell'Università di Bologna, opportunamente modificata, si ottiene un indice sintetico finale denominato “Express Building Circularity Indicators” (EBCI). Il metodo di analisi proposto come strumento a supporto del processo progettuale è stato validato tramite l’applicazione a diverse soluzioni di facciata per l’efficientamento energetico di un fabbricato selezionato come caso di studio italiano dal progetto Europeo Horizon 2020 "DRIVE 0 – Driving decarbonization of the EU building stock by enhancing a consumer centred and locally based circular renovation process". I risultati ottenuti hanno consentito di verificare la replicabilità del processo digitalizzato a diverse soluzioni costruttive e l’affidabilità del metodo di valutazione del grado di circolarità.

Prognostic Value Of Dna And Mrna E6/e7 Of Human Papillomavirus In The Evolution Of Cervical Intraepithelial Neoplasia Grade 2.

This study aimed at evaluating whether human papillomavirus (HPV) groups and E6/E7 mRNA of HPV 16, 18, 31, 33, and 45 are prognostic of cervical intraepithelial neoplasia (CIN) 2 outcome in women with a cervical smear showing a low-grade squamous intraepithelial lesion (LSIL). This cohort study included women with biopsy-confirmed CIN 2 who were followed up for 12 months, with cervical smear and colposcopy performed every three months. Women with a negative or low-risk HPV status showed 100% CIN 2 regression. The CIN 2 regression rates at the 12-month follow-up were 69.4% for women with alpha-9 HPV versus 91.7% for other HPV species or HPV-negative status (P < 0.05). For women with HPV 16, the CIN 2 regression rate at the 12-month follow-up was 61.4% versus 89.5% for other HPV types or HPV-negative status (P < 0.05). The CIN 2 regression rate was 68.3% for women who tested positive for HPV E6/E7 mRNA versus 82.0% for the negative results, but this difference was not statistically significant. The expectant management for women with biopsy-confirmed CIN 2 and previous cytological tests showing LSIL exhibited a very high rate of spontaneous regression. HPV 16 is associated with a higher CIN 2 progression rate than other HPV infections. HPV E6/E7 mRNA is not a prognostic marker of the CIN 2 clinical outcome, although this analysis cannot be considered conclusive. Given the small sample size, this study could be considered a pilot for future larger studies on the role of predictive markers of CIN 2 evolution.

Solid Lipid Nanoparticles For Hydrophilic Biotech Drugs: Optimization And Cell Viability Studies (caco-2 & Hepg-2 Cell Lines).

Insulin was used as model protein to developed innovative Solid Lipid Nanoparticles (SLNs) for the delivery of hydrophilic biotech drugs, with potential use in medicinal chemistry. SLNs were prepared by double emulsion with the purpose of promoting stability and enhancing the protein bioavailability. Softisan(®)100 was selected as solid lipid matrix. The surfactants (Tween(®)80, Span(®)80 and Lipoid(®)S75) and insulin were chosen applying a 2(2) factorial design with triplicate of central point, evaluating the influence of dependents variables as polydispersity index (PI), mean particle size (z-AVE), zeta potential (ZP) and encapsulation efficiency (EE) by factorial design using the ANOVA test. Therefore, thermodynamic stability, polymorphism and matrix crystallinity were checked by Differential Scanning Calorimetry (DSC) and Wide Angle X-ray Diffraction (WAXD), whereas the effect of toxicity of SLNs was check in HepG2 and Caco-2 cells. Results showed a mean particle size (z-AVE) width between 294.6 nm and 627.0 nm, a PI in the range of 0.425-0.750, ZP about -3 mV, and the EE between 38.39% and 81.20%. After tempering the bulk lipid (mimicking the end process of production), the lipid showed amorphous characteristics, with a melting point of ca. 30 °C. The toxicity of SLNs was evaluated in two distinct cell lines (HEPG-2 and Caco-2), showing to be dependent on the concentration of particles in HEPG-2 cells, while no toxicity in was reported in Caco-2 cells. SLNs were stable for 24 h in in vitro human serum albumin (HSA) solution. The resulting SLNs fabricated by double emulsion may provide a promising approach for administration of protein therapeutics and antigens.

Integrated Proteomics Identified Up-regulated Focal Adhesion-mediated Proteins In Human Squamous Cell Carcinoma In An Orthotopic Murine Model.

Understanding the molecular mechanisms of oral carcinogenesis will yield important advances in diagnostics, prognostics, effective treatment, and outcome of oral cancer. Hence, in this study we have investigated the proteomic and peptidomic profiles by combining an orthotopic murine model of oral squamous cell carcinoma (OSCC), mass spectrometry-based proteomics and biological network analysis. Our results indicated the up-regulation of proteins involved in actin cytoskeleton organization and cell-cell junction assembly events and their expression was validated in human OSCC tissues. In addition, the functional relevance of talin-1 in OSCC adhesion, migration and invasion was demonstrated. Taken together, this study identified specific processes deregulated in oral cancer and provided novel refined OSCC-targeting molecules.

Hypertrophic Adenoid Is A Major Infection Site Of Human Bocavirus 1.

Human bocavirus 1 (HBoV1) is associated with respiratory infections worldwide, mainly in children. Similar to other parvoviruses, it is believed that HBoV1 can persist for long periods of time in humans, probably through maintaining concatemers of the virus single-stranded DNA genome in the nuclei of infected cells. Recently, HBoV-1 was detected in high rates in adenoid and palatine tonsils samples from patients with chronic adenotonsillar diseases, but nothing is known about the virus replication levels in those tissues. A 3-year prospective hospital-based study was conducted to detect and quantify HBoV1 DNA and mRNAs in samples of the adenoids (AD), palatine tonsils (PT), nasopharyngeal secretions (NPS), and peripheral blood (PB) from patients undergoing tonsillectomy for tonsillar hypertrophy or recurrent tonsillitis. HBoV1 was detected in 25.3% of the AD samples, while the rates of detection in the PT, NPS, and PB samples were 7.2%, 10.5%, and 1.7%, respectively. The viral loads were higher in AD samples, and 27.3% of the patients with HBoV had mRNA detectable in this tissue. High viral loads and detectable mRNA in the AD were associated with HBoV1 detection in the other sample sites. The adenoids are an important site of HBoV1 replication and persistence in children with tonsillar hypertrophy. The adenoids contain high HBoV1 loads and are frequently positive for HBoV mRNA, and this is associated with the detection of HBoV1 in secretions.

The Effect Of Celastrol, A Triterpene With Antitumorigenic Activity, On Conformational And Functional Aspects Of The Human 90kda Heat Shock Protein Hsp90α, A Chaperone Implicated In The Stabilization Of The Tumor Phenotype.

Hsp90 is a molecular chaperone essential for cell viability in eukaryotes that is associated with the maturation of proteins involved in important cell functions and implicated in the stabilization of the tumor phenotype of various cancers, making this chaperone a notably interesting therapeutic target. Celastrol is a plant-derived pentacyclic triterpenoid compound with potent antioxidant, anti-inflammatory and anticancer activities; however, celastrol's action mode is still elusive. In this work, we investigated the effect of celastrol on the conformational and functional aspects of Hsp90α. Interestingly, celastrol appeared to target Hsp90α directly as the compound induced the oligomerization of the chaperone via the C-terminal domain as demonstrated by experiments using a deletion mutant. The nature of the oligomers was investigated by biophysical tools demonstrating that a two-fold excess of celastrol induced the formation of a decameric Hsp90α bound throughout the C-terminal domain. When bound, celastrol destabilized the C-terminal domain. Surprisingly, standard chaperone functional investigations demonstrated that neither the in vitro chaperone activity of protecting against aggregation nor the ability to bind a TPR co-chaperone, which binds to the C-terminus of Hsp90α, were affected by celastrol. Celastrol interferes with specific biological functions of Hsp90α. Our results suggest a model in which celastrol binds directly to the C-terminal domain of Hsp90α causing oligomerization. However, the ability to protect against protein aggregation (supported by our results) and to bind to TPR co-chaperones are not affected by celastrol. Therefore celastrol may act primarily by inducing specific oligomerization that affects some, but not all, of the functions of Hsp90α. To the best of our knowledge, this study is the first work to use multiple probes to investigate the effect that celastrol has on the stability and oligomerization of Hsp90α and on the binding of this chaperone to Tom70. This work provides a novel mechanism by which celastrol binds Hsp90α.

Study Of The Homogeneity Of Drug Loaded In Polymeric Films Using Near-infrared Chemical Imaging And Split-plot Design.

Split-plot design (SPD) and near-infrared chemical imaging were used to study the homogeneity of the drug paracetamol loaded in films and prepared from mixtures of the biocompatible polymers hydroxypropyl methylcellulose, polyvinylpyrrolidone, and polyethyleneglycol. The study was split into two parts: a partial least-squares (PLS) model was developed for a pixel-to-pixel quantification of the drug loaded into films. Afterwards, a SPD was developed to study the influence of the polymeric composition of films and the two process conditions related to their preparation (percentage of the drug in the formulations and curing temperature) on the homogeneity of the drug dispersed in the polymeric matrix. Chemical images of each formulation of the SPD were obtained by pixel-to-pixel predictions of the drug using the PLS model of the first part, and macropixel analyses were performed for each image to obtain the y-responses (homogeneity parameter). The design was modeled using PLS regression, allowing only the most relevant factors to remain in the final model. The interpretation of the SPD was enhanced by utilizing the orthogonal PLS algorithm, where the y-orthogonal variations in the design were separated from the y-correlated variation.

Doehlert Design-desirability Function Multi-criteria Optimal Separation Of 17 Phenolic Compounds From Extra-virgin Olive Oil By Capillary Zone Electrophoresis.

In Brazil, the consumption of extra-virgin olive oil (EVOO) is increasing annually, but there are no experimental studies concerning the phenolic compound contents of commercial EVOO. The aim of this work was to optimise the separation of 17 phenolic compounds already detected in EVOO. A Doehlert matrix experimental design was used, evaluating the effects of pH and electrolyte concentration. Resolution, runtime and migration time relative standard deviation values were evaluated. Derringer's desirability function was used to simultaneously optimise all 37 responses. The 17 peaks were separated in 19min using a fused-silica capillary (50μm internal diameter, 72cm of effective length) with an extended light path and 101.3mmolL(-1) of boric acid electrolyte (pH 9.15, 30kV). The method was validated and applied to 15 EVOO samples found in Brazilian supermarkets.

Lawsonia Inermis-mediated Synthesis Of Silver Nanoparticles: Activity Against Human Pathogenic Fungi And Bacteria With Special Reference To Formulation Of An Antimicrobial Nanogel.

Lawsonia inermis mediated synthesis of silver nanoparticles (Ag-NPs) and its efficacy against Candida albicans, Microsporum canis, Propioniabacterium acne and Trichophyton mentagrophytes is reported. A two-step mechanism has been proposed for bioreduction and formation of an intermediate complex leading to the synthesis of capped nanoparticles was developed. In addition, antimicrobial gel for M. canis and T. mentagrophytes was also formulated. Ag-NPs were synthesized by challenging the leaft extract of L. inermis with 1 mM AgNO₃. The Ag-NPs were characterized by Ultraviolet-Visible (UV-Vis) spectrophotometer and Fourier transform infrared spectroscopy (FTIR). Transmission electron microscopy (TEM), nanoparticle tracking and analysis sytem (NTA) and zeta potential was measured to detect the size of Ag-NPs. The antimicrobial activity of Ag-NPs was evaluated by disc diffusion method against the test organisms. Thus these Ag-NPs may prove as a better candidate drug due to their biogenic nature. Moreover, Ag-NPs may be an answer to the drug-resistant microorganisms.