Research Centre in Digitalization and Intelligent Robotics

Funder

Organizational Unit

Publications

Scale-up of a sorption process working with molecularly imprinted adsorbents for enrichment of winemaking residues and improvement of bioactivity

Publication . Gomes, Catarina; Duarte, Cristina N.; Martins, Cláudia D.; Amaral, Joana S.; Igrejas, Getúlio; Pereira, Maria João; Costa, Mário Rui; Dias, Rolando

This work presents the scale-up of a sorption process for the fractionation and enrichment of bioactive compounds in winemaking residues using molecularly imprinted adsorbents. The process works with hydroalcoholic solvents and the improvement of the bioactivity of the produced fractions, comparatively to the raw extracts, is demonstrated. The proposed approach was experimentally validated through the designing and running of a pilot size sorption prototype for the automation of the fractionation method. The synthesis of the molecularly imprinted adsorbents was made at the gram-scale and spent diatomaceous earth, used for wine filtration, was considered as a possible source of bioactive compounds in winemaking residues. The different fractions produced were evaluated for their antioxidant activity through three different assays, namely radical scavenging activity, reducing power and inhibition of lipid peroxidation. The results obtained show the improvement of the bioactivity of most of the fractions comparatively to the original diatomaceous earth extract. The most enriched fraction is estimated to have a total phenolic content c.a. 3.8 times higher than the original extract. The radical scavenging activity, the reducing power and the inhibition of lipid peroxidation for this fraction were measured to be 6.4, 4.2 and 4.5 times higher, respectively, than the initial diatomaceous earth extract. This work provides new insights on biomass valorisation and circular bioeconomy by combining in the same research materials development, process design and application to real extracts with proved improvement of the bioactivity of purified products.

2024Journal article

Open access

Application of Pattern Recognition Techniques for MathE Questions Difficulty Level Definition

Publication . Azevedo, Beatriz Flamia; Souza, Roberto Molina de; Pacheco, Maria F.; Fernandes, Florbela P.; Pereira, Ana I.

Active learning is a modern educational strategy that involves students in the learning process through diverse interactive and participatory activities. The MathE platform is an international online platform created to support students and lecturers in the Mathematics teaching and learning process. This platform offers a tool to aid and engage students, ensuring new and creative ways to encourage them to improve their mathematical skills. The study proposed in this paper refers to a comprehensive investigation of the patterns that may exist within the set of questions available on the MathE platform. The objective is to investigate how to evaluate the student’s opinions about the question’s difficulty levels based on the variables extracted from student answers collected through surveys applied among the platform’s users. Moreover, a comparative study between variables is performed using correlation and hypothesis tests. Furthermore, based on the results obtained for samples of different sizes, it was possible to define the most appropriate number of answers that should be considered to categorize the question’s difficulty level. The results demonstrated that the variables extracted could be used to carry out the question level, and 30 answers are the most appropriate number of questions that must be used to categorize the question level.

2024Conference object

Restricted access

A neural network approach in WSN real-time monitoring system to measure indoor air quality

Publication . Brito, Thadeu; Lima, José; Biondo, Elias; Nakano, Alberto Yoshiro; Pereira, Ana I.

Indoor Air Quality (IAQ) pertains to the air quality within a specific space and is directly linked to the well-being and comfort of its occupants. In line with this objective, this research presents a real-time system dedicated to monitoring and predicting IAQ, encompassing both thermal comfort and gas concentration. The system initiates with a data acquisition, wherein a set of sensors captures environmental parameters and transmits this data for storage in a database. The measured parameters are analyzed by a neural network algorithm that predicts anomalies based on historical data. The neural network model generated predictions from 75.9% to 98.1% (depending on the parameter) of precision during regular situations. After that, a test with smoke in the same place was done to validate the model, and the results showed it could detect anomalies. Finally, prediction data are stored in a new database and displayed on a dashboard for monitoring in real-time measured and prediction data.

2023Conference object

Restricted access

Impact of EMG Signal Filters on Machine Learning Model Training: A Comparison with Clustering on Raw Signal

Publication . Barbosa, Ana Carolina; Ferreira, Edilson Santos; Grilo, Vinicius F.S.B.; Mattos, Laercio; Lima, José

Our current society faces challenges in integrating individuals with disabilities, making this process difficult and painful. People with disabilities (PwD) are often mistakenly considered incapable due to the difficulties they face in daily tasks due to the lack of adapted means and tools. In this context, assistive technologies play a crucial role in improving the quality of life for these individuals. However, assistive technologies still have various limitations, making research in this area essential to enhance existing solutions and develop new approaches that meet individual needs, aiming to promote inclusion and equal opportunities. This paper presents a research project that focuses on the study of electromyography (EMG) signal processing generated by individuals who have undergone amputations. These signals are essential in assistive technologies, such as myoelectric prostheses. The study focuses on the impact of different filters and machine learning training methods on this processing. The results of this study have the potential to provide relevant findings for the development of more efficient assistive technologies. By understanding the processing of EMG signals and applying machine learning techniques, it is possible to improve the accuracy and response speed of prosthetics, increasing the functionality and naturalness of movements performed by users, as well as paving the way for the emergence of new technologies.

2024Conference object

Restricted access

The impact of educational robots as learning tools in specific technical classes in undergraduate education

Publication . Santos, Tatiana M.B.; Amorim, Johann S.J.C.C.; Carneiro, Mirella M. de O.; Campos, Victor F.; Manhães, Aline; Lima, José; Pinto, Milena F.

The use of mobile robots in the classroom has gained increasing attention in recent years due to their potential to enhance student engagement and facilitate personalized learning. This research presents the insertion of mobile robots as a hands-on learning experience in Control and Servomechanisms II and Signal Processing II classes. This work also addresses the challenges and limitations of using mobile robots in the classroom, including technical difficulties. The students were evaluated during the code implementation in the practical exercises. Besides, a form was provided to them in order to assess the impact of these robots as part of the pedagogical practice. From the students’ positive feedback, it was possible to conclude that the mobile robots were well-accepted. Besides, the robots enhanced Control Systems classes and improved students’ learning outcomes.

2023Conference object

Open access