This research to practice paper describes two NSF-funded projects: The “Program for Engineering Access, Retention, and LIATS Success“ (PEARLS) and the “Academic and Socioemotional Support Ecosystem for Talented low-income STEM Students“ (ASSETS). These initiatives aim to support low-income, academically talented students pursuing STEM degrees, offering academic, socio-emotional, and financial assistance. Both projects provide mentoring, workshops, mental health services, and scholarships. Outcomes have shown increased engagement, retention, and graduation rates among participants compared to non-participants. Retention and persistence rates among PEARLS participants reached 97% and 96.3% respectively, showcasing the model effectiveness in keeping students focused on their studies. Graduation rates saw remarkable improvements, with on-time graduation rates soaring to 4.17 times that of non-participating students. The paper presents a robust framework for an integrated support ecosystem, emphasizing scalability and replicability across diverse institutions. By addressing the challenges faced by underprivileged students during environmental crises, this framework aims to foster equity and inclusion in higher education, ensuring all students have the support they need to succeed.

This paper analyzes, from a servingness perspective, the experiences of students in a Hispanic Serving Institution (HSI) implementing personalized mentoring strategies as part of an NSF-funded, S-STEM program. The program objective was to increase the retention and success of low-income academically talented students. We discuss the results of using Individual Development Plans (IDP) to help students establish post-graduation goals and action plans to reach them. The IDPs also maintained a flexible structure that allowed students for morphing both their plans and envisioned outcomes, as they progressed in their chosen academic programs. Student outcome expectations were assessed at three stages of their development: in their first year, as juniors in what can be considered a mid-point of their development, and during their fourth year of program participation.

The prevalent gap between students from different socioeconomic statuses (SES) affects multiple aspects of our social functionality. Among its effects, the SES gap directly impacts the opportunities to which students have access while in college and beyond. This paper discusses the efforts and partial results of a program aimed at bridging such differences for low-income, academically talented students (LIATS) in a Hispanic Serving Institution (HSI). The reported approach leverages the relations with a group of industry partners to provide LIATS with scholarships, professional mentorship, work-shops, and on-the-job training opportunities while sharing with the in-dustry the students' e-portfolios, professional profiles, and resumes, developing a symbiotic relationship where both benefit. After three years of interactions, the results show how these opportunities helped LIATS develop their skills, leadership, and competitiveness as future STEM professionals. The experience also demonstrates that profes-sional growth opportunities are critical for engaging LIATS in real-life contexts where they collaborate and interact with industry part-ners, and for providing them with opportunities that help to bridge the SESgap.

The rapid progress of modern AI tools for automatic speech recognition and machine translation is leading to a progressive cost reduction to produce publishable subtitles for educational videos in multiple languages. Similarly, text-to-speech technology is experiencing large improvements in terms of quality, flexibility and capabilities. In particular, state-of-the-art systems are now capable of seamlessly dealing with multiple languages and speakers in an integrated manner, thus enabling lecturer’s voice cloning in languages she/he might not even speak. This work is to report the experience gained on using such systems at the Universitat Politècnica de València (UPV), mainly as a guidance for other educational organizations willing to conduct similar studies. It builds on previous work on the UPV’s main repository of educational videos, MediaUPV, to produce multilingual subtitles at scale and low cost. Here, a detailed account is given on how this work has been extended to also allow for massive machine dubbing of MediaUPV. This includes collecting 59 h of clean speech data from UPV’s academic staff, and extending our production pipeline of subtitles with a state-of-the-art multilingual and multi-speaker text-to-speech system trained from the collected data. Our main result comes from an extensive, subjective evaluation of this system by lecturers contributing to data collection. In brief, it is shown that text-to-speech technology is not only mature enough for its application to MediaUPV, but also needed as soon as possible by students to improve its accessibility and bridge language barriers.

This Research to Practice Full Paper presents how modular design techniques aided by an Outcome-Based educational framework, can be incorporated in an Embedded Systems Design laboratory to improve student learning. Teaching embedded systems design concepts and enhancing students' skills in this area are important tasks for universities in order to provide an up to date education. To achieve this, the laboratory objective, content, pedagogical methods, and assessment activities were aligned using an Outcome-Based educational framework to ensure proper student learning. The modular approach was applied to pedagogical methods through the design of a set of progressive laboratory experiments and six electronic educational modules. Using this approach effective laboratory experiments that promoted better student learning, in the area of embedded systems design, were developed. As a result, the overall laboratory student performance was improved and therefore the proposed methodology was validated.

This paper presents an evaluation of the new JEDEC standard JEP173. The JEP173 establishes a characterization procedure to reliably assess the dynamic ON-resistance of GaN lateral power transistors. DC and pulsed measurement setups were developed to evaluate the proposed methods for hard and soft switching conditions. Several devices were tested under a wide range of test conditions. We found that the proposed procedures in JEP173 allow for accurate acquisition of the ON-resistance under several test conditions. However our experiment found the standard, in its current version, does not account for the stress voltage and temperature effects on the dynamic RDSON, leading to under-characterization of parts.

Citizen science is becoming mainstream in a wide variety of real-world applications in astronomy or bioinformatics, in which, for example, classification tasks by experts are very time consuming. These projects engage amateur volunteers that are tasked to manually classify unannotated examples. As a result, we obtain a larger volume of labelled data that, however, contains a great level of uncertainty due to the wide range of expertise of the volunteers. Handling that inherent uncertainty is key to building robust and fast machine learning models that maximise the outcome of citizen science projects. In this work, we introduce a preliminary approach that first transforms the original results from a citizen science project to handle the uncertainty, and then uses this as input to a fuzzy k-nearest neighbour classifier. We leverage citizen science results in such a way that it naturally speeds up the learning and classification phases of the fuzzy classifier, and improves the classification performance. As a case study, we will focus on the Galaxy Zoo project that consisted of galaxy image classification. Our experimental results show that an appropriate use of citizen science data enables a faster and more robust classification using the fuzzy k-nearest neighbour classifier.

Citizen Science is coming to the forefront of scientific research as a valuable method for large-scale processing of data. New technologies in fields such as astronomy or bio-sciences generate tons of data, for which a thorough expert analysis is no longer feasible. In contrast, communities of volunteers coordinated by the Internet are showing a great potential in completing such analysis in a reasonable time. However, this approach brings uncertainty and the spread of biases within the data, since amateur participants are usually non-experts on the subject and count with variable skills and expertise. This means lack of accuracy in results coming from Citizen Science projects. This work presents a novel approach to handle uncertainty in Citizen Science. We focus on leveraging this uncertainty in the data pursuing a refinement of results. We distinguish between two types of uncertainty: a first one due to the lack of consensus between amateurs, and another one quantified by amateurs themselves during the course of the project. We test our method using the Galaxy Zoo, a project which aims for the labelling of a huge dataset of galaxy images. Considering available expert classifications to validate our experiments, the proposed method is able to improve current accuracy and classify a greater number of images.

Impacting college readiness and reinforcing recruitment and retention methods are objectives shared by virtually all engineering schools. This paper describes a dual enrollment program for high-school students that supports a recruitment, retention, and college readiness improvement plan initiated by the UPRM College of Engineering. Elements of synchronicity, hybrid populations, and assisted teaching are combined with a distance learning modality to define the backbone of this initiative. A small sample provides preliminary statistics on the potential results of this model, while yielding valuable feedback and insight for driving its scaled implementation.

This paper describes a project experience in a microprocessor interfacing course, where computer engineering (CE) and electrical engineering (EE) students were joined to develop a project designing and implementing a Digital Controller for a Three Degree of Freedom Helicopter (3DOFH). This is a highly non-linear problem brought in by the EE students from the Process Instrumentation and Control Laboratory (PICL). Besides serving as the course project for the students, the motivation for taking such a project was to create a base platform using embedded microprocessors where control students could acquire signal conditioning and embedded software design skills in a more realistic platform than that provided by virtual instrument environments. This paper describes the course setting, design approach, and experience gained by the students, establishing a collaboration modality that could be emulated to bring multidisciplinary projects into traditional courses.

Online video lecture repositories are rapidly growing and becoming established as fundamental knowledge assets. However, most lectures are neither transcribed nor translated because of the lack of cost-effective solutions that can give accurate enough results. In this paper, we describe a system architecture that supports the cost-effective transcription and translation of large video lecture repositories. This architecture has been adopted in the EU project transLectures and is now being tested on a repository of more than 9000 video lectures at the Universitat Politecnica de Valencia. Following a brief description of this repository and of the transLectures project, we describe the proposed system architecture in detail. We also report empirical results on the quality of the transcriptions and translations currently being maintained and steadily improved.

This paper discusses the tradeoffs of using two different approaches for designing and implementing digital control systems for electrical engineering course projects. The usage of an off-the-shelf microcontroller unit (MCU) is explored and contrasted to the more traditional approach based on personal computers (PCs), data acquisition cards (DaqCs), and programs like RTW, LabView, and Matlab/Simulink. We discuss how an MCU-based approach was developed and tested in the design and implementation of a two-loop digital feedback control system for a Ball-and-Beam mechanical system. Results show that by using an MCU it is possible to achieve a performance comparable to that of using PC's and data acquisition cards with the added advantage of exposing students to an experience closer to the actual way most digital control systems are implemented in practice.

The achievement of effective implementations to multi-FPGA architectures is greatly dependent on the process of partitioning. Although several automated high-level partitioning (HLP) methods have been reported (Srinivasan, et al., 2001) most of them are designed to solve general partitioning problems, and tend to apply generic local optimization techniques that miss out on alternate formulations that become apparent only with knowledge of the algorithm's functionality. The algorithmic formulation of discrete signal transforms (DST) especially that of the DFT has been extensively studied. Automated computational algebra platforms for the algorithmic manipulation of fast transform algorithms have been proposed, as well as automated methods to optimize DST implementations to general purpose processor platforms (Puschel et al., 2001) However, these methods have yet to be successfully adapted to automated partitioning methodologies for dedicated distributed hardware platforms

This paper introduces a high-level partitioning methodology which uses formulation-level discrete signal transform properties to provide improved results for their partitioning to multi-FPGA architectures. We review the global optimization scheme, the various methodology processes, and explain how their designs were influenced by characteristics of the discrete signal transforms and the target architecture. To illustrate our methodology's solution quality, we present results of partitioning several FFT sizes to a Berkeley emulation engine 2 multi-FPGA module.

No abstract available.