New generation of interns: engineering students share their research experiences after their stay at Inria Chile

After two months, the 14 students from six Chilean universities, selected at the end of 2024 to participate in the Summer Internship Program at Inria Chile, presented the results they achieved and the key takeaways from their experience. During this period, the students were integrated into five of the institute's projects alongside the team: OcéanIA, CIEN, AI-Powered Modeling of Stellar Atmospheres, EMISTRAL, and Vendimia 5.0.

This edition of the Summer Internship Program included students from universities in Santiago and other Chilean regions: Universidad de Chile, Pontificia Universidad Católica de Chile, Universidad Técnica Federico Santa María, Universidad Austral de Chile, Universidad de Valparaíso, and Pontificia Universidad Católica de Valparaíso.

During the event organized on February 27 and 28 at Inria Chile, the students had the opportunity to present their work to their supervisors, researchers, and collaborators. The presentations covered the research areas they participated in, as well as the results and challenges faced during their internships.

For Nayat Sánchez Pi, Director of Inria Chile, "The Summer Internship Program at Inria Chile is a program we have been organizing for several years now, and it has become increasingly successful, with applications coming from many universities across Chile. At Inria Chile, we aim to train the future talents in digital sciences and artificial intelligence, and we firmly believe that one of the best ways to achieve this is by involving young students in science and research projects of excellence."

Luis Martí, Scientific Director of Inria in Chile, added, "Our interns have had the opportunity to understand how we work in a multidisciplinary environment. This exposes students to areas outside their particular expertise or interests, enriching them through the various projects we develop at Inria Chile."

This year, out of more than 100 applications received, 14 students were selected to study a specific topic in one of our projects:

• In the project AI-Powered Modeling of Stellar Atmospheres:

  • Sol Covacich, Civil Engineering in Computation student, School of Engineering, Pontificia Universidad Católica de Chile, “Machine Learning in Astronomy: Development of a Machine Learning Model for Calculating Opacity and Radiative Pressure in Stellar Atmospheres”

  • Luis Llanca, Civil Engineering in Informatics student, Universidad Austral de Chile, “Machine Learning in Astronomy: Development of a Machine Learning Model for Calculating Opacity and Radiative Pressure in Stellar Atmospheres”

• In the project CIEn - Collaboration Inria Environment:

  • Alex Garnica, Civil Informatics Engineering student, Universidad Austral de Chile, selected for “Template for Collaboration Visualization Platforms in the CIEN Project”

• In the project EMISTRAL: An Environmental Monitoring and Inspection Sailboat via Transfer, Reinforcement, and Autonomous Learning:

  • Moisés Donoso, Electrical Engineering student, School of Engineering, Pontificia Universidad Católica de Chile, “Autonomous Control of a Robotic Sailboat Using Reinforcement Learning”

• In the project OcéanIA: AI, the Ocean, and Climate Change:

  • Juan Pablo Barías, Civil Engineering student, School of Engineering, Pontificia Universidad Católica de Chile, “Improvement of the Repository for Creating Computer Vision Models for Plankton Image Classification”

  • Emilio Díaz, Civil Engineering in Computation student, Faculty of Physical and Mathematical Sciences, Universidad de Chile, “Multi-modal Models (metadata-image) for Plankton Image Classification”

  • Cristóbal Godoy, Civil Mathematical Engineering student, Universidad de Chile, “Modeling Navier-Stokes Equations Using Physics-based Deep Learning in the OcéanIA Project”

  • Lilly Guevara, Civil Informatics Engineering student, Universidad Técnica Federico Santa María, selected for the application of pre-trained LLMs with genomes to decipher the role of the marine microbiome in carbon capture in the context of climate change

  • Julio Jofré, Civil Electrical Engineering student, Faculty of Physical and Mathematical Sciences, Universidad de Chile, “Out-of-distribution Computer Vision Models for Plankton Image Classification”

  • Camila Reyes, Civil Engineering student, Faculty of Physical and Mathematical Sciences, Universidad de Chile, “Application of Pre-trained LLMs with Proteomes to Elucidate the Role of the Marine Microbiome in Carbon Capture in the Context of Climate Change”

  • Ricardo Soto, Mathematical Engineering student, Pontificia Universidad Católica de Chile, “Study of Uncertainty Quantification Methods for Simulating PDEs Using Deep Learning” 

• In the project Vendimia 5.0: Human-Machine Planning to Optimize Grape Harvesting Decisions and Winemaking Operations:

  • Patricio Figueroa, Civil Informatics Engineering student, Pontificia Universidad Católica de Valparaíso, “Use of Sentinel 2 Satellite Images for Precision Agriculture-Oriented ML Models”

  • Diego Mena, Civil Electrical Engineering student, Universidad de Chile, “Development of Computer Vision Models for Detecting the Ripening State of Vines”

  • Brandon Torres, Civil Informatics Engineering student, Universidad de Valparaíso, “Optimization and Performance Analysis of Artificial Intelligence Models Using Profilers”

Learn about the testimonials of some of the interns here:

“I had the opportunity to implement a technique to save computational resources in the training of large models. It was a significant process, as I worked alongside other interns on the same tool, with the constant support of my supervisor, who was always there to help us. I learned the importance of teamwork, even in academia, and that even when you don't know how to achieve a result, you have to trust in your ability to succeed.”

Juan Pablo Barías, School of Engineering, Pontificia Universidad Católica de Chile, intern in the OcéanIA project.

“I was part of the OcéanIA project, which aims to understand the role of the marine microbiome in absorbing carbon dioxide. My role was to assign functions to some genomes whose functions were unknown. My experience was very positive; there was always a supportive environment, both from my supervisor and my colleagues. I wasn't used to conducting experiments and getting poor results, so I learned how to deal with that frustration and try again or find another strategy. It really opened my mind.”

Lilly Guevara, Universidad Técnica Federico Santa María, intern in the OcéanIA project.

“I was involved in the project evaluating distance metrics for comparing phylogenetic trees. It was a very rewarding experience that provided me with a lot of knowledge, not just theoretical, but also in handling tools for collaborative work. I was able to delve into research methods and become resilient to certain obstacles we often encounter, and when we manage to overcome them, it is very gratifying. I want to apply this new knowledge in my postgraduate studies, and this research environment has certainly prepared me for that.”

Luis Llanca, Faculty of Engineering Sciences, Universidad Austral de Chile, intern in the AI-Powered Modeling of Stellar Atmospheres project.

“My project focused on optimizing the analysis of artificial intelligence models using profilers. I concentrated on identifying the inefficiencies in AI models and generating optimizations for them. The experience was very rewarding, as the mentors were constantly supportive and provided potential solutions. Additionally, we worked with an agile methodology that allowed us to progress very efficiently. This experience helped me understand the importance of teamwork.”

Brandon Torres, Faculty of Engineering, School of Computer Engineering, Universidad de Valparaíso, intern in the Vendimia 5.0 project

“I was working and researching the use of Sentinel 2 satellite images for Machine Learning projects. We collected satellite data and conducted analyses of vegetation indexes to understand the characteristics of the population. From there, we developed ML models to predict or analyze the various data obtained. I sincerely believe that this experience was very enriching as it was my first internship, and it helped me understand what it means to work in a research-focused institute. I found the methodology to be very effective, learning how to work in this type of organization, how to structure tasks, reduce uncertainty, and achieve results.”

Patricio Figueroa, Faculty of Engineering, School of Computer Engineering, Pontificia Universidad Católica de Valparaíso, intern in the Vendimia 5.0 project

“I was developing the foundations for training computer vision models capable of identifying the ripeness of grapes. I acquired new skills that can be applied in both research and practical contexts. I learned how to conduct thorough research for the task at hand, as opposed to relying on the first search engine result. In this field, it's crucial to conduct exhaustive research to identify the problems you want to solve. I plan to apply what I've learned to my thesis project, and in the long term, I want to use these skills and tools in the various jobs I undertake.”

Diego Mena, Faculty of Physical and Mathematical Sciences, Universidad de Chile, intern in the Vendimia 5.0 project