Karen Torres-Medina

Karen Torres-Medina

she/her

A versatile and proactive Mechatronics Engineer with expertise in AI, robotics, TinyML, and IoT. Passionate about creating impactful solutions through innovation and technology.

Karen Torres-MedinaRobotics, AI, TinyML, and Autonomous Vehicles

Welcome to the personal website of Karen Torres-Medina, a Mechatronics Engineer, AI Researcher, and Robotics Enthusiast with a passion for advancing technology through autonomous systems, TinyML, and IoT. Currently, I am working as a Research Technician and Researcher at the Smart Environments Technologies Lab (TAI) at ESPOL, where I focus on cutting-edge research in TinyML, robotics, and autonomous vehicles.

This site showcases my academic journey, professional achievements, and ongoing research in robotics and artificial intelligence. Through projects involving unmanned autonomous vehicles, edge devices, and machine learning, I aim to contribute to the development of smarter, more efficient systems that can revolutionize industries such as healthcare, transportation, and robotics.

My Research Focus

I am currently developing TinyML-based solutions for embedded systems, working on autonomous vehicles, and exploring the integration of AI and robotics for real-world applications. Some of the key areas of my research include:

  • Unmanned Autonomous Vehicles: Developing systems for obstacle detection, SLAM (Simultaneous Localization and Mapping), and autonomous navigation.
  • TinyML: Implementing machine learning models directly on low-power, resource-constrained edge devices for real-time processing.
  • IoT: Exploring Internet of Things (IoT) solutions for smart environments and sustainable technology.
  • AI in Robotics: Utilizing AI to improve the functionality and autonomy of robotic systems, particularly in applications such as prosthetics, robotics in healthcare, and autonomous vehicles.

Teaching & Mentoring

In addition to my research, I have had the opportunity to teach robotics to high school and university students. My classes have been centered around engaging students in hands-on projects using platforms like STEAM and introducing them to foundational topics in AI, machine learning, and robotics. I believe in fostering creativity and critical thinking through practical exercises that bridge the gap between theory and real-world applications.

Research Experience

Research Technician and Researcher at TAI - ESPOL

Guayaquil, Ecuador | 2024 – Present

Research Technician and Researcher at TAI - ESPOL

Guayaquil, Ecuador | Sept 2024 – Present
At TAI - ESPOL, I focus on the development of prototypes and software for advanced research projects in the Smart Environments Technologies Lab. My work is heavily centered around the hardware design and development of printed circuit boards (PCBs) and embedded systems. I am responsible for designing and testing hardware components, ensuring that they integrate seamlessly into the overall research systems.

I also oversee the operation and maintenance of hardware prototyping equipment, ensuring that all tools and devices are in optimal working condition to support ongoing experimental work. This is crucial for maintaining the accuracy and reliability of the systems we develop.

In addition, I create datasets and develop applications using cutting-edge artificial intelligence (AI) and TinyML techniques. These solutions are applied to machine learning models that are used for real-time data processing and decision-making in IoT systems. I work closely with my team to design intelligent systems capable of adapting to dynamic environments.

I assist in preparing research project proposals and conduct literature reviews, helping to shape the direction of research within the group. This includes identifying gaps in existing knowledge and proposing innovative solutions to address those challenges.

As part of my role, I also direct projects and supervise prototyping within a multidisciplinary team. This involves coordinating with various specialists, ensuring that project milestones are met, and that all work is aligned with the overall research objectives.

Finally, I actively contribute to the academic side of our projects by authoring and publishing scientific articles. I assist in the preparation and submission of papers to various conferences and journals, helping to communicate our research findings to the wider scientific community.

Research Assistant (Ad Honorem) - ESPOL

Guayaquil, Ecuador | 2021 – 2023
As an ad honorem research assistant at ESPOL, I participated in several high-impact research projects focused on robotics and autonomous navigation. My work involved developing intelligent control strategies for autonomous sailboats, incorporating TensorFlow, YOLO, and SLAM to enable the boat to autonomously navigate and avoid obstacles. I also worked on preparing research papers, contributing to the IEEE Ibero-American Congress of Mechanical Engineering where I presented a paper on autonomous sailboat navigation. This experience helped me refine my skills in AI, machine learning, and robotics, while also improving my ability to communicate complex technical concepts through research publications and conference presentations.

Mechatronic Design Assistant - Robotilsa S.A.

Guayaquil, Ecuador | Feb 2022 – Jul 2022
At Robotilsa S.A., I participated in mechatronic design projects where I collaborated in the development and testing of robotic systems for various industrial applications. My work focused on system integration and mechanical design, applying my skills in AI to optimize performance and enhance the functionality of robotic systems. I also worked on creating Printed Circuit Boards (PCBs) for robotic systems and flexible manufacturing systems, improving the efficiency and reliability of automated production lines.

Key Projects and Achievements

  • Autonomous Sailboat Navigation: Developing a system for unmanned autonomous vehicles to navigate autonomously using SLAM and obstacle detection for sailboats using AI, SLAM, and YOLO for obstacle detection and autonomous movement.

  • Robotic Prosthesis Development: Designed and implemented a robotic prosthesis using myoelectric sensors and oximetry, applying AI to interpret sensor data and control the prosthetic movement. This work was part of my thesis, where I utilized reinforcement learning to optimize the control of the prosthesis.

  • AI-Based Pneumonia Detection System: Developed a CNN-based system for pneumonia detection, applying deep learning techniques to analyze medical data and provide accurate diagnosis.

Thesis: Robotic Prosthesis with EMG and Oximetry Sensors

For my thesis, I developed a robotic prosthesis controlled by EMG signals (electromyographic signals) and oximetry sensors. The challenge was to train a reinforcement learning agent to interpret these signals and generate precise control commands for the prosthesis. The key steps included:

  • Data collection: EMG signals were gathered and paired with the intended prosthetic movements.
  • Signal processing: The raw EMG signals were filtered, normalized, and segmented into time windows.
  • Reinforcement learning: Defined the learning environment where the agent (neural network) interacted with the prosthesis, using reward functions to guide learning.
  • Evaluation: The agent was trained using PPO, TRPO, and SAC algorithms, achieving real-time control and precise manipulation of the prosthetic.

These projects, combined with my academic background and hands-on experience, have shaped my expertise in robotics, AI, and TinyML, allowing me to contribute to innovative solutions in autonomous systems, prosthetics, and AI-driven technologies.

Future Goals

I am deeply interested in continuing to push the boundaries of TinyML and AI in embedded systems. My future projects will focus on integrating machine learning into small, efficient devices for a variety of applications, from robotics to healthcare solutions. I am also committed to advancing the development of autonomous systems.