Lab Updates

Excited to welcome a new semester with both familiar faces and new members joining the team. Together, we’re diving into projects that span from low-rank approximations of convolutional layers to automating knowledge distillation and quantization across entire model libraries, and even exploring applications like portable orchid identification and mini load distribution on NVIDIA Jetson devices.

Proud of the passion and creativity this group brings every semester—looking forward to seeing how far we can push the boundaries of Edge Computing this time around!

April 14, 2025 — A huge thank you to Manuel Ortiz Bey, Senior Solutions Architect at Amazon Web Services (AWS), for taking the time to speak to our Cloud Computing Infrastructure class!

Manuel has always been incredibly generous in sharing his expertise, and his talk provided students with valuable real-world insights into cloud architecture and industry best practices. His continued support of our university and courses is truly appreciated.

It’s always great to have professionals like Manuel helping bridge the gap between academia and industry. Looking forward to more collaborations in the future!

Edge tags: #CloudComputing #AWS #Education #IndustryCollaboration #ThankYou

March 17–19, 2025 — Last week, we had a proud moment for the Edge Computing Group at University of Puerto Rico–Mayagüez! 🎉

At ISICN 2025, our undergraduate researchers showcased their work on Machine Learning Model Compression across multiple healthcare applications.

• Luis Fernandez presented on Model Compression for Wearable Devices in Skin Cancer Screening.
• JACOB M DELGADO-LOPEZ discussed Computer Vision for Real-Time Monkeypox Detection on Embedded Systems.
• Sebastián A. Cruz-Romero analyzed Post-Training Quantization for CNN-Based Conjunctival Pallor Anemia Detection.

This was a fantastic experience for our team! A huge congratulations to our students for their hard work and dedication. The future of Edge AI is bright!

Edge tags: #EdgeComputing #MachineLearning #AI #ModelCompression #ISICN2025

March 2025 — 🌱 Bringing Technology to Agriculture: Supporting Sigatoka Disease Research 🌱

This week, our students supported ongoing research led by the Agricultural Extension Service and Professor Wanda Almodóvar on early detection of Sigatoka disease in plantains. 🍌🔬

Their work focuses on understanding and mitigating the disease’s impact on Puerto Rican farmers. Our role is to explore how spectral imaging and edge computing can assist in developing a low-power portable device to detect the disease before it becomes visible to the human eye. This could provide farmers with an early warning system, helping them take action before their crops are affected.

Team:

• Led by Phytopathology Professor Wanda Almodóvar
• Agricultural Extension Service: Vilmaris Bracero, Bryan Hernandez
• Microbiology Student: Wesley Mercado
• Edge Computing and Model Compression: CSE Students Luis Fernandez and Eithan M. Capella Muñiz

A special thank you to Nelson Tubens, the owner of the farm, for allowing us to conduct research on-site. This collaboration is a perfect example of how technology and agriculture can work together to solve real-world challenges.

We’re excited to continue supporting this and other efforts and look forward to seeing how technology can enhance agricultural resilience in Puerto Rico.

Collaboration across fields, impact across communities…

Edge tags: #Research #AgriculturalTechnology #EdgeComputing #AIforGood #PuertoRico

March 3–7, 2025 — We’re proud to share that one of our undergraduate researchers presented at the Society for Industrial and Applied Mathematics (SIAM) Conference on Computational Science & Engineering 2025 in Fort Worth, TX.

As part of the Building Engagement Program, our student received support to present the work “Evaluating Performance of Neural Network Quantization for Conjunctival Pallor Anemia Detection.” This research advances our lab’s mission of bringing effective AI to low-resource settings by making deep learning models lighter and more efficient while maintaining diagnostic accuracy.

We extend our sincere thanks to the Sustainable Horizons Institute for sponsoring this participation, and to mentor and Guided Affinity Group Lead Malena Español and the Inverse Problems & Applications Group for their tremendous guidance and support.

To everyone who engaged with our student at the conference—thank you for your thoughtful discussions, insightful feedback, and new collaborations. We look forward to continuing to push the boundaries of Edge AI for healthcare.

Edge tags: #ComputationalScience #SIAMCSE2025 #Quantization #EdgeAI #HealthcareAI

We’re excited to announce our Spring 2025 lab team! This semester brings together passionate students from various engineering disciplines to work on cutting-edge computing solutions. Our new members will be contributing to projects in biomedical engineering, agricultural sciences, and transportation management.

The team includes both undergraduate and graduate students who will be working on innovative edge computing applications. Each member brings unique perspectives and skills that will help us push the boundaries of what’s possible with edge computing technology.

Stay tuned for updates on their research projects and achievements throughout the semester!

Andrea’s groundbreaking research focuses on developing edge computing solutions for disaster resilience and emergency response systems. Her work addresses critical challenges in communication and data processing during natural disasters when traditional infrastructure may be compromised.

The research explores how edge computing can provide reliable, low-latency communication and data processing capabilities in disaster scenarios. This includes developing robust algorithms for resource allocation, real-time data analysis, and coordination between emergency response teams.

This work has significant implications for improving disaster response effectiveness and could save lives in emergency situations.

Sebastian presented an insightful overview of Apple’s CoreML framework and its applications in edge computing. The presentation covered the fundamentals of CoreML, including model conversion, optimization techniques, and deployment strategies for iOS devices.

The talk highlighted how CoreML enables on-device machine learning inference, reducing latency and improving privacy by keeping data processing local. Sebastian demonstrated practical examples of deploying custom models for real-time image recognition and natural language processing applications.

This presentation opened up new possibilities for our lab’s research in mobile edge computing and privacy-preserving AI applications.

Jacob delivered an excellent lecture series on TensorRT optimization for edge computing applications. The series covered fundamental concepts of deep learning model optimization, including quantization, pruning, and model compression techniques specifically designed for resource-constrained edge devices.

Students learned practical techniques for deploying neural networks on edge devices while maintaining performance and accuracy. The hands-on sessions included real-world examples of optimizing models for medical imaging, agricultural monitoring, and transportation applications.

This lecture series has become a cornerstone of our edge computing curriculum, providing students with essential skills for deploying AI models in edge environments.