Gabriel Guerra
Electrical and Computer EngineeringPhD student in Electrical Engineering at the University of Saskatchewan (2024-present), supervised by Prof. Seok-Bum Ko. His research applies deep learning to natural language processing, particularly named-entity recognition over scientific text, to surface electromyography (sEMG) signals, and to AI for industry including predictive maintenance. He came to the Ko Lab from robotics, with published work on multi-robot systems, trajectory planning over IoT architectures, and robotics platforms for teaching.
Background and Publications
Education
- PhD, Electrical Engineering, University of Saskatchewan, 2024-present. Deep learning for surface electromyography and natural language processing; transformer and recurrent architectures. Supervised by Prof. Seok-Bum Ko.
- MSc, Smart Systems (Computer Engineering), Universidad de Salamanca, Spain, 2023. Robotics, artificial intelligence and data mining.
- BSc, Mechatronics Engineering, Universidad de las Fuerzas Armadas ESPE, Ecuador, 2021. Mechanical design, electronics and software development.
Publications
UV-SAFE: automated UV disinfection of market products
A. Erazo, L. Allauca, G. Guerra, S.-B. Ko. IEEE Seventh Ecuador Technical Chapters Meeting (ETCM), 2023.
- An automated device that disinfects market products with ultraviolet light
- Experiments on regular and irregular objects fixed the optimal placement of the UV lamps inside the chamber
- Ten seconds per product, reaching a log-reduction dose
Multi-robot system for collaborative work with trajectory planning over IoT
G. Guerra, L. Allauca, L. Escobar, X. Sanchez-Sanchez, S. Puma-Araujo, R. Ramirez-Mendoza. Advances in Automation and Robotics Research, Springer, 2021.
- Omnidirectional robots on Mecanum wheels, built in modular layers for sensors, power and control
- A central broker coordinates the robots and corrects their actions across an IoT architecture
- Overhead computer vision detects and classifies the workspace, feeding trajectory correction
- Developed entirely with open-access software, following the VDI 2206 mechatronic design standard
Multi-robot platform with cyber-physical features for education
L. Escobar, C. Moyano, G. Aguirre, G. Guerra, L. Allauca, D. Loza. IEEE ANDESCON, 2020.
- A flexible, scalable and easily replicable omnidirectional mobile robot
- Several robots form an open-source platform managed as a cyber-physical system over MQTT
- ROS on a Raspberry Pi controls each robot; an Arduino handles sensors and actuators
- Built to teach the principles of cyber-physical systems
6-DOF anthropomorphic robot for remote robotics teaching
G. Guerra et al. Advances in Science, Technology and Engineering Systems Journal, 2021.
Coating 3D-printed gyroid structures with dragon's blood extract
L. Allauca, G. Guerra, H. Lara. IEEE Fourth Ecuador Technical Chapters Meeting (ETCM), 2019.
- Two coating methods compared on 3D-printed gyroid lattices: direct perfusion and dip coating
- Coated with the Amazonian herbal remedy sangre de drago (dragon's blood)
- Image analysis found a non-linear relationship between porosity and coated area, peaking at 70% porosity for perfusion and 90% for dip coating
- A proof of concept aimed at tissue regeneration and defect repair
