Mexican expert designs high-level service robots

The robots are able to identify people by memorizing their facial features in dynamic environments

Mexican expert designs high-level service robots
The UNAM expert spent a whole decade in his robotics project - Photo: Taken from DGCS UNAM's website
English 11/09/2020 15:50 Newsroom & Agencies Mexico City Actualizada 16:07
Guardando favorito...

Leer en español

An expert from the National Autonomous University of Mexico (UNAM) was able to make robots identify a person by memorizing their facial features and remember the environment where they saw them for the first time in order to track missing people or recognize criminals.

Arturo Rodríguez García, from the School of High Studies Aragón, has spent a whole decade with this project with the objective to make robots able to detect, remember, recognize, understand, and interact with people in order to perform everyday tasks, just as service robots do.

This kind of robot is different “from industrial ones, that is, the ones that are inside a Factory, such as robotic arms,” as explained by the expert, since the first ones are designed to perform activities in offices, homes, and hospitals.

The research was divided into three phases. The first one, and which earned him his master’s degree in Computer Science and Engineering (PCIC) was named “Golem,” and consisted of providing the robot with the capacity to identify people in a certain environment and choose one so that it walked behind them.

Recommended: Temi, the AI robot fighting COVID-19 in Mexico

During Golem-II, the second phase, Rodríguez García implemented an algorithm that helped robots to remember people and then describe them. “I’ve presented the results of the first stages of the project in RoboCup tournaments in Turkey, the Netherlands, Japan, and in different versions in Germany,” he added.

This opened the way to Golem-III, with which he concluded his Ph.D., where the robots were able to identify a person through the description of a verbal picture, considering specific characteristics like the color of a shirt or the kind of complexion.

“An automatic search system was developed and implemented based on descriptions in order to explore dynamic environments and find the objective. This can be useful in emergencies, such as finding a lost child or an alleged criminal,” said the UNAM professor.

For Rodríguez García, robotics is essential for the education of Computer Engineering, who he teaches, so he has created digital didactic resources to apply them on investigation issues related to artificial intelligence and robotics so as to encourage students to implement their own applications.

mp

Guardando favorito...
 

Noticias según tus intereses