Adda Avendaño
Due to the multiple benefits that can be obtained from drones, now these unmanned aerial vehicles (UAVs) are not only used as purely recreational devices but are also used in other professional fields for soil analysis, surveillance, forest fire detection, technical inspections in large infrastructures and support in the agricultural, mining or archaeological sectors, among many others.
"If one single drone can support various human tasks, the drone swarm can offer greater advantages for the benefit of society because it consists of a group of UAVs operating with an autonomous and intelligent coordinated flight to perform a joint mission," said Rodolfo Vera Amaro, professor, and researcher at the Telematics Engineering Academy of the Unidad Profesional Interdisciplinaria en Ingeniería y Tecnologías Avanzadas (UPIITA).
The Doctor in Advanced Technology from UPIITA stated that his research project began as a mathematical model to predict the random movement of endangered animals, in which he would use a drone to collect the information stored in the devices attached to it.
"During a research stay at Virginia Tech in the United States, I discovered the potential of my project; but with a coordinated, intelligent, and autonomous set of drones, and the multiple benefits it offers, the work is more efficient: energy would be better used, it would have more coverage and the collection of information would be much faster," he added.
Returning to UPIITA, Dr. Vera Amaro focused on the implementation of a drone swarm system with autonomous and coordinated flight for remote and web monitoring of endangered animals through a wireless sensor network or WSN, a project approved by the Secretariat of Research and Postgraduate Studies (SIP) of the Instituto Politécnico Nacional (IPN).
The Master of Science in Telecommunications Engineering from the Escuela Superior de Ingeniería Mecánica y Eléctrica (ESIME), Unit Zacatenco, pointed out that, unlike the group of exhibition drones that make figures in the sky by means of a pre-established programming, the swarm is endowed with certain intelligence, according to a bio-inspired algorithm in which each member acts together and in relation to a leader.
"You have a leader drone, and the other drones will be his followers, and the actions they perform will be done without the need for any control or pre-established trajectory, they will simply go where the leader goes to perform a common or general task, but at the same time each drone could perform a specific task, which will depend on the mission to be performed," he said.
He added that drones communicate by telemetry, and their GPS coordinates are shared by means of wireless transceivers, that is, very fast transmitter-receiver modules. Although the transceiver system used by the polytechnic professor is by means of radio frequency, it is possible to change it for other more efficient systems, such as Lora, Xbee, even with 4G/LTE/5G technology, less Wi-Fi or Bluetooth because they are not energy efficient and do not have great coverage in open spaces or in moving vehicles.
The Telecommunications Engineer from ESIME Zacatenco commented that the type of communication has to do with the needs of the application or the environment where it is located. "For example, if we are in an agricultural activity, the drone swarm would have to fly with local communication because in a field there is usually no Wi-Fi signal or 4G cellular coverage. But whatever the case, the communication system is Peer to Peer (P2P or peer-to-peer)."
He also said that to complete a mission, flight times can be reduced with a swarm of drones, because if it is a 20-kilometer sample collection, time and energy can be saved if 10 drones do it instead of one, which must be constantly recharged at a base station.
According to Valeria León Morales, a student in the last semester of Telematics Engineering at UPIITA, who is in charge of introducing artificial intelligence in Dr. Vera's project, unmanned aerial vehicles are normally programmed manually, but during the development of the research, the idea arose that drones could make intelligent decisions, that is, that their trajectory would not be predetermined by any programming.
"At the moment, the project is focused on three drones: a leader and two followers, to which an artificial intelligence (AI) algorithm has been implemented using LSTM (Long-Short Term Memory) recurrent neural networks, whose main characteristic is that they can remember previous states and predict what the leader's next step will be," he added.
To make the aerial vehicles make their own decisions in real-time and keep their spatial structure, it was necessary to feed the training algorithm with about 15,000 data to the LSTM neural networks and to use more than 5,000 different data to test the drones' performance. It should be clarified that the training is done in two simulation stages: first in Matlab and the second in Python, to finish with the implementation on real drones.
With the advice of professor and researcher Rodolfo Vera, three research lines are being carried out for the use that can be made of the drone swarm, completely built by the polytechnic scientist, one of them, the main one, is for the surveillance of endangered animals and their monitoring by sensor networks.
A second proposal is precision agriculture, in which a swarm can better distribute irrigation loads, or spread pesticides strategically, led by a guide drone, to optimize the time occupied by a single autonomous drone that must return every so often for irrigation material or battery, and which is currently already being used in the Mexican countryside.
A third line of research is being conducted by Telematics Engineering students Jesús Iván García Argüelles, Alejandro Alemán Pérez and Ángel Leonardo Montiel Cruz, who have focused on maintaining communication during natural disasters, such as earthquakes, floods or landslides.
"What we propose is to provide Wi-Fi coverage through the swarm of drones; for this, we are going to take it from a place where the signal is not down, and through a hopping model transfer the signal from the leader and vehicle to vehicle to the area that requires network coverage," said Alejandro Alemán.
Selección Gaceta Politécnica #163. 2023, June 30th. IPN Imagen Institucional: Read the full magazine in spanish here
INSTITUTO POLITÉCNICO NACIONAL
D.R. Instituto Politécnico Nacional (IPN). Av. Luis Enrique Erro S/N, Unidad Profesional Adolfo López Mateos, Zacatenco, Alcaldía Gustavo A. Madero, C.P. 07738, Ciudad de México. Conmutador: 55 57 29 60 00 / 55 57 29 63 00.
Esta página es una obra intelectual protegida por la Ley Federal del Derecho de Autor, puede ser reproducida con fines no lucrativos, siempre y cuando no se mutile, se cite la fuente completa y su dirección electrónica; su uso para otros fines, requiere autorización previa y por escrito de la Dirección General del Instituto.
